Assessing biodiversity in farming landscapes:

a cross-taxonomic approach to conservation planning

Ding Li, Yong

A thesis submitted for the degree of

Doctor of Philosophy

of The Australian National University

October 2017

This thesis is a Thesis by Compilation, as set out in ANU’s Higher degree by research -

thesis by compilation and thesis by creative works procedure

(https://policies.anu.edu.au/ppl/document/ANUP_003405). This thesis is my own work

except where otherwise acknowledged (refer to preface and acknowledgements).

Ding Li, Yong

29 October 2017

i

Preface

This thesis is based on the following publications either published or submitted,

which will be referred to by their roman numerals throughout the text:

I. Yong, D. L., Barton, P. S., Okada, S., Crane, M., & Lindenmayer, D. B. (2016).

Birds as surrogates for mammals and reptiles: Are patterns of cross-

taxonomic associations stable over time in a human-modified landscape?

Ecological Indicators, 69, 152-164.

II. Yong, D. L., Barton, P. S., Okada, S., Crane, M., Cunningham, S. A., &

Lindenmayer, D. B. (Submitted). Conserving bee and beetle assemblages in

farming landscapes: the role of landscape context and structure on cross-

taxonomic congruence. Biodiversity and Conservation.*

III. Yong, D. L., Barton, P. S., Ikin, K., Evans, M. J., Crane, M., Okada, S.,

Cunningham, S. A., & Lindenmayer, D. B. (In review). Validating cross-

taxonomic surrogates for biodiversity conservation in farming landscapes - a

multi-taxa approach. Biological Conservation.*

IV. Yong, D. L., Barton, P. S., Cunningham, S. A., & Lindenmayer, D. B.

(Submitted). Effects of anthropogenic disturbance on cross-taxonomic

congruence patterns in terrestrial ecosystems. Diversity and Distributions.

*Elements of publications II and III were also presented as an oral presentation at

the ICE 2016 XXV International Congress of Entomology in Orlando, USA (with an

accepted abstract; see Appendix I) and as an invited seminar, "Measuring insect

diversity in the Australian countryside - a surrogate approach to conservation

planning", at the Wildlife Ecology and Conservation Department, University of

Florida.

ii

I performed the majority of the work in the publications that constitute this thesis.

This includes the majority of the field data collection, laboratory work, statistical

analysis and writing. The design of the Nanangroe Natural Experiment was

developed by David. B. Lindenmayer, Ross B. Cunningham, Chris MacGregor and

colleagues in 1997, in collaboration with a number of private landowners and the

State Forests of New South Wales (now Forestry Corporation of New South Wales)

which manages the pine plantations in the area. In addition to the regular field

sampling of vertebrate taxa conducted at sites in the Nanangroe landscape, I

extended the data collection to invertebrates and other vegetation structure

variables. I was assisted during field work by Sachiko Okada and Mason Crane, who

supported me with field logistics. I performed the majority of the laboratory work on

processing the invertebrate samples, but was assisted by Michael Neave and

Maldwyn John Evans during the sorting and identification phases. Lastly, the co-

authors of each paper provided detailed comments and feedback during the drafting

of each paper.

Verification by my co-authors that the above statement is true (April 2017):

Dr. Philip S. Barton

Dr. Mason Crane

Prof. Saul A. Cunningham

Dr. Maldwyn John Evans

iii

Dr. Karen Ikin

Prof. David B. Lindenmayer

Ms. Sachiko Okada

iv

Acknowledgments

I am especially grateful to my supervisors, Philip Barton, Saul Cunningham and

David Lindenmayer for their constant guidance, patience and encouragement

throughout my candidature. Their tremendous support helped me to overcome a

myriad of challenges that I have encountered during my PhD journey, and included

anything from learning how to sample beetles to maintaining the stamina needed to

draft lengthy, coherent and readable manuscripts. Their important feedback on

various aspects of my work has certainly made me a far better researcher than when

I started three years ago.

As my research formed part of the Nanangroe natural experiment long-term

study, I am especially indebted to the ANU field team based at Gundagai and Tumut

for their assistance, care and support. Most of all, I thank Sachiko Okada, Mason

Crane and their families for hosting me during my series of trips to Gundagai to

survey or work at the Nanangroe study sites. Without Sachiko’s help, I would not be

able to efficiently get to so many bird sampling points on time, or set up and retrieve

my insect traps at the widely scattered sites across Nanangroe during daylight hours

and still find time to prepare insects at night. Mason helped me with many aspects

of logistics and gave me a comfortable place to stay during my stints to Gundagai. I

am grateful to the staff at the Forestry Corporation of New South Wales (Tumut) for

granting me the permission to work at sites within the pine plantations under their

management, and keeping me updated on logging activities within the Nanangroe

plantations. I also thank the landowners who facilitated my access to study sites in

their private property. They include the Luff, Greeny, Keating and Skinner families.

I wish to extend my gratitude to a number of people who assisted me with the

sorting and identification of bee and beetle specimens. Michael Batley (Australian

Museum), Michael Schwarz (Flinders University), John Ascher (National University

v

of Singapore) and Michael Neave (CSIRO Ecosystems, Black Mountain) provided

invaluable advice on how to prepare and mount bee specimens, and interpret

complicated insect keys to get each individual identified to the species-level.

Michael Batley hosted me during my working trip to the Australian Museum to

identify difficult bee species. Pia Lentini (University of Melbourne) provided useful

advice on how to set up blue vane traps to sample bees. M. John Evans assisted me

with the sorting of beetles and gave sound advice on how to identify beetles

efficiently, which diversity was overwhelming at times. Clive Hilliker helped me with

the preparation of maps. Thanks also go to Wade Blanchard, Maldwyn John Evans,

Claire Foster, Mark Hall, Karen Ikin, Geoffrey Kay, Katherina Ng, Jennifer Pierson,

Ayesha Tulloch and Martin Westgate for many engaging discussions on Australian

natural history, sampling methodology and statistical approaches for analysing my

dataset. I am also indebted to a number of my peers in academia who provided me

with encouragement and support during the course of my PhD candidature. They

include Kwek Yan Chong, Luke Gibson, J. Berton C. Harris, Fangyuan Hua, Lian Pin

Koh, Benjamin Lee, Brett Scheffers, David S. Wilcove and the late Navjot S. Sodhi.

My field work was supported by a research grant from the Robert Lesslie

Endowment, while academic travel (to the U.S.A.) to present my research findings at

the International Entomological Congress was supported by an Australian National

University Vice-Chancellor’s Travel Grant.

Finally, I thank Shiori Shakuto for her support and care during the most

daunting parts of my studies here in Australia. My dad and mom back in Singapore

tolerated my absence for the past three years and I too, am extremely grateful for

their unwavering support during this period.

vi

Abstract

Surrogates of biodiversity are necessary tools for guiding the effective conservation

of biodiversity. One of the best known approaches to assessing biodiversity is cross-

taxonomic surrogacy, which is underpinned by the hypothesis that selected taxa (i.e.

the surrogate) can provide useful and commensurate information on other

components of biodiversity (i.e. the target). In this thesis, I examined the

effectiveness of cross-taxonomic surrogacy by assessing cross-taxonomic congruency

patterns among ecologically important vertebrate and invertebrate taxa, and with

respect to time and different landscape contexts.

Using a long-term dataset, I first assessed patterns of cross-taxonomic

congruence between three vertebrate groups over a 15-year period. My analyses

revealed that patterns of cross-taxonomic congruency were inconsistent over time,

varied among the taxa compared, and across different landscape contexts. Bird and

mammal diversity were weakly concordant, but strengthened with time. However,

there was little association between either birds or mammals, and reptiles. My

findings suggested that cross-taxonomic surrogacy has limited effectiveness in

heavily disturbed landscapes such as Nanangroe where ecological communities are

expected to exhibit high temporal variation.

Second, I examined the responses of two ecologically important insect groups

(wild bees, beetles) to landscape context. Here, I found that species richness of bee

assemblages showed no clear responses to different landscape contexts, unlike

beetle assemblages. These patterns persisted even when both insect assemblage was

partitioned into functionally-defined groups. Further analyses showed that both

groups were responding to different landscape and vegetation components. My

findings here demonstrated that wild bee diversity is weakly congruent with beetle

diversity, and that surrogacy relationships between even charismatic insects should

not be assumed without rigorous testing.

vii

Third, I examined how sets of woodland patches prioritised to best conserve

each surrogate group (bird, herpetofauna, bee, beetle, tree) represented the other

four groups using a complementarity-based approach. Thereafter, I compared these

findings with correlation-based analysis to determine patterns of cross-taxonomic

congruence. I found that patch sets selected to optimise representation of the

surrogate varied in how it incidentally represented other taxa. Beetles achieved the

highest incidental representation of other taxa while bees and trees performed the

worst. Yet, beetles were the most costly taxa to conserve given the large number of

patches needed to meet beetle targets, an ecological consequence of the high

diversity and compositional turnover of beetle assemblages. My findings show that

species diversity of any taxa should be a pertinent consideration in identifying cross-

taxonomic surrogates to prioritise sites for biodiversity conservation.

Fourth, I performed a meta-analytical review of the global surrogate literature to

assess the effects of anthropogenic disturbance on cross-taxonomic surrogacy in

terrestrial systems. Drawing from a dataset compiled from 146 studies, my analyses

revealed that anthropogenic disturbance plays an important role in shaping patterns

of cross-taxonomic congruence, especially at landscape and regional scales. Spatial

scale was an important predictor of cross-taxonomic patterns, but only at very large

scales. In conclusion, my findings caution against extrapolating cross-taxonomic

surrogates across landscapes subjected to different levels of disturbance and spatial

scales to assess biodiversity.

Focussing on ubiquitous, human-modified landscapes, my work underscores a

number of practical and theoretical issues concerning the use of cross-taxonomic

surrogacy. By collectively or individually examining the roles of time, landscape

context and habitat structure with respect to diverse groups of vertebrate and

invertebrate taxa, my thesis makes explicit the need to consider important ecological

processes that can better guide the use of biodiversity surrogates in conservation.

viii

Table of contents

Preface…………………………………………………………………………………..................... i

Acknowledgments………………………………………………………………………………….. iv

Abstract………………………………………………………………………………………………….. vi

Extended context statement………………………………………………………………….. 1

Paper I. Birds as surrogates for mammals and reptiles: Are patterns of

cross-taxonomic associations stable over time in a human-modified

landscape? .....................................................................................................

25

Paper II. Conserving bee and beetle assemblages in farming landscapes:

the role of landscape context and structure on cross-taxonomic

congruency ……………………………………………………………………………………………..

81

Paper III. Validating cross-taxonomic surrogates for biodiversity

conservation in farming landscapes - a multi-taxa approach…………….......

131

Paper IV. Effects of anthropogenic disturbance on cross-taxonomic

congruence patterns in terrestrial ecosystems ……………………………………….

175

Appendix I. Submitted abstract for ICE 2016 XXV………………………………….. 220

Appendix II. List of cross-taxonomic surrogacy papers compiled for

meta-analysis, classified by continent …………………………………………………….

221

ix

1

Extended context statement

Introduction

The loss of biodiversity is an issue of global concern because of its significant

impacts on the world’s ecosystems (e.g. Hooper et al. 2012) and human wellbeing

(e.g. Dirzo et al. 2015). The rate of biodiversity decline continues (Butchart et al.

2010) in spite of better recognition of the global biodiversity crisis and the need to

set urgent conservation priorities (Wilson et al. 2006; Turner et al. 2007). However,

the assessment and monitoring of biodiversity to guide conservation priority setting

and management is regularly confronted with impediments such as funding

limitations (Pearson 1994; Duelli & O’brist 2003), insufficient data on species

distributions (Possingham et al. 2007; Guisan et al. 2013) and limited taxonomic

expertise (Ward & Larivière 2004; Favreau et al. 2006; Rodrigues & Brooks 2007).

Together, these challenges have fostered the development of numerous surrogate

approaches to enable the expedient assessment of biodiversity and organise

conservation activities (Caro 2010), such as designing networks of reserves through

systematic conservation planning (e.g. Margules & Pressey 2000; Ball et al. 2009) and

biodiversity monitoring (e.g. Lindenmayer & Likens 2010; Kessler et al. 2011).

The concept of surrogacy is grounded on the presumption that a measured

subset of biodiversity provides useful information on broader biodiversity patterns

or changes (Larsen et al. 2012; Barton et al. 2015; Lindenmayer et al. 2015). Surrogate

approaches arise out of a need for expediency in conservation; they therefore

functions as shortcuts by acting as a proxy for other components of biodiversity

without the need for direct measure (Balmford et al. 1996; Rodrigues & Brooks 2007;

Lindenmayer & Likens 2011). Many surrogate approaches are based on species data,

including cross-taxonomic surrogates (e.g. Kati et al. 2004; Gallardo et al. 2011),

2

selected indicator species (Roberge & Angelstam 2004; Branton & Richardson 2011;

Nicholson et al. 2013), higher-taxonomic diversity (Heino & Soininen 2007) and

species functional traits (e.g. Vandewalle et al. 2010). Other surrogate approaches are

based on landscape or environmental data (e.g. Dauber et al. 2003; Brin et al. 2009;

Barton et al. 2014). If validated with robust datasets and used with explicit

(conservation management) goals, surrogate approaches can indeed provide feasible

alternatives to direct measurements of biodiversity (Lindenmayer & Likens 2011).

One of the best known applications of surrogates in conservation are cross-

taxonomic surrogates. Cross-taxonomic surrogates are based on the principle that

one taxonomic group (i.e. the surrogate) co-vary with the diversity of other groups

(Lovell et al. 2007; Sætersdal & Gjerde 2011; Westgate et al. 2014). Such congruence in

diversity patterns across taxa (i.e. cross-taxonomic congruence) may be driven by

similar responses of each taxa to environmental gradients (e.g. Heino 2010),

structuring effects due to biotic interactions between species assemblages (e.g.

Larsen et al. 2012) or direct evolutionary and ecological relationships (mammals-

dung beetles) (e.g. Nichols et al. 2009). Despite the great amount of effort invested

into testing cross-taxonomic surrogacy, the outcomes have been mixed and not

necessarily conclusive (Grenyer et al. 2006; Possingham et al. 2007; Fattorini et al.

2012). Studies have revealed that the effectiveness of cross-taxonomic surrogates can

vary across habitats (e.g. Abensperg-traun et al. 1996; Wesner & Belk 2012), at

different scales of investigation (Ricketts et al. 2002; Hess et al. 2006; Cushman et al.

2010; Westgate et al. 2014) and across biogeographic regions (Jenkins et al. 2013).

Because agricultural and other anthropogenic activities have already altered a

large proportion of the Earth’s land surface and natural vegetation cover (e.g. Foley

et al. 2005), the effective conservation of biodiversity will surely require the

extension of conservation initiatives into human-modified landscapes such as

plantations and farmland (Lindenmayer & Hobbs 2004; Norris 2008; Koh & Gardner

3

2010; Sodhi et al. 2010). Alongside this is a need to better understand how

anthropogenic disturbance and temporal variation can influence cross-taxonomic

surrogacy (Wolters et al. 2006; Rooney & Azeria 2015; Yong et al. 2016) to better

understand biodiversity responses and plan for its conservation. Yet, few studies

have explicitly examined cross-taxonomic surrogacy in the context of land use

change and anthropogenic disturbance (Wolters et al. 2006; de Andrade et al. 2014).

Furthermore, there is a paucity of studies that have examined the response of

multiple insect groups to landscape modification (e.g. Tscharntke et al. 2002;

Gardner et al. 2009) or with respect to surrogacy (Ward & Larivière 2004; Lovell et

al. 2007), despite the diverse ecological roles played by insects as pollinators,

herbivores, ecosystem engineers and as prey for other taxa (Losey & Vaughan 2006)

and their enormous diversity (e.g. Stork et al. 2015; Hochkirch 2016). Clearly, there

remains considerable knowledge gaps on how cross-taxonomic surrogates perform

in human-modified environments under a diversity of land uses, and in relation to

different insect assemblages.

My thesis aimed to adopt a multi-taxa approach covering both invertebrates (i.e.

insects) and vertebrates to assess the broader impacts of landscape modification on

cross-taxonomic surrogacy. Specifically, I tested the role of landscape context,

vegetation structure and temporal variation in driving diversity patterns of different

taxonomic groups, and how this in turn influenced patterns of cross-taxonomic

congruence. I also aimed to test and integrate different analytical approaches in

quantifying cross-taxonomic surrogacy, and how this may lead to practical

implications for conservation management actions.

In Paper I, I addressed the temporal dimension of cross-taxonomic surrogacy. I

aimed to investigate how patterns of cross-taxonomic congruence among three well-

studied vertebrate groups (i.e. bird, mammal, reptile) varied over time, and with

respect to shifts in species richness, composition and abundances for each group in

4

the Nanangroe landscape. Understanding the temporal consistency of cross-

taxonomic patterns is important because it has immediate implications on the

contexts in which cross-taxonomic surrogates can be used to monitor biodiversity

effectively. I based my analysis on a dataset for these three taxa that was collected

over a 15-year period (1999-2013), in addition to field work carried out in the spring

of 2013-2014. Shifts in species assemblages arising from temporal variation, if found

to be inconsistent across taxa, could be an important driver of variation in cross-

taxonomic patterns over time. I then examined how the diversity (species

composition) of each group correlated to a set of vegetation and landscape structural

variables. Shared or similar responses to the environment is frequently hypothesised

to drive cross-taxonomic congruence, and therefore forms an important conceptual

basis to understanding cross-taxonomic surrogacy.

In Paper II, I adopted a multi-taxa approach to compare how two ecologically

important group of insects responded to the different landscape contexts and

vegetation structure in a human-modified landscape. I incorporated a functional-

dimension into my study design by further examining how insect groups defined by

their shared functional attributes (e.g. nesting requirements, dispersal capability)

would respond differently to landscape context. Because cross-taxonomic

congruency is widely hypothesised to arise when different taxa respond to the

landscape change and habitat structure in similar ways, determining the

environmental variables important to different groups could provide an important

first step to identifying consistent species surrogates of biodiversity. In Paper III, I

again adopted a multi-taxa approach to assess cross-taxonomic surrogacy between

five taxa that were simultaneously sampled in the Nanangroe landscape. In this

study, I used three conceptually different approaches to quantify cross-taxonomic

surrogacy. I first used a correlative approach to measure cross-taxonomic

congruence between pairs of taxa. I then compared this with an incidental

representation-based approach to assess how sets of sites optimised to conserve one

taxon could benefit a suite of other taxa. Next, I used a hierarchical clustering

5

approach to assess the degree of spatial overlap between sets of sites selected for

each taxa. Integrating the findings of different analytical approaches to examine

cross-taxonomic surrogacy could yield insights, both practical and theoretical, that

could guide the more efficient use of cross-taxonomic surrogates.

In Paper IV, I aimed to investigate the role of disturbance, spatial scale, time and

their interactions in explaining the congruence in cross-taxonomic patterns across

terrestrial ecosystems around the world. Given that ecological communities of many

human-modified landscapes are dynamic and undergo significant variation with

time due to various ecological processes (e.g. extirpations, colonisations, and

succession), disturbance and its interaction with space and time are important, yet

seldom studied ecological components in the context of cross-taxonomic surrogacy.

In this study, I performed a systematic review of the surrogate literature with a focus

on cross-taxonomic surrogates. I quantified the levels of disturbance in the study

sites, as well as the spatio-temporal attributes of each study. I then used a meta-

regression approach to assess how these attributes can explain the variation in cross-

taxonomic congruence. Findings from this component of my thesis will pose a

number of implications on the use of cross-taxonomic surrogates for conservation

management in landscapes subjected to different levels of disturbance.

Methodology

My study area was the Nanangroe experimental landscape, a large area (c. 30,000

hectares) of mostly cleared box gum grassy woodland in south-eastern Australia

(New South Wales) west of the Australian Capital Territory (ACT). The landscape

consists of grazing pasture, dotted with over 100 remnant patches of woodland

ranging in area from 0.25 to 21.1 hectares (Fischer & Lindenmayer 2002; Lindenmayer

et al. 2008). Establishment of plantations of the exotic Monterey or radiata pine

(Pinus radiata) in the late 1990s resulted in significant changes to the landscape,

6

leading to many of these remnant woodland patches being now surrounded by a

novel, pine plantation-dominated landscape matrix. Since 1999, regular surveys of

the vertebrate biota and vegetation structure in this landscape has established a

large, longitudinal dataset, allowing important ecological questions with a temporal

dimension to be asked and tested.

To determine the biotic composition in these woodland patches and how

different groups of taxa can function as cross-taxonomic surrogates of biodiversity, I

carried out standardised surveys of a broad representation of the biota in the

Nanangroe landscape. To do so, I conducted systematic surveys of both invertebrate

and vertebrate groups in woodland patches in two kinds of landscape contexts: (1)

woodland patches in the agricultural (grazing) matrix and (2) woodland patches in

the pine plantation matrix. Further surveys were carried out in a number of sites in

pine plantations as habitat contrasts. Three vertebrate groups, namely birds,

mammals and herpetofauna were sampled using line transect and point count

surveys in each woodland patch.

In addition to vertebrate surveys, I also carried out surveys of the wild bee and

ground-active beetle fauna. Both insect groups were selected because of their large

contribution to species diversity in many types of landscapes (Stork & Grimbacher

2006), and their diverse ecological and functional roles (Kremen et al. 2002). Wild

bee assemblages were surveyed using blue vane traps, an approach to sampling bees

that have been increasingly used in recent years (e.g. Lentini et al. 2012). The

ground-active beetle fauna was sampled using fenced, pit-fall traps. All taxa

surveyed were identified to the species level with the exception of beetles which

were sorted first to family, and then to the morphospecies-level (sensu Oliver &

Beattie 1996). More details on the sampling methodology for each taxon can be

found in papers I, II and III.

7

I used three main statistical approaches to test for cross-taxonomic surrogacy

between pairs of the five taxa sampled. In papers I, II and III, I employed a

combination of regression (including multiple regression in matrix form) and

ordination analysis to test for correlative associations between different taxa, and in

relation to a diversity of landscape and vegetation structural measures. Model

selection was performed using Bayesian model averaging. These analyses were

implemented on the ‘lme4’ (Bates et al. 2014), ‘BMA’ (Raftery et al. 2014), ‘vegan’

(Oksanen et al. 2013) and ‘ecodist’ (Goslee & Urban 2007) packages available on the

R platform (R Core Team 2013). In paper III, I employed a heuristic approach to

select near-optimal sets of sites to meet a range of representation targets for each

taxon. This analyses is implemented with the simulated annealing algorithm and

was carried out using Marxan, a software developed for spatial conservation

prioritisation and reserve-selection (Ball et al. 2009; Beyer et al. 2016). In paper IV, I

used a variety of meta-regression models implemented on ‘metafor’ (Vietchbauer

2014), an R package specifically designed for meta-analyses.

Summary of outcomes

Paper I. Birds as surrogates for mammals and reptiles: are patterns of cross-taxonomic

associations stable in a human-modified landscape?

In this study, I assessed changes in cross-taxonomic associations over time between

birds, a well-known group of surrogates (Di Minin & Moilanen 2014; Ikin et al. 2016)

and two target vertebrate taxa that demand high sampling effort (mammals,

reptiles). Specifically, I investigated, (1) temporal changes in cross-taxonomic

congruency across three animal taxa, (2) explored how temporal variation in

composition and species richness of each taxon can account for pairwise cross-

taxonomic congruency, and (3) identified habitat structural variables that influenced

the species composition of each taxon.

8

Using available data on the three taxa, and new data collected over spring 2013-

2014, I found that cross-taxonomic associations between taxa were inconsistent over

time, and differed between the landscape contexts. Correlations of species richness

and composition were consistently positive and strongest between birds and

mammals, strengthening over time. This contrasted with pairwise comparisons

involving bird and mammals with reptiles, which were mostly weak and negative.

Additionally, I found that temporal variation in diversity differed in rate and extent

among the three groups while vegetation structure components significantly

correlated with each group were seldom shared. This study draws attention to the

seldom investigated role of temporal variation in shaping cross-taxonomic

associations between different taxa, especially in a heavily disturbed landscape

immediately following perturbation (i.e. plantation establishment). I concluded that

in dynamic landscapes such as Nanangroe, taxon-specific shifts in diversity over

time can influence the strength, direction and consistency of cross-taxonomic

associations. From a conservation perspective, these findings poses a ‘temporal’

problem to the use of popular taxonomic surrogates such a birds to account for

other biota when assessing biodiversity to inform conservation management.

Paper II. Comparing bee and beetle assemblages in modified landscapes: testing the

effectiveness of a surrogate-based conservation approach

Identifying the shared responses of different insect taxa to landscape modification is

a necessary first step to understand how the conservation of selected key groups of

insects can benefit other groups as potential surrogates (e.g. Schweiger et al. 2005).

In this study, I investigated how two ecologically important groups of insects,

namely wild bees and ground-active beetles, responded to the different landscape

contexts that has arisen from land use change. By assessing how diversity (species

richness, species composition) of bees and ground-active beetles responded to

different landscape contexts and components of the vegetation structure, I test the

hypothesis that cross-taxonomic congruency between different insect taxa is a

consequence of shared responses to changes in landscape structure. Both wild bee

9

and ground-active beetle assemblages were sampled in woodland patches in two

landscape contexts: surrounded by pine plantation (the pine matrix) or by cleared,

grazing land (the agricultural matrix), and in pine monoculture in the Nanangroe

landscape.

My study revealed that wild bee species richness was not significantly different

across woodland remnants in different landscape contexts after controlling for

spatial effects. There were no significant differences in bee assemblages across

different landscape contexts even after I partitioned the bee dataset into sub-groups

defined by shared functional traits (i.e. nesting substrate requirement). Unlike bees,

total beetle species richness, and richness of the functionally-defined sub-groups

varied significantly across the landscape contexts. My analyses also showed that

landscape context exerted a stronger effect on species composition than species

richness for both wild bees and ground-active beetles. Such compositional

differences will not be apparent if only species richness was used as a metric of

diversity. My ordination analysis suggested that some landscape and habitat

variables were useful in predicting the diversity of either group, but few were shared.

I concluded that wild bee and ground-active beetle assemblages are poor

surrogates for each other in agricultural landscapes because of, (1) the low cross-

taxonomic congruency between them, and (2) the highly dissimilar responses of the

two insect groups to landscape context and vegetation structure. However, I suggest

that where needed, sets of landscape and habitat variables such as native tree cover

could be used as habitat-based surrogates for either insect group. My study

highlights the need to consider: (1) taxon-specific responses to landscape context, (2)

different metrics of cross-taxonomic surrogacy and (3) differences in ecological

attributes across insect taxa when managing landscapes for insect conservation.

Ultimately, agricultural landscapes managed solely to conserve focal pollinator

10

assemblages (e.g. wild bees) should not be assumed to benefit other insect

assemblages without rigorous testing.

Paper III. Validating cross-taxonomic surrogates for biodiversity conservation in

farming landscapes - a multi-taxa approach

In this study, I evaluated the effectiveness of five taxa as surrogates for other taxa in

woodland patches using a, (1) correlative (e.g. Sauberer et al. 2004; Lovell et al.

2007), (2) incidental representation-based (e.g. Sætersdal et al. 2004; Albuquerque &

Beier 2016) and, (3) hierarchical clustering-based approach (e.g. Wiens et al. 2008;

Ikin et al. 2016). I used occurrence data of five taxa comprising two groups of

vertebrate (birds, herpetofauna), two groups of invertebrate (bees, beetles) and one

plant group (trees). I first compared species richness and compositional turnover

patterns across taxa. I then assessed how species richness and composition co-varied

between pairs of taxa using a correlative approach. To validate the cross-taxonomic

surrogates identified this way, I then quantified how well each taxon incidentally

represented other taxa in their best patch sets using a complementarity-based site

selected approach implemented with the simulated annealing algorithm. Last, I used

hierarchical clustering to determine the extent of similarity and spatial overlap

between patch sets for each taxa, and across a wide range of representation targets.

Across taxa, my study revealed that the beetle assemblage exhibited the highest

compositional heterogeneity among the taxa studied, while the bee and tree

assemblages showed the lowest heterogeneity. I found significant correlations

between some taxa-pairs for species richness (bird-bee) and composition (e.g.

beetle-herpetofauna, beetle-bee, bird-bee), suggesting that birds could act as good

surrogates for bees and vice versa. However, no single taxon was well-correlated

with all other taxa. With the complementarity-based site selection approach, I found

that sets of woodland patches selected to optimise beetle representation performed

11

best in incidentally representing other taxa across the range of targets. For instance,

sets of patches selected for beetles represented birds, bees, herpetofauna and trees

better than patch sets selected for other taxa. However, beetle-focused patch sets

were consistently the costliest among the taxa and required maintaining the highest

amount of woodland patches. By comparison, bees and trees were the worst

performing taxa in incidentally representing other taxa, although both were far less

costly. However, an increase in the representation targets for bee and tree

occurrences paralleled consistent increment in representation of other taxa. This

suggested that both have the potential to be efficient surrogates for other taxa.

This study indicates that taxon-specific patterns of compositional turnover have

an important bearing on how sets of woodland patches should be prioritised for

conservation in farming landscapes. Taxa with high species diversity such as beetles

will show higher compositional heterogeneity, and thus demand the retention of

more habitat patches across any landscape in systematic conservation (reserve-

selection and design) scenarios to capture their diversity. Such practical

considerations for conservation may not be made explicit in biodiversity surrogate

studies using a purely correlative approach. Therefore, while it is attractive to select

species-rich taxa as surrogates because of the wider range of habitats and ecological

conditions captured by landscapes prioritised with these groups (Larsen et al. 2012;

Ikin et al. 2016), due consideration need to be made for practical issues such as the

increased allocation of land (i.e. through land sparing) for conserving biodiversity in

landscapes prioritised for agricultural production.

Paper IV. Global meta-analysis reveals effects of anthropogenic disturbance on cross-

taxonomic congruency patterns across terrestrial ecosystems

In this study, I synthesised the global literature on studies of cross-taxonomic

surrogates using a meta-analytical approach to investigate how three ecologically

informative attributes: (1) anthropogenic disturbance, (2) spatial scale and the (3)

12

duration of a study’s data collection period, can influence the cross-taxonomic

congruency patterns that underpin many aspects of biodiversity surrogacy. To do so,

I performed a systematic literature review on the ISI Web of Science and Scopus

databases using a priori defined search terms. After checking each study for

relevance to cross-taxonomic surrogacy in terrestrial systems, I extracted and

compiled data in two measures of effect sizes (correlations of species richness and

species composition), and information on disturbance, spatial scale and study

duration from each study. I then performed meta-regression analysis to test how

well each of these study attributes predicted patterns of cross-taxonomic

congruence for both effect size measures.

My systematic review identified a total of 146 studies, which yielded 1,633

measures of pairwise correlations of species richness, and 1,030 measures of pairwise

correlations of species composition. Meta-regression analyses revealed that the level

of disturbance in a landscape was a significant predictor of cross-taxonomic

congruence in both species richness and composition measures. I found that

patterns of cross-taxonomic congruence was highest in landscapes that are most

heavily disturbed and lowest in relatively undisturbed landscapes (e.g. reserves,

national parks). Interactions between spatial scale and disturbance revealed that

cross-taxonomic congruency declined with increasing levels of disturbance at the

smallest spatial scale of study (local). However, at larger spatial scales (landscape,

region), cross-taxonomic congruency increased with disturbance level. Spatial scale

alone was a relatively weak predictor of cross-taxonomic congruency, becoming

important only at large (global) or small (local) scales. I also showed that the

duration of a study negatively predicts cross-taxonomic congruency for both species

richness and composition. Interactions with disturbance revealed that cross-

taxonomic congruency declined with study duration at low disturbance levels, but

increased at higher disturbance levels.

13

My analyses demonstrate the importance of anthropogenic disturbance effects

in shaping cross-taxonomic congruency patterns at different spatial and temporal

scales. While it is unclear what processes are driving these variation, possible

explanations include shifts in species assemblages leading to species extirpations,

biotic homogenisation, or colonisation by generalist species. Alternatively, it may be

that environmental gradients created by disturbance may enhance the species pool

across the landscape, promoting cross-taxonomic congruency (Rooney & Azeria

2015). Clearly, a better understanding of the ecological processes underpinning these

patterns is needed to guide biodiversity conservation initiatives in the world’s

modified agricultural landscapes.

Conclusion

Given resource limitations, biodiversity surrogates will remain as vital tools for

conservation practitioners to assess biodiversity in a wide variety of contexts and

conservation management scenarios in the foreseeable future. By examining

patterns of cross-taxonomic congruency over time, under different landscape

contexts and across multiple taxa, my thesis weaves together a number of critical

ecological considerations often overlooked in studies of cross-taxonomic surrogates,

especially in the context of human-modified landscapes. First, taxon-specific,

temporal variation in diversity can account for inconsistent cross-taxonomic

congruence over time, especially in transformed landscapes. Thus, cross-taxonomic

surrogates should not be assumed to be consistently effective over time. Taxa that

show high temporal variation in their diversity patterns should be cautiously used as

surrogates as they may not consistently predict the diversity of other target taxa.

Second, cross-taxonomic congruence patterns between some of the best known, and

most easily sampled insect groups can be limited given the taxon-specific responses

to different components of the landscape. Wild bees for instance, have been the

focus of many conservation initiatives. However, the differences between how bee

14

and beetle assemblages respond to landscape change suggests that bees are likely

poor surrogates for other insects, and that landscapes targeted at conserving only

bees may not necessarily benefit other important insect groups. Third, the diversity

of any given taxonomic group has a direct bearing on how well it can be used as a

(coarse-filter) surrogate. While species-rich taxa can represent a greater diversity of

environmental conditions that capture more of other biota, conserving them will

inevitably require the retention of more, and larger sites. Using species-rich groups

as surrogates for conservation planning may thus conflict with agricultural

production in landscapes prioritised for farming activities. Fourth, the role of

disturbance, while seldom explored with respect to surrogacy, can interact with

spatial scale to influence the degree of co-variation in species richness and

composition between different taxa. This means that cross-taxonomic surrogates

identified in one landscape should not be assumed to perform well in landscapes

subjected to different levels of disturbance, and across different spatial scales.

Collectively, my findings raise a number of considerations on how different species

groups should be used as surrogates for other biodiversity to guide conservation

management of human-modified landscapes. In particular, the fact that cross-

taxonomic relationships can vary in time, space, disturbance and landscape contexts

implies that the use of specific group(s) of taxa as surrogates must be carefully

considered in relation to the goals of any project or initiative attempting to assess

biodiversity for conservation outcomes. Lastly, an important scope for future work is

to compare cross-taxonomic surrogates effectiveness with respect to spatial scale,

time and disturbance, with alternative surrogate approaches focused on landscape

and environmental measures.

15

References cited

Abensperg-traun, M., Arnold, G. W., Steven, D. E., Smith, G. T., Atkins, L., Viveen, J.

J. & Gutter, M. (1996) Biodiversity indicators in semi-arid, agricultural Western

Australia. Pacific Conservation Biology, 2, 375-388.

Albuquerque, F. S., Beier, P. (2016) Downscaling patterns of complementarity to a

finer resolution and its implications for conservation prioritization. Ecology and

Evolution, 6(12), 4032-4040.

Andelman, S. J. & Fagan, W. F. (2000) Umbrellas and flagships: efficient

conservation surrogates or expensive mistakes? Proceedings of the National

Academy of Sciences U.S.A., 97(11), 5954-5959.

Ball, I. R., Possingham, H. P., & Watts, M. (2009). Marxan and relatives: Software for

spatial conservation prioritisation. In: Moilanen, A., Wilson, K. A. & Possingham,

H. P. eds. Spatial conservation prioritisation: Quantitative methods and

computational tools. Oxford University Press, Oxford, UK.

Balmford, A., Green, M. J. B. & Murray, M. G. (1996) Using higher-taxon richness as a

surrogate for species richness: I. Regional tests. Proceedings of the Royal Society of

London B: Biological Sciences, 263, 1267-1274.

Barton, P. S., Westgate, M. J., Lane, P. W., MacGregor, C., & Lindenmayer, D. B.

(2014) Robustness of habitat‐based surrogates of animal diversity: a multitaxa

comparison over time. Journal of Applied Ecology, 51(5), 1434-1443.

Barton, P. S., Pierson, J. C., Westgate, M. J., Lane, P. W. & Lindenmayer, D. B. (2015)

Learning from clinical medicine to improve the use of surrogates in ecology.

Oikos, 124 (4), 391-398.

Bates, D., Maechler, M., Bolker, B. & Walker, S. (2014) lme4: Linear mixed-effects

models using Eigen and S4. R package version 1.1-7, <URL: http://CRAN.R-

project.org/package=lme4>.

16

Beyer, H. L., Dujardin, Y., Watts, M. E., & Possingham, H. P. (2016) Solving

conservation planning problems with integer linear programming. Ecological

Modelling, 328, 14-22.

Branton, M. & Richardson, J. S. (2011) Assessing the Value of the Umbrella-Species

Concept for Conservation Planning with Meta-Analysis. Conservation Biology, 25,

9-20.

Brin, A., Brustel, H. & Jactel, H. (2009) Species variables or environmental variables

as indicators of forest biodiversity: a case study using saproxylic beetles in

Maritime pine plantations. Annals of Forest Science, 66(3), 1-11.

Butchart, S. H., Walpole, M., Collen, B., Van Strien, A., Scharlemann, J. P., Almond,

R. E., Baillie, J. E. M., Bomhard, B., Brown, C., Bruno, J., Carpenter, K. E., Carr, G.

M., Chanson, J., Chenery, A., Csirke, J., Davidson, N. C., Dentener, F., Foster, M.,

Galli, A., Galloway, J. N., Genovesi, P., Gregory, R. D., Hockings, M., Kapos, V.,

Lamarque, J., Leverington, F., Loh, J., McGeoch, M., McRae, L., Minasyan, A.,

Morcillo, M. H., Oldfield, T. E. E., Pauly, D., Quader, S., Revenga, C., Sauer, J. R.,

Skolnik, B., Spear, D., Stanwell-Smith, D., Stuart, S. N., Symes, A., Tierney, M.,

Tyrrell, T. D., Vie, J. & Watson, R. (2010) Global biodiversity: indicators of recent

declines. Science, 328 (5982), 1164-1168.

Caro, T. (2010) Conservation by proxy: indicator, umbrella, keystone, flagship, and

other surrogate species. Island Press, U.S.A.

Cushman, S. A., McKelvey, K. S., Noon, B. R. & McGarigal, K. (2010) Use of

abundance of one species as a surrogate for abundance of others. Conservation

Biology, 24, 830-840.

Dauber, J., Hirsch, M., Simmering, D., Waldhardt, R., Otte, A. & Wolters, V. (2003)

Landscape structure as an indicator of biodiversity: matrix effects on species

richness. Agriculture, Ecosystems and Environment, 98, 321-329.

De Andrade, R. B., Barlow, J., Louzada, J., Mestre, L., Silveira, J., Vaz-de-Mello, F. Z.,

& Cochrane, M. A. (2014) Biotic congruence in humid tropical forests: a multi-taxa

17

examination of spatial distribution and responses to forest disturbance. Ecological

Indicators, 36, 572-581.

Di Minin, E. & Moilanen, A (2014). Improving the surrogacy effectiveness of

charismatic megafauna with well‐surveyed taxonomic groups and habitat types.

Journal of Applied Ecology, 51(2), 281-288.

Dirzo, R., Young, H. S., Galetti, M., Ceballos, G., Isaac, N. J. & Collen, B. (2014)

Defaunation in the Anthropocene. Science, 345 (6195), 401-406.

Duelli, P. & O’brist, M. K. (2003) Biodiversity indicators: the choice of values and

measures. Agriculture, Ecosystems and Environment, 98, 87-98.

Fattorini, S., Dennis, R. L. H. & Cook, L. M. (2012) Use of cross-taxon congruence for

hotspot identification at a regional scale. PLoS ONE, 7, e40018.

Favreau, J. M., Drew, C. A., Hess, G. R., Rubino, M. J., Koch, F. H. & Eschelbach, K. A.

(2006) Recommendations for assessing the effectiveness of surrogate species

approaches. Biodiversity and Conservation 15 (12), 3949-3969.

Fischer, J. & Lindenmayer, D. B. (2002). Small patches can be valuable for

biodiversity conservation: two case studies on birds in southeastern Australia.

Biological Conservation, 106(1), 129-136.

Fischer, J., Lindenmayer, D. B. & Manning, A. D. (2006) Biodiversity, ecosystem

function, and resilience: ten guiding principles for commodity production

landscapes. Frontiers in Ecology and the Environment, 4(2), 80-86.

Foley J. A., DeFries, R., Asner, G. P., Barford, C., Bonan, G., Carpenter, S. R., Chapin,

F. S., Coe, M. T., Daily, G. C., Gibbs, H. K., Helkowski, J. H., Holloway, T.,

Howard, E. A., Kucharik, C. J., Monfreda, C., Patz, J.A., Prentice, I. C.,

Ramankutty, N. & Snyder. P. K. (2005) Global consequences of land use. Science

309(5734): 570-574.

18

Gallardo, B., Gascon, S., Quintana, X. & Comin, F. A. (2011) How to choose a

biodiversity indicator – redundancy and complementarity of biodiversity metrics

in a freshwater ecosystem. Ecological Indicators, 11, 1177-1184.

Gardner, T. A., Barlow, J., Chazdon, R., Ewers, R. M., Harvey, C. A., Peres, C. A., &

Sodhi, N. S. (2009) Prospects for tropical forest biodiversity in a human‐modified

world. Ecology Letters, 12(6), 561-582.

Goslee, S. C., & Urban, D. L. (2007) The ecodist package for dissimilarity-based

analysis of ecological data. Journal of Statistical Software, 22(7), 1-19.

Grenyer, R., Orme, C. D. L., Jackson, S. F., Thomas, G. H., Davies, R. G., Jones, K. E.,

Olson, V. A., Ridgely, R. S., Rasmussen, P.C., Ding, T., Bennett, P., Blackburn,

T.M., Gaston, K. J., Gittleman, J. L. & Owens, I. P. F. (2006) Global distribution

and conservation of rare and threatened vertebrates. Nature, 444, 93-96.

Guisan, A., Tingley, R., Baumgartner, J. B., Naujokaitis‐Lewis, I., Sutcliffe, P. R.,

Tulloch, A. I., Regan, T. J., Brotons, L., McDonald-Madden, E., Mantyka-Pringle,

C., Martin, T.G., Rhodes, J. R., Maggini, R., Setterfield, S. A., Elith, J., Schwartz, M.

W., Wintle, B. A., Broennimann, O., Austin, M., Ferrier, S., Kearney, M. R.,

Possingham, H. P. & Buckley, Y. M. (2013) Predicting species distributions for

conservation decisions. Ecology Letters, 16(12), 1424-1435.

Heino, J. (2010) Are indicator groups and cross-taxon congruence useful for

predicting biodiversity in aquatic ecosystems? Ecological Indicators, 10(2), 112-117.

Heino, J. & Soininen, J. (2007) Are higher taxa adequate surrogates for species-level

assemblage patterns and species richness in stream organisms? Biological

Conservation, 137, 78-89.

Hess, G. R., Bartel, R. A., Leidner, A. K., Rosenfeld, K. M., Rubino, M. J., Snider, S. B.

& Ricketts, T. H. (2006) Effectiveness of biodiversity indicators varies with extent,

grain, and region. Biological Conservation, 132(4), 448-457.

Hochkirch, A. (2016). The insect crisis we can't ignore. Nature, 539(7628), 141.

19

Hooper, D. U., Adair, E. C., Cardinale, B. J., Byrnes, J. E., Hungate, B. A., Matulich, K.

L., Gonzalez, A., Duffy, J. E., Gamfeldt, L. & O’Connor, M. I. (2012) A global

synthesis reveals biodiversity loss as a major driver of ecosystem change. Nature,

486, 105-108.

Ikin, K., Yong, D. L. & Lindenmayer, D. B. (2016) Effectiveness of woodland birds as

taxonomic surrogates in conservation planning for biodiversity on farms.

Biological Conservation, 204, 411-416.

Jenkins, C. N., Guenard, B., Diamond, S. E., Weiser, M. D. & Dunn, R. R. (2013)

Conservation implications of divergent global patterns of ants and vertebrate

diversity. Diversity and Distributions, 19, 1084-1093.

Kati, V., Devillers, P., Dufrene, M., Legakis, A., Vokou, D. & Lebrun, P. (2004)

Testing the value of six taxonomic groups as biodiversity indicator at a local scale.

Conservation Biology, 18, 667-675.

Kessler, M., Abrahamczyk, S., Bos, M., Buchori, D., Putra, D. D., Robbert Gradstein,

S., Hohn, P., Kluge, J., Orend, F., Pitopang, R., Saleh, S., Schulze, C. H., Sporn, S.

G., Steffan-Dewenter, I., Tjitrosoedirdjo, S. S. & Tscharntke, T. (2011) Cost‐

effectiveness of plant and animal biodiversity indicators in tropical forest and

agroforest habitats. Journal of Applied Ecology, 48 (2), 330-339.

Koh, L. P. & Gardner, T. A. (2010) Conservation in human-modified landscapes. In

Sodhi, N. S., & Ehrlich, P. R. eds. Conservation biology for all. Oxford University

Press, U.K.

Kremen, C., Williams, N. M. & Thorp, R. W. (2002) Crop pollination from native

bees at risk from agricultural intensification. Proceedings of the National Academy

of Sciences U.S.A., 99(26), 16812-16816.

Larsen, F. W., Bladt, J., Balmford, A. & Rahbek, C. (2012) Birds as biodiversity

surrogates: will supplementing birds with other taxa improve effectiveness?

Journal of Applied Ecology, 49(2), 349-356.

20

Lentini, P. E., Martin, T. G., Gibbons, P., Fischer, J. & Cunningham, S. A. (2012)

Supporting wild pollinators in a temperate agricultural landscape: maintaining

mosaics of natural features and production. Biological Conservation, 149(1), 84-92.

Lindenmayer, D. B. & Hobbs, R. J. (2004) Biodiversity conservation in plantation

forests - a review with special reference to Australia. Biological Conservation, 119,

151-168.

Lindenmayer, D. B. & Likens, G. (2011) Direct measurement versus surrogate

indicator species for evaluating environmental change and biodiversity loss.

Ecosystems, 14, 57-59.

Lindenmayer, D. B., Cunningham, R. B., MacGregor, C., Tribolet, C., & Donnelly, C.

F. (2001) A prospective longitudinal study of landscape matrix effects on fauna in

woodland remnants: experimental design and baseline data. Biological

Conservation, 101(2), 157-169.

Lindenmayer, D. B., Fischer, J., Felton, A., Crane, M., Michael, D., Macgregor, C.,

Montague-Drake, R., Manning, A. & Hobbs, R. J. (2008) Novel ecosystems

resulting from landscape transformation create dilemmas for modern

conservation practice. Conservation Letters, 1(3), 129-135.

Lindenmayer, D., Barton, P., Pierson, J. eds. (2015) Indicators and Surrogates of

Biodiversity and Environmental Change. CSIRO Publishing, Australia.

Losey, J. E. & Vaughan, M. (2006) The economic value of ecological services provided

by insects. Bioscience, 56(4), 311-323

Lovell, S., Hamer, M., Slotow, R. & Herbert, D. (2007) Assessment of congruency

across invertebrate taxa and taxonomic levels to identify potential surrogates.

Biological Conservation, 139, 113-125.

Margules, C. R. & Pressey, R. L. (2000) Systematic conservation planning. Nature,

405, 243-253.

21

Nichols, E., Gardner, T. A., Peres, C. A. & Spector, S. (2009) Co‐declining mammals

and dung beetles: an impending ecological cascade. Oikos, 118(4), 481-487.

Nicholson, E., Lindenmayer, D.B., Karin, F. & Possingham, H. P. (2013) Testing the

focal species approach to making conservation decisions for species persistence.

Diversity and Distributions, 19, 530-540.

Norris, K. (2008) Agriculture and biodiversity conservation: opportunity knocks.

Conservation Letters 1(1), 2-11.

Oksanen, J., Blanchet, F. G., Kindt, R., Legendre, P., Minchin, P. R., O'Hara, R. B.,

Simpson, G. L., Solymos, P., Stevens, M. H. H. & Wagner, H. (2013) Vegan:

Community Ecology Package. R package version 2.0-10. http://CRAN.R-

project.org/package=vegan.

Oliver, I., & Beattie, A. J. (1996) Invertebrate morphospecies as surrogates for

species: a case study. Conservation Biology, 10(1), 99-109.

Pearson, D. L. (1994) Selecting indicator taxa for the quantitative assessment of

biodiversity. Philosophical Transactions of the Royal Society B: Biological Sciences,

345, 75-79.

Possingham, H. P., Grantham, H. & Rondinini, C. (2007) How can you conserve

species that haven't been found? Journal of Biogeography, 34 (5), 758-759.

R Core Team. (2013) R: A language and environment for statistical computing. R

Foundation for Statistical Computing, Vienna, Austria. URL http://www.R-

project.org/.

Raftery, A., Hoeting, J., Volinsky, C., Painter, I. & Yeung, K. Y. (2014) BMA: Bayesian

Model Averaging. R package version 3.18.1. http://CRAN.R-

project.org/package=BMA

Ricketts, T. H., Daily, G. C. & Ehrlich, P. R. (2002) Does butterfly diversity predict

moth diversity? Testing a popular indicator taxon at local scales. Biological

Conservation 103, 361-370.

22

Roberge, J.M. & Angelstam, P. (2004) Usefulness of the umbrella species concept as a

conservation tool. Conservation Biology 18, 76-85.

Rodrigues, A. S. & Brooks, T. M. (2007) Shortcuts for biodiversity conservation

planning: the effectiveness of surrogates. Annual Reviews of Ecology, Evolution

and Systematics, 38, 713-737.

Rooney, R. C., & Azeria, E. T. (2015) The strength of cross‐taxon congruence in

species composition varies with the size of regional species pools and the

intensity of human disturbance. Journal of Biogeography, 42(3), 439-451.

Sætersdal, M. & Gjerde, I. (2011) Prioritising conservation areas using species

surrogate measures: consistent with ecological theory? Journal of Applied Ecology,

48, 1236-1240.

Sætersdal, M., Gjerde, I., Blom, H. H., Ihlen, P. G., Myrseth, E. W., Pommeresche, R.,

John Skartveit, J., Solhøy, T. & Aas, O., (2004) Vascular plants as a surrogate

species group in complementary site selection for bryophytes, macrolichens,

spiders, carabids, staphylinids, snails, and wood living polypore fungi in a

northern forest. Biological Conservation 115(1), 21-31.

Sauberer, N., Zulka, K. P., Abensperg-Traun, M., Berg, H. M., Bieringer, G.,

Milasowszky, N., Moser, D., Plutzar, C., Pollheimer, M., Tröstl, R., Zechmeister,

H. & Grabherr, G. (2004) Surrogate taxa for biodiversity in agricultural landscapes

of eastern Austria. Biological Conservation, 117(2), 181-190.

Schweiger, O., Maelfait, J. P., Wingerden, W. V., Hendrickx, F., Billeter, R.,

Speelmans, M., Augenstein, I., Aukema, B., Aviron, S., Bailey, D., Bukacek, R.,

Burel, F., Diekötter, T., Dirksen, J., Frenzel, M., Herzog, F., Liira, J., Roubalova, M.

& Bugher, R. (2005) Quantifying the impact of environmental factors on

arthropod communities in agricultural landscapes across organizational levels

and spatial scales. Journal of Applied Ecology, 42(6), 1129-1139.

Sodhi, N. S., Koh, L. P., Clements, R., Wanger, T. C., Hill, J. K., Hamer, K. C., Clough,

Y., Tscharntke, T., Posa, M. R. C., & Lee, T. M. (2010) Conserving Southeast Asian

23

forest biodiversity in human-modified landscapes. Biological Conservation,

143(10), 2375-2384.

Stork, N. E., & Grimbacher, P. S. (2006) Beetle assemblages from an Australian

tropical rainforest show that the canopy and the ground strata contribute equally

to biodiversity. Proceedings of the Royal Society of London B: Biological Sciences,

273(1596), 1969-1975.

Stork, N. E., McBroom, J., Gely, C. & Hamilton, A. J. (2015) New approaches narrow

global species estimates for beetles, insects, and terrestrial arthropods.

Proceedings of the National Academy of Sciences U.S.A., 112(24), 7519-7523.

Tscharntke, T., Steffan-Dewenter, I., Kruess, A. & Thies, C. (2002) Contribution of

small habitat fragments to conservation of insect communities of grassland–

cropland landscapes. Ecological Applications, 12(2), 354-363

Tulloch, A. I., Chadès, I., Dujardin, Y., Westgate, M. J., Lane, P. W. & Lindenmayer,

D. (2016) Dynamic species co–occurrence networks require dynamic biodiversity

surrogates. Ecography, 39, 1185-1196.

Turner, W. R., Brandon, K., Brooks, T. M., Costanza, R., Da Fonseca, G. A. & Portela,

R. (2007) Global conservation of biodiversity and ecosystem services. BioScience,

57 (10), 868-873.

Vandewalle, M., De Bello, F., Berg, M.P., Bolger, T., Doledec, S., Dubs, F., Feld, C. K.,

Harrington, R., Harrison, P. A., Lavorel, S., da Silva, P. M., Moretti, M., Niemela,

J., Santos, P. & Sattler, T. (2010) Functional traits as indicators of biodiversity

response to land use changes across ecosystems and organisms. Biodiversity and

Conservation, 19, 2921-2947.

Viechtbauer, W. (2010) Conducting meta-analyses in R with the metafor package.

Journal of Statistical Software, 36(3), 1-48.

Ward, D.F. & Lariviere, M-C. (2004) Terrestrial invertebrate surveys and rapid

biodiversity assessments in New Zealand: lessons from Australia. New Zealand

Journal of Ecology, 28, 151-159.

24

Wesner, J.S. & Belk, M.C. (2012) Habitat relationships among biodiversity indicators

and co‐occurring species in a freshwater fish community. Animal Conservation,

15, 445-456.

Westgate, M. J., Barton, P. S., Lane, P. W. & Lindenmayer, D. B. (2014) Global meta-

analysis reveals low consistency of biodiversity congruence relationships. Nature

Communications, 5, 3899.

Wiens, J. A., Hayward, G. D., Holthausen, R. S. & Wisdom, M. J. (2008) Using

surrogate species and groups for conservation planning and management.

Bioscience, 58(3), 241-252.

Wilson, K. A., McBride, M. F., Bode, M. & Possingham, H. P. (2006) Prioritizing

global conservation efforts. Nature, 440, 337-340.

Wolters, V., Bengtsson, J., & Zaitsev, A. S. (2006) Relationship among the species

richness of different taxa. Ecology, 87(8), 1886-1895.

Yong, D. L., Barton, P. S., Okada, S., Crane, M., & Lindenmayer, D. B. (2016). Birds as

surrogates for mammals and reptiles: Are patterns of cross-taxonomic

associations stable over time in a human-modified landscape? Ecological

Indicators, 69, 152-164.

25

I. Yong, D. L., Barton, P. S., Okada, S., Crane, M., & Lindenmayer, D. B. (2016).

Birds as surrogates for mammals and reptiles: Are patterns of cross-

taxonomic associations stable over time in a human-modified landscape?

Ecological Indicators, 69: 152-164.

26

Birds as surrogates for mammals and reptiles: are patterns of

cross-taxonomic associations stable over time in a human-

modified landscape?

Abstract

Cross-taxonomic surrogates can be feasible alternatives to direct measurements of

biodiversity in conservation if validated with robust data and used with explicit

goals. However, few studies of cross-taxonomic surrogates have examined how

temporal changes in composition or richness in one taxon can drive variation in

concordant patterns of diversity in another taxon, particularly in a dynamic and

heavily modified landscape. We examined this problem by assessing changes in

cross-taxonomic associations over time between the surrogate (birds) and target

vertebrate taxa (mammals, reptiles) that demand high sampling effort, in a

heterogeneous mosaic landscape comprising pine monoculture, eucalypt woodland

remnants and agricultural land. Focussing on four study years (1999, 2001, 2011, 2013)

from a dataset spanning 15 years, we: (1) investigated temporal changes in cross-

taxonomic congruency among three animal taxa, (2) explored how temporal

variation in composition and species richness of each taxon might account for

variation in cross-taxonomic congruency, and (3) identified habitat structural

variables that are strongly correlated with species composition of each taxon. We

found the strength of cross-taxonomic congruency varied between taxa in response

to both landscape context and over time. Among the three taxa, overall correlations

were weak but were consistently positive and strongest between birds and

mammals, while correlations involving reptiles were usually weak and negative. We

also found that stronger species richness and composition correlations between

birds and mammals were not only more prevalent in woodland remnants in the

agricultural matrix, but they also increased in strength over time. Temporal shifts in

species composition differed in rate and extent among the taxa even though these

27

changes were significant over time, while important habitat structural correlates

were seldom shared across taxa. Our study highlights the role of the landscape

matrix and time in shaping animal communities and the resulting cross-taxonomic

associations in the woodland remnants, especially after a major perturbation event

(i.e. plantation establishment). In such dynamic landscapes, differing and taxon-

specific shifts in diversity over time can influence the strength, direction and

consistency of cross-taxonomic correlations, therefore posing a ‘temporal’ problem

for the use of surrogates like birds in monitoring and assessments of biodiversity,

and conservation management practices.

1. Introduction

Land-use change increasingly threatens biodiversity globally by driving habitat loss

and degradation (Sala et al. 2000; Reidsma et al. 2006; Sayer et al. 2013). As a result,

there is an urgent need to understand how diverse groups of biota respond to land-

use modification across various scales (e.g., Mattison & Norris 2005; Haines-Young

2009). Such knowledge is integral to informing decisions on how sites should be

conserved and managed (Meir et al. 2004; Vandewalle et al. 2010). However,

resource and taxonomic limitations impose enormous difficulties on sampling large

suites of taxonomic groups (Lawton et al. 1998; Schulze et al. 2004; Gardner et al.

2008) to understand broad changes in biodiversity patterns. This has resulted in

multiple surrogate approaches being developed to act as proxies for components of

biodiversity not able to be directly measured (Prendergast & Eversham 1997; Caro

2010; Lindenmayer et al. 2015), or biota that are costly or logistically difficult to

survey within time frames available for decision-making (Favreau et al. 2006).

Species-based surrogates of biodiversity are a common type of surrogate (e.g., Caro

2010), and are based on the hypothesis that the occurrence or diversity of a surrogate

or indicator taxon reflects the occurrence (i.e. co-occurrence) or diversity (i.e.

28

richness, composition) of other sets of target taxa (Rohr et al. 2006; Rondinini et al.

2006; Gaspar et al. 2010). The best examples of these species surrogates include

cross-taxonomic surrogates (e.g., Kati et al. 2004; Gallardo et al. 2011; Gaspar et al.

2010; Fattorini et al. 2012), biodiversity indicator species or species groups (e.g., Nally

& Fleishman 2002; Roberge & Angelstam 2004; Branton & Richardson 2011), and

higher-taxonomic groups (e.g., Báldi 2003; Heino & Soininen 2007).

Species surrogates of diversity in conservation have several empirical and conceptual

shortcomings (e.g., Andelman & Fagan 2000; Heink & Kowarik 2010). First, studies of

cross-taxonomic relationships have yielded mixed results in terms of the strength

and direction of congruency across different taxa, often varying with the analytical

approaches used (Gioria et al. 2011), even when landscape contexts and scales are

broadly similar (Wolters et al. 2006; Lewandowski et al. 2009). At small to

intermediate spatial scales of study, cross-taxonomic congruency of species richness

was found to be weak in some studies (e.g., Kati et al. 2004; Lovell et al. 2007) but

strong in others (e.g., Negi & Gadgil 2002). Such divergent findings are further

exacerbated by the fact that these surrogates are often used to predict occurrence

and diversity of target taxa with different ecological attributes (e.g. dispersal ability,

habitat requirements, life histories) (e.g., Ricketts et al. 1999). Second, many studies

testing surrogacy relationships with respect to a biodiversity target are not clearly

defined within a theoretical framework, thus weakening the ecological basis for

using a surrogate (Belovsky et al. 2004; Lindenmayer & Likens 2010). Many studies

emphasise the identification of cross-taxonomic surrogate associations, but fail to

define the surrogate relationships clearly, or under a robust framework that

incorporate cause-effect relationships and predictive strength (Barton et al. 2015).

Others like Hunter et al. (2016) has pointed out controversies arising from surrogate

concept as a result of differing goals of surrogate application in conservation. Third,

many studies of surrogates are ‘snapshot’ investigations and fail to tackle the

problem of how species surrogates perform over time, or with respect to temporal

variability in ecological processes (Anderson 2001; Favreau et al. 2006; Magurran et

al. 2010). For any biodiversity surrogate to function as a useful tool for conservation,

29

it should consistently predict diversity patterns or responses of other species over

time (Rodrigues et al. 2000). Understanding of how biodiversity surrogates perform

over time (Favreau et al. 2006) is constrained by the paucity of long-term datasets,

with the result that few studies (e.g., Thomson et al. 2007) have examined how long-

term shifts in the composition of animal communities associated with landscape

modification may affect cross-taxonomic congruency (see Table 1 for definitions).

Biodiversity patterns in general, and individual species in particular, respond to the

extent of landscape modification in different and diverse ways (Fischer &

Lindenmayer 2007). Typically, modification of a landscape leads to changes in

habitat spatial configuration and structure (e.g., patch size, matrix quality, edge

effects), which impact animal communities differently, depending on individual

species’ ecological needs and their ability to disperse across the wider landscape

(Dormann et al. 2007; Driscoll et al. 2013). Over time, species composition in a biotic

community can be affected by dynamic changes in landscape configuration and

vegetation structure or habitat recovery post-disturbance (e.g., Guedo & Lamb 2013).

While it remains unclear how shifts in community composition of one taxon

changes relative to other taxa, a taxonomic group can act as a good surrogate for

others if it undergoes turnover (see Table 1 for definition) in species richness or

compositional patterns that are consistent and congruent with other taxa over space

and time. For instance, strong patterns of congruency between turnover of

invertebrate and macroalgal diversity highlight the potential of macroalgae

assemblages to act as biodiversity surrogates for fish and invertebrates (Thomson et

al. 2014)

In this study, we investigated temporal variation in cross-taxonomic congruency (see

Table 1 for definitions) of diversity between pairs of three taxa, and explored how, (a)

temporal shifts in diversity and, (b) habitat correlates specific to each animal taxon

can drive variation in the extent of cross-taxonomic congruency. We used a large

dataset that has been collected over a period of 15 years in a dynamic, human-

modified landscape that has undergone rapid transformation from a woodland-

30

agriculture mosaic to large tracts of pine monoculture (Lindenmayer et al. 2001;

Lindenmayer et al. 2008). We focussed on birds, mammals and reptiles as these taxa

are not only frequently used in conservation assessments (e.g., Westgate et al. 2014),

but are also species-rich in our study landscape (See Table A6, A7, A8 for list of

species). In addition, sampling these three taxa demands very different amounts of

effort and resources given the nature of field surveys. For instance, birds can be

easily surveyed and have found to be popular and cost-effective surrogates in

inventories of biodiversity (e.g., Lawton et al. 1998; Gardner et al. 2008) whereas

sampling reptile diversity not only involves a very different methodology, but also

demands specialist knowledge (e.g., McDiarmid et al. 2011). For mammals, the

nocturnal habits and cryptic behaviour of many species (e.g., Suter et al. 2000)

means effort-intensive night surveys and baited traps are needed to survey them.

Differences in natural history across taxa, and disparate sampling effort to be

invested in different taxonomic groups underscores the need for viable biodiversity

surrogates, which could facilitate more optimal use of resources in inventorying

biodiversity.

The aim of our study was to evaluate congruence in diversity and species

composition measures between birds, mammals and reptiles over time, and thus

uncover evidence for consistent cross-taxonomic surrogacy (Table 1 for definitions),

as the quantification of cross-taxonomic congruency is a critical step in identifying

surrogates (Gioria et al. 2011) . To quantify cross-taxonomic congruency, we used

metrics of correlation between species richness and species composition, as both

measures are frequently adopted in studies of cross-taxon surrogates (e.g., Kati et al.

2004; Sauberer et al. 2004; Gaspar et al. 2010; Cabra-García et al. 2012) and

collectively can offer a comprehensive evaluation of cross-taxonomic congruency (Su

et al. 2004; Gioria et al. 2011). To address our study aims, we posed three questions:

1. Based on the strength and direction of associations between pairs of taxa,

what is the extent of variation in cross-taxonomic congruence patterns at the

species richness and composition levels over 15 years?

31

Given the limited vagility of reptiles, small spatial requirements (Stow et al. 2014)

and the limited effect posed by habitat fragmentation on lizard communities (e.g.,

Jellinek et al. 2004) compared to birds or mammals, we predicted that reptiles were

likely to show low congruency in diversity patterns with either mammals or birds.

Cross-taxonomic congruency patterns are often derived from measures of diversity

and thus determined by temporal shifts in the diversity of different taxonomic

groups relative to each other. To, (a) explore the extent of temporal variation in

diversity across the taxonomic groups and, (b) determine how different habitat

structural variables in remnant woodlands can influence each taxon in our study, we

asked:

2. In terms of species richness, abundance and composition, what is the extent

of temporal change in three animal taxa over 15 years?

3. Are the habitat structure variables that best predict patterns of species

composition common to all three taxa?

Based on our findings, we discuss how variation in the predictive strength of

surrogates for other aspects of biodiversity (e.g., other taxonomic groups) can be

influenced by taxon-specific temporal shifts and habitat conditions, as well as

implications for the use of cross-taxonomic surrogates in conservation assessments,

inventorying and monitoring.

2. Materials and methods

2.1. Study region

Our study was conducted in the Nanangroe region (34°57'54''S, 148°28'46''E) near

Jugiong and Gundagai, Central New South Wales, Australia. Nanangroe is a dynamic

landscape spanning c. 30,000 ha of agricultural (i.e., grazing) land and exotic tree

plantations. Nanangroe was established as a long-term natural experiment to

understand how animal communities respond to differing landscape treatments over

32

time (Lindenmayer et al. 2008). Much of the original Eucalyptus-dominated, box-

gum grassy woodland landscape has been cleared for agriculture in the past two

centuries (Yates & Hobbs 1997), leaving what is best described as a variegated

landscape consisting of distinct patches and strips of remnant woodlands of varying

tree densities (McIntyre & Barrett 1992) surrounded by a larger matrix of pastures

grazed by livestock. These woodland remnants are dominated by five Eucalyptus

species: white box (E. albens), red box (E. polyanthemos), yellow box (E. melliodora),

red stringybark (E. macrorhyncha) and Blakely’s red gum (E. blakelyi), while the

understorey supports a diverse community of native and introduced grasses and

forbs.

Prior to the commencement of the Nanangroe Natural Experiment in 1999, 52

Eucalypt-woodland remnants were identified using two landscape contexts and four

patch sizes classes (0.5-0.9 ha; 1.0-2.4 ha; 2.5-4.9 ha; 5.0-10 ha). In 1998, the

agricultural matrix landscape surrounding these 52 woodland remnants was

transformed by the establishment of dense plantations of the exotic Monterey Pine

Pinus radiata (hereafter these remnants are referred to as “woodland remnants in

pine matrix”). In addition, sampling points in 56 patches of Eucalypt-woodland

remnants of broadly similar vegetation classes and areas were established in

surrounding agricultural land (hereafter these remnants are referred to as

“woodland remnants in agricultural matrix”), mostly on farms under private

ownership (see Table A1 in the supplementary material for definitions on landscape

contexts). Additionally, 10 sites in cleared and grazed paddocks and 10 sites in pine

plantations were established as “controls”. Inclusive of these two sets of control sites,

there were a total of four landscape contexts examined in our study.

2.2. Animal sampling

Permanent transects were marked and established at all 128 study sites prior to the

commencement of the study in 1999. In woodland remnants exceeding one hectare

33

in area, a straight 200m long transect was established. For a few small remnants less

than one hectare in area, a ‘dog-legged’ 200m or 150m transect was established.

We sampled bird diversity and abundance at each site using three, five-minute point

counts along each transect, which were conducted between 05:00–10:00hrs during

early-middle spring (October-November). At each point count, observers recorded

the numbers of individual species heard or seen within a 50m radius. Each point was

re-sampled by a different observer on another day during the survey period to

minimize bias as a result of weather and variable detection skills by different

observers. Bird surveys were conducted in the years, 1999, 2000, 2001, 2003, 2005,

2007, 2009, 2011, 2012 and 2013.

To survey reptile abundance and diversity, we conducted standardised, area-

constrained searches at two points along each transect once a year between late

winter to early spring (October–November). During the establishment of the

transects, artificial substrates consisting of corrugated metal sheets (c. 1.0m x 1.0m),

hardwood timber sleepers (c. 1.0m long, 0.2m thick) and roof tiles (c. 0.3m x 0.3m)

were placed at the 0m and 100m points along each transect to simulate

microhabitats for small terrestrial reptiles like snakes, skinks and other lizards.

Active searches for reptiles were completed by turning over logs, rocks and the

artificial substrates throughout the sites. Standardised reptile surveys were

conducted in the years: 1999, 2000, 2001, 2003, 2006, 2011 and 2013.

Finally, we surveyed mammal diversity and abundance using standardised,

nocturnal spotlighting searches along each transect, on nights of good weather (i.e.

no rain, storms). Mammal spotlight surveys were conducted in the years: 1999, 2001,

2004, 2005, 2009, 2011 and 2013. Additional details on our mammal and reptile

surveys have been described in Lindenmayer et al. (2001) and (2008).

2.3. Vegetation sampling

34

To describe the habitat structure at each study site, we conducted vegetation surveys

at all study sites once every four years. A total of 34 vegetation variables was

measured at each site to capture the variation in vegetation structure from the

ground to the canopy. We averaged measures taken from each of three sampling

points to obtain mean values for all habitat structure variables at every site. A full

list of the vegetation variables is available in the supplementary information section

(Table A2).

2.4. Data analysis

2.4.1. Data selection

We used species data from surveys of birds, mammals and reptiles completed in

1999, 2001, 2011 and 2013. Each of these years were selected for our analysis as they

included data where all three taxonomic groups were simultaneously surveyed in the

same year and season, and therefore minimized the influence of temporal effects on

our dataset.

2.4.2. Tests for correlations of species richness between different taxa over time

(Question 1)

We used Spearman’s rank correlations to test for cross-taxonomic congruence in

species richness patterns over time between pair-wise combinations of the three taxa

for each of four study years and each landscape context class (including both control

sites). The strength of correlation of species richness between two taxa is often used

as a proxy of cross-taxonomic associations (e.g., Hess et al. 2006; Wolters et al.

2006). Spearman’s correlation was chosen over Pearson’s correlation as the metric of

correlation strength as species richness was relatively low across sites, particularly

for mammals, and is thus likely to be distributed non-normally. We also calculated

correlations between birds, and pooled species richness of mammals and reptiles

combined. Using 1,000 bootstrap replicates, we calculated the 95% confidence

35

interval for all Spearman’s correlations. The strength of the Spearman’s correlation

coefficient (ρ), which is used as a measure of congruency of species richness between

two taxa was interpreted as follows: correlation values of ≥ 0·50 were considered to

be strong, between 0.10 to 0·30 to be moderate, and correlations ≤ 0·10 to be weak

(see Lamoreux et al. 2006).

2.4.3. Test for correlations of species composition over time (Question 1)

We used partial Mantel tests to investigate the strength of cross-taxonomic

congruence in species composition between pair-wise combinations of animal taxa

for each year of four study years. Partial Mantel tests were used because the data

were not independent and Mantel tests are able to address the problem of partial

dependence in dissimilarity matrices (Legendre & Legendre 1998), and have

previously been used to identify correlations between pairs of taxa (e.g., Su et al.

2004; Gioria et al. 2011; Gaspar et al. 2012). Abundance values for all species were

square-root transformed to reduce the potential over-influence of highly abundant

species on among-site dissimilarity values. We quantified species composition using

Bray-Curtis dissimilarity metric between pairs of sites for all landscapes contexts.

The advantage of partial Mantel tests over simple Mantel tests is that they can

measure the correlation between two matrices (Paszkowski & Tonn 2000; Su et al.

2004) after considering variation associated with a matrix of spatial (Euclidean)

distances, thus accounting for potential problems of spatial autocorrelation.

Significance of all partial Mantel tests was assessed using a Monte Carlo procedure

with 999 permutations. Mantel and Spearman’s correlations were implemented

using the ‘ecodist’ package in R version 1.2.9, while confidence intervals for

Spearman’s correlations were estimated using 1,000 bootstraps in the

‘RVAideMemoire’ package (R Development Core Team 2013).

2.4.4. Test of species composition of two taxa as predictors over time (Question 1)

36

We completed multiple regressions on distance matrices (MRM) (Lichstein 2007) to

test if species composition of two taxa based on dissimilarity matrices can

collectively better predict composition of a target taxa selected a priori. Unlike

partial Mantel tests which are limited to comparing pairs of taxa, this approach

allows multiple taxa to be used as predictor variables. MRM involves regressing the

response matrix using more than one explanatory matrix, while each matrix contains

all combinations of pair-wise distances between n number of sample units. We

chose not to use bird data as the response variable in any of our MRM models. This

was because birds are usually the surrogate taxon in conservation of other

components of biodiversity (e.g., Blair 1999; Sauberer et al. 2004; Larsen et al. 2012)

given the relative ease of collecting bird data compared to data of other taxa.

Additionally, we factored geographic distance into our models as a predictor matrix,

since spatial data derived from geographic coordinates are often available along with

species datasets and can be used to reveal ecologically meaningful effects (e.g.,

strong spatial influences on composition may reveal dispersal limitations imposed

by space).

We constructed a set of candidate models using all possible combinations of bird,

reptile and mammal composition and spatial distances as predictor variables, while

only mammal or reptile composition was treated as the response. MRM analysis was

completed only for species data collected from woodland remnant sites in pine and

agricultural sites as there were too few data for analysis in the control sites due to

low species abundance and richness. As with our partial Mantel tests, we square-

root transformed the animal count data, and used the Bray-Curtis metric to calculate

pair-wise species dissimilarity. The statistical significance of each MRM model was

assessed with 999 permutations.

2.4.5. Analysis of shifts in animal communities over time (Question 2)

37

We plotted site-level, mean species richness and mean abundance for each taxon in

both landscape contexts and control sites to assess temporal changes in species

richness and abundance over the four study years. To visualise changes in

community composition between the four landscape contexts over the four study

years, we first performed non-metric multidimensional scaling (NMDS) analysis

using the R function ‘metaMDS’ to ordinate site counts in species space for all three

groups and the two main landscape contexts (woodland remnants in agriculture and

pine). For each landscape context, all ordinations of each taxon were presented

together in each plot, but separated by year using coloured polygons. We then used

the multiple response permutation procedure (MRPP) as a non-parametric test for

significant differences in species compositional changes over time. MRPP generates

the effect size statistic A, which provides a measure of within-group heterogeneity,

and a measure of significance P. The significance of the effect size A was assessed

using 999 permutations. Bray-Curtis dissimilarity was used as the measure for

species composition in both the NMDS and MRPP analyses.

To explore how different habitat structural variables influenced each of the three

taxa in ordination space, we fitted vectors for all habitat variables measured in each

ordination, to identify those that were significantly correlated to the two NMDS axes

for each taxon. The R function ‘envfit’ available in the vegan package computes

vectors or factor averages of environmental variables fitted to the ordination matrix.

The significance of these fitted vectors was then assessed using 999 permutations.

Habitat correlates that were significant at P < 0.05, and marginally significant 0.05 <

P ≤ 0.1 were retained for further consideration.

2.4.6. Evaluating the influence(s) of habitat structural correlates on animal

communities (Question 3)

We were interested in identifying habitat structural variables consistently associated

with species composition among the three taxa in the Eucalypt-woodland remnants.

38

We constructed a series of candidate ‘global’ models using multiple regressions on

distance matrices for each taxa, and using the full set of habitat structural variables

to explore how the different variables influenced each taxon. Only habitat variables

not strongly correlated with others (Pearson’s r < 0.5) were retained in the MRM

analysis after an initial screening of the full set of variables in a correlogram matrix.

NMDS and MRPP analyses were completed using the ‘vegan’ package in R version

2.2-1 (R Development Core Team 2013) while MRM analysis was carried out using the

‘ecodist’ package in R version 1.2.9 (R Development Core Team 2013)

3. Results

3.1. What is the extent of variation in cross-taxonomic congruence patterns at the

richness and composition over 15 years (Question 1)?

3.1.1. Change in correlations of species richness over 15 years

We found that correlations of species richness varied between different pairs of taxa

and across landscape contexts, but increased in strength and significance over the 15

years (Figure 1, Table A3). In woodland remnants in the agricultural matrix, species

richness was weakly and negatively correlated between reptiles and birds, but none

of these correlations were significant (see supplementary material). Mammal species

richness was weakly and negatively correlated with that of birds in 1999, but the

correlations became positive and strengthened over time, with mammal species

richness being significantly correlated with bird species richness in 2011 and 2013

(Spearman’s ρ = 0.306 with P = 0.022; Spearman’s ρ = 0.350 with P = 0.01) but not in

1999 and 2001. Additionally, a linear model relating year to correlations of species

richness for bird–mammal congruency was significant (model adjusted R-square =

0.998, coefficient estimate = 0.0297, P = 0.0007). Species richness correlations

39

between birds and reptiles, and pooled mammal and reptile richness were weak and

insignificant for all years except in 2011 (Spearman’s ρ = 0.308 with P = 0.022).

In woodland remnants in the pine plantation matrix, bird species richness was

consistently and positively correlated with that of mammals, and the strength of

these correlations increased with time, with correlations in 2013 being marginally

significant (Spearman’s ρ = 0.277 with P = 0.065). In addition, bird species richness

was positively correlated with pooled mammal and reptile species richness in later

years, being significantly so in 2013 (Spearman’s ρ = 0.300 with P = 0.04).

3.1.2. Change in correlations of species composition over 15 years

We found that partial Mantel correlations between distance matrices of animal

groups were often weak and insignificant (Table 2, Figure 2). For instance, in

woodland remnants in the agricultural matrix, bird and reptile composition was

negatively correlated in all study years except in 2001. Bird and mammal

composition were mostly positively correlated over the four study years, although

only correlations in later years – 2011 and 2013 were moderately strong and

significant (Mantel R = 0.306 with P = 0.002; Mantel R = 0.168 with P = 0.008). None

of the correlations between reptile and mammal composition were strong or

significant, and fluctuated between being weakly positive and negative over time.

In woodland remnants in the pine plantation matrix, bird and reptile composition

were positively correlated only in 1999 (Mantel R = 0.1912 with P = 0.035), but

negatively correlated in all other years. Although consistently positive, we found

that correlations of bird and mammal composition were weak and insignificant

across all study years except 2001 (Mantel R = 0.279 with P = 0.012). None of the

correlations between mammals and reptiles were significant, and were mostly

negative. Overall, we found that while correlations involving reptiles were usually

40

negative and weak (Figure 2), correlations between mammal and bird species

composition were consistently positive, and appeared to have strengthened over

time, at least for woodland remnants in the agricultural matrix (2011 Mantel R =

0.306 with P = 0.002). Such a trend did not apply for woodland remnants in the pine

matrix, as correlation strength peaked in 2001, but declined thereafter.

3.1.3. Change in predictive strength of two taxa for a single target animal group over 15

years

We found that MRM models incorporating distance matrices of birds, reptiles and

spatial distances were able to predict mammalian composition, albeit weakly for

woodland remnants sites in the agricultural matrix, but relationships declined in

predictive strength between 2011 (R2 = 0.182) and 2013 (R2 = 0.0565) (Table 3). Of

three explanatory variables including reptile composition and spatial distances, bird

species composition explained 85.9% of the variation in mammal species

composition in 2011 but only 42.5% in 2013, although bird composition remained

significant as a predictor in both years. Reptile composition and spatial distance

were weak and non-significant predictors in all candidate models explaining

mammal composition in 2011 and 2013. For candidate models using birds, mammals

and spatial distance to predict reptile composition, bird and mammal composition

never emerged as significant predictor, being weakly but positively correlated in

most years. However, spatial distance appeared to be a significant and relatively

important predictor of reptile composition, explaining 48.3% and 55.9% of the

variation of reptile composition in 2011 and 2013 respectively (Table 3).

All candidate models incorporating species compositional and spatial distances for

woodland remnants in the pine matrix explained very little variation in either reptile

or mammal composition. Although bird species composition explained 31.0% of the

variation in mammalian composition in 2013, it was not a significant predictor in

other years, and in fact was negatively correlated in 2011. Neither mammal nor bird

41

species composition with spatial distances were useful predictors of reptile species

composition in woodland remnants in the pine matrix, although mammal species

composition was marginally significant as a predictor in 2011 and 2013 (0.05 < P <

0.1).

3.2. What is the extent of temporal changes in three animal communities over 15

years (Question 2)?

3.2.1. Changes in species richness and abundance of three taxa over 15 years

Across the study landscape, we found that mammal and reptile species richness

showed clear increases over the four study years, while bird species richness

increased marginally between 1999 and 2011, but declined in 2013 (Figure 3,

Supplementary Tables A6, A7, A8). Species richness and abundance, and their

change over 15 years in both pine and agricultural control sites were limited

especially for mammals and reptiles, and consistently lower than corresponding

woodland sites in either landscape contexts. At a site level, we found weak and

insignificant patterns of change in bird species richness for woodland remnants in

the pine matrix over time (Figure 3) while mean site abundance increased from 57.7

to 68.9 individuals (Mann-Whitney U = 693, Z = 1.65, P > 0.05). By comparison, bird

species richness in woodland remnants in the agricultural matrix increased more

rapidly over time, from 12.1 species in 1999 to 15.0 species per site (Mann-Whitney U

= 971, Z = -3.231, P < 0.05) in 2013. The trends in reptile richness and abundance over

time for woodland remnants in the pine matrix were less clear compared to those in

the agricultural matrix, but changed somewhat faster, and were significant. For

example, mean reptile species richness in woodland remnants in agriculture

increased from 0.357 species per site in 1999 to 2.52 species in 2013 (Mann-Whitney

U = 244, Z = 5.771, P < 0.001) while mean reptile richness in woodland remnants in

pine increased from 0.54 to 2.33 species over the same period (Mann-Whitney U =

971, Z = -3.231, P < 0.001). Unlike birds or reptiles, both mammal species richness and

42

abundance showed consistent increases in the two landscape contexts over the study

period. For example, mean mammal species richness for woodland remnants in the

pine matrix increased from 0.475 to 1.27 species in 2013 (Mann-Whitney U = 517, Z =

3.368, P < 0.001), while mean abundance increased from 0.775 individuals in 1999 to

2.4 individuals (Mann-Whitney U = 531, Z = 3.245, P < 0.05). Likewise, mean

mammal richness for woodland remnants in the agricultural matrix doubled over

the same period, from 0.518 species in 1999 to 1.03 species per site in 2013 (Mann-

Whitney U = 1076.5, Z = -2.6009, P < 0.01).

3.2.2. Changes in community composition of three taxa over 15 years

In woodland remnants in the pine matrix, we found that points representing

mammal species composition in ordination space clustered towards the negative

end of NMDS axis 1 in 1999, but became less clustered in subsequent years, and

shifted positively along the axis (Figure 4a). The MRPP results indicated that

mammal assemblages differed over the four years (A = 0.0464, P < 0.01). Points

representing reptiles were well spread in ordination space in 1999 (Figure 4b), but

became increasingly clustered towards the positive end of NMDS axis 1 in later years,

with these changes in species assemblage being significantly different over time (A =

0.0504, P < 0.01). Similarly, points representing birds were sparsely clustered in 1999,

but subsequently clustered closely towards the negative end of NMDS axis 1 in 2011

and 2013 (Figure 4c). Such a change in the bird assemblage over time was also found

to be significant in our MRPP analysis (A = 0.0741, P < 0.01) and suggests that bird

composition in these woodland remnants were become increasingly similar over the

15 years.

For woodland remnants in the agriculture matrix, points representing mammal

species were sparsely clustered in ordination space, and appeared to be even less so

in 2013 and 2011 than in 1999 and 2001 (Figure 4d). Although changes in the mammal

assemblage over time were significant, they were weaker than the species

43

compositional changes observed for the other two groups (A = 0.0193, P < 0.05), and

for all animal groups in woodland remnants in the pine matrix.

We did not plot the ordinations for reptiles due to the very large scatter of points.

However, we noted that reptile assemblages in woodland remnants in the

agricultural matrix changed significantly over 15 years (A = 0.103 P < 0.001) (see also

Figure A2 in the supplementary material for cluster dendrograms representing

differences in Bray-Curtis dissimilarity across the study years). Likewise, the bird

assemblage in these woodland remnants differed significantly over 15 years (A =

0.0222, P < 0.001). The change in bird species composition is shown in the positive

shift in clusters of points representing bird species composition in ordination space

along NMDS axis 1 (Figure 4e).

3.3. Are the habitat variables that drive shifts in species composition over time

shared among the three taxa (Question 3)?

3.3.1. Significant habitat structure variables correlated with each taxonomic group

We found that the habitat variables strongly correlated with species communities

differed among taxa and between woodland remnants in the two key landscape

contexts (Table 4, see also Table A5), and few variables were shared. Bird species

composition was correlated with more habitat structure variables than either

reptiles or mammals for woodland remnants in both landscape contexts. ‘Blackberry

cover’ was a recurrent explanatory variable for birds and reptiles, correlating

strongly with at least one NMDS axis for each group. In woodland remnants

surrounded by pine, ‘crown structure’ (R2 = 0.163, P < 0.05) and ‘basal count’ (R2 =

0.272, P < 0.01) were moderately and significantly correlated with both NMDS axes

for birds, while ‘blackberry cover’ and ‘woodland strata’ appeared only weakly

correlated. While ‘blackberry cover’ (R2 = 0.292, P < 0.01), ‘dominant cover’ (R2 =

44

0.226, P < 0.05) and ‘shrub stem count’ (R2 = 0.154, P < 0.05) appeared to be

important correlates for reptile species composition in woodland remnants

surrounded by pine, we found that no habitat variables were strongly and

significantly correlated to either NMDS axes for mammals.

Bird species composition in the woodland remnants in the agricultural matrix was

significantly correlated with eight habitat structure variables, with two of these

variables shared with mammals (‘exposed rock’, ‘blackberry cover’). While reptiles

were found not to be significantly correlated with any habitat variables in the

landscape contexts, we found that mammal species composition in woodland

remnants in the agricultural matrix was strongly correlated with four variables, with

‘blackberry cover’ (R2 = 1.00, P < 0.05) again being very strongly and positively

correlated with NMDS axis 1, and negatively with NMDS axis 2.

4. Discussion

4.1. Overview

We assessed the strength and direction of cross-taxonomic correlations in species

richness and composition between three vertebrate taxa that feature frequently in

conservation assessments (e.g., Lawton et al. 1998; Schulze et al. 2004; Westgate et

al. 2014). We then compared these associations over each of four study years spread

over 15 years to assess whether cross-taxonomic congruency was consistent over

time, a requisite of a good biodiversity surrogate. Below, we discuss our key findings

and outline some of the implications of our research for the use of cross-taxonomic

surrogates in dynamic landscapes undergoing rapid transformation.

45

4.2. Variation in cross-taxonomic congruency patterns over time

Whether based on a species richness or a species composition approach, we found

that the strength of congruency between pairs of animal taxa varied with the taxon

examined, landscape context, and over time (Question 1). Between pairs of taxa, we

found that correlations from both approaches ranged from being very weak to

moderate, and that correlations in either species richness or composition tended to

be positive and stronger between birds and mammals than between either group

and reptiles. We also found that species richness and composition correlations

increased in strength over time for woodland remnants in both agricultural and pine

matrixes. The prevalence of stronger and significant associations between taxa in

woodland remnants in the agricultural matrix in the later years compared to

woodland remnants in the pine matrix underscores the role played by the matrix in

shaping animal communities in remnant woodland patches (e.g., Ricketts 2001),

possibly by influencing the dispersal of different species (e.g., Dormann et al. 2007;

Driscoll et al. 2013). For instance, the stronger cross-taxonomic associations may

arise from greater dispersal into, and out of these woodland patches by species in all

three taxa through the comparatively more open agricultural matrix. Additionally,

the effects of the pine plantation matrix on animal communities in the Eucalypt-

woodland patches embedded within may be further accentuated by the limited food

resources available (e.g. flowering plants, arthropods) and a different set of

microclimatic conditions resulting from the dense pine cover.

4.3. Change in species richness and composition of animal communities over time

We found that species richness and abundances of mammals and reptiles showed

larger shifts than birds over time for woodland remnants in both pine and

agricultural matrixes. We also found that the extent of temporal change in species

composition differed with taxa, and between the two major landscape contexts

46

(Question 2). Our findings of taxa-specific shifts in diversity and abundance here

mirror the variation in congruency across taxa described earlier, and add yet another

line of evidence to the influences exerted by the landscape matrix on shaping the

animal communities occurring within these habitat patches. It is likely that

woodland remnants in the pine matrix showed lower cross-taxonomic congruence a

decade after the initial disturbance period (when pine monoculture was established)

because the dense pine plantation matrix may have acted as a barrier to the

dispersing reptiles and mammals (Mortelliti et al. 2014), thus influencing species

richness and composition of both taxa over time. Our finding here highlights the

problem posed by differential turnover in species diversity across taxonomic groups

to cross-taxonomic surrogacy because it compromises the temporal consistency

required if these surrogates are to be used in conservation monitoring and

biodiversity assessments.

4.4. Differing habitat structure variables correlated with animal taxa

Our analyses of the influence of habitat structure variables on animal taxa indicated

that the explanatory variables that fit best with the NMDS axes were different for

each group (Table 3) at the landscape scale, although one variable was frequently

shared (blackberry index). When an MRM approach was used to evaluate the

relative influence of habitat structure correlates, we again found that there were few

or no shared correlates between any two taxa. Other studies of cross-taxonomic

surrogates have also reported such differences of explanatory variables across taxa

(e.g., Dauber et al. 2003). Blackberry (Rubus fruticosus sp. agg.) is widely recognized

as one of the most invasive plant species across Australia (Dehaan et al. 2013) and

has increasingly spread across our study sites. Blackberry forms dense patches in

woodland remnants along creeks in our study sites and is likely to have modified

habitats and microclimatic conditions for many terrestrial species. This may account

47

for its strong correlation with the composition of all three animal taxa as revealed in

our ordination analysis (Figure 4).

We hypothesize that the broadly differing set of correlated habitat structure

variables identified in our analyses is the outcome of divergent habitat requirements

of birds, reptiles and mammals at the landscape scale. Bird species composition was

predicted by more habitat variables than reptiles or mammals in both landscape

contexts. This pattern is likely due to the fact that while the majority of reptile

species (e.g. skinks) and mammal species are more affected by habitat structural

variables on the ground, bird species composition are more strongly affected by a

larger set of habitat variables associated with trees (e.g. stand height, number of

trees, number of strata), due to the arboreal behaviour of many species (Barton et al.

2014). The differential associations of each taxa with specific sets of habitat

attributes and their changing relationship over time, may explain the weak cross-

taxonomic congruency observed in our study, and has also been highlighted by

other studies of cross-taxonomic associations (e.g., Dauber et al. 2003; Azeria et al.

2009; Heino et al. 2009).

4.5. Implications for the use of cross-taxonomic surrogates in conservation

Our findings have several key implications for the use of some vertebrate taxa,

particularly birds, as surrogates or broad indicators for the diversity of other taxa in

conservation. First, variation in species richness and composition over time and

among the taxa studied suggests that species richness and compositional approaches

to quantifying surrogates of species diversity should be applied cautiously. Our

finding that stronger cross-taxonomic associations in composition and species

richness occurred in woodland remnants in the agricultural matrix alludes to the

role played by the landscape matrix in shaping animal communities, either by

limiting or promoting species dispersal (e.g., Driscoll et al. 2013). Differences in

48

dispersal ability and spatial requirements may have influenced cross-taxonomic

associations at the landscape scale, and may explain why both birds and mammals

were better correlated with each other, but were often weakly and negatively

correlated with reptiles, which are not only predominantly terrestrial but less vagile,

and thus have smaller spatial requirements (e.g. Stow et al. 2014).

Second, our finding of stronger and more positive associations between bird and

mammal diversity in both landscape contexts over time suggests that animal

communities can become increasingly similar and more stable, possibly in response

to changes in vegetation structure as woodland remnants regenerate and mature in

the years following initial disturbance (e.g., change of the landscape matrix when

pines were planted). Strengthening of these cross-taxonomic relationships may be

also paralleled by increases in mean species richness at the site level for both taxa

(Figure 3). Communities in heavily modified landscapes are likely to show lower

community stability and higher temporal turnover in species composition. However

these communities can become more stable with time post-disturbance (Leibold

2009) and with increased overall species diversity (van Ruijven & Berendse 2007).

We hypothesise that increased community stability and higher diversity at the

landscape scale may have a role in driving stronger cross-taxonomic congruency at

the species richness and composition levels observed in our study in 2011 and 2013,

and suggest that cross-taxonomic surrogates may not be very useful for assessing

biodiversity in landscapes that have recently been subject to heavy anthropogenic

disturbance.

Third, our findings suggest that high rates of taxa-specific turnover and among-

group differences in habitat correlates, can affect the degree of congruency in

diversity patterns between different taxa. For example, birds showed significant

shifts in species composition over time, but with little increase in richness or

abundance. By contrast, the reptile communities showed significant temporal

49

turnover, and increases in overall diversity and abundance (Figure 3). Differing rates

of temporal turnover shown by change in Bray-Curtis dissimilarity over the study

period may account for the large variation in the Mantel correlations over time.

While many studies have explored cross-taxonomic congruency using large sets of

species data (e.g., Schulze et al. 2004; Grenyer et al. 2006; Stoch et al. 2009), we note

that few have examined congruency patterns in relation to temporal changes in

species richness, abundance and composition. This temporal problem continues to

persist because most surrogate studies are based on short-termed datasets (Favreau

et al. 2006). Therefore, we suggest that strong congruencies observed between two

taxonomic groups at one point in time may be ephemeral, especially in highly

disturbed landscapes undergoing change. Our results thus offer some support to the

predictions by Prendergast & Eversham (1997) that differential responses to the

environment (in this case, habitat structure variables), may be responsible for

driving weakly congruent patterns of diversity. From a conservation standpoint, the

use of one or few taxa as cross-taxonomic surrogates, especially birds, is likely to be

problematic since it could inherently fail to represent diversity patterns of other taxa

(e.g., Dauber et al. 2003) and their responses to changing habitat structure (e.g.,

Barton et al. 2014).

4.6. Ecological basis of surrogacy relationships and scope for future research

Snapshot-type studies of cross-taxonomic surrogates are ubiquitous in the literature

but lack a temporal dimension, thus failing to take into consideration ecological

processes that take time to manifest (e.g. Bond 2001, Favreau et al. 2006). Since

many ecological patterns and processes are highly dynamic in time and space

(Morgan et al. 1994), short-term studies will inherently fail to capture the temporal

variability of communities and their effects on cross-taxonomic comparisons.

Moreover, many such surrogate studies are also conducted at scales too large for

surrogacy patterns to be meaningful for conservation (Westgate et al. 2014), often at

50

a continental to global scale. However, Grenyer et al. (2006) and others (e.g. Weibull

et al. 2003) have noted that congruency between taxa tends to be highly scale

dependent; levels of congruency may be particularly low if these patterns are

measured at the fine spatial resolutions relevant to conservation. There is thus a

need for more studies of cross-taxonomic surrogacy at these fine spatial scales which

these surrogates are to be applied.

Our findings of stronger associations at the species richness and composition level

between mammals and birds, both which are known to be better dispersers and have

larger spatial requirements than reptiles, underscores the role of dispersal and

spatial scale in shaping animal communities (e.g., Howeth & Leibold 2010). These

ecological factors needs to be considered when identifying species surrogates for

conservation application in dynamic landscapes. Our findings also raise problems

for the efficacy of using biodiversity surrogates in dynamic, human-modified

landscapes because cross-taxonomic congruency changes over time with temporal

shifts in diversity (e.g., Wolters et al. 2006).

Finally, an immediate goal for ecologists studying indicators of biodiversity should

be to identify clearer links between different taxonomic groups and in relation to

underlying ecological processes, to ensure that taxa used as surrogates are grounded

within a more robust, science-driven framework that considers causal links that

allows for validation across spatial and temporal contexts (e.g., Lindenmayer &

Likens 2011; Barton et al. 2015). Identifying shared responses and relationships to

landscape and habitat structure variables between species, and between different

taxa could be a first step in understanding these associations in a mechanistic

manner. This, in turn, needs to be coupled with a better understanding of how

temporal processes may alter these relationships, although doing so will demand

greater investments into collecting long-term data.

51

Acknowledgments

We thank three anonymous reviewers and Roland Achtziger whose comments have

helped to improve this manuscript greatly. DLY thanks the Lesslie Foundation for

support, and Martin Westgate for useful comments during the preparation of the

manuscript. DBL was funded by an Australian Research Council Laureate

Fellowship. We thank all the landowners for retaining patches of woodland and

granting us permission to conduct wildlife and vegetation surveys on their farm

properties.

Literature cited

Andelman, S. J., & Fagan, W. F. (2000). Umbrellas and flagships: efficient

conservation surrogates or expensive mistakes? Proceedings of the National

Academy of Sciences 97(11): 5954-5959.

Anderson, M. J. (2001). A new method for non‐parametric multivariate analysis of

variance. Austral ecology 26(1): 32-46.

Azeria, E. T., Fortin, D., Lemaître, J., Janssen, P., Hébert, C., Darveau, M., &

Cumming, S. G. (2009). Fine‐scale structure and cross‐taxon congruence of bird

and beetle assemblages in an old‐growth boreal forest mosaic. Global Ecology and

Biogeography 18(3): 333-345.

Báldi, A. (2003). Using higher taxa as surrogates of species richness: a study based on

3700 Coleoptera, Diptera, and Acari species in Central-Hungarian reserves. Basic

and Applied Ecology 4(6): 589-593.

52

Barton, P. S., Westgate, M. J., Lane, P. W., MacGregor, C., & Lindenmayer, D. B.

(2014). Robustness of habitat‐based surrogates of animal diversity: a multitaxa

comparison over time. Journal of Applied Ecology 51(5): 1434-1443.

Barton, P. S., Pierson, J. C., Westgate, M. J., Lane, P. W., & Lindenmayer, D. B. (2015).

Learning from clinical medicine to improve the use of surrogates in ecology.

Oikos 124(4): 391-398.

Belovsky, G. E., Botkin, D. B., Crowl, T. A., Cummins, K. W., Franklin, J. F., Hunter,

M. L., Joern, A., Lindenmayer, D. B., MacMahon, J. A., Margules, C. R., & Scott, J.

M. (2004). Ten suggestions to strengthen the science of ecology. BioScience 54(4):

345-351.

Blair, R. B. (1999). Birds and butterflies along an urban gradient: surrogate taxa for

assessing biodiversity? Ecological Applications 9(1): 164-170.

Bond, W. (2001). Keystone Species–Hunting the Snark? Science 292(5514): 63-64.

Branton, M., & Richardson, J. S. (2011). Assessing the Value of the Umbrella‐Species

Concept for Conservation Planning with Meta‐Analysis. Conservation Biology

25(1): 9-20.

Cabra-García, J., Bermúdez-Rivas, C., Osorio, A. M., & Chacón, P. (2012). Cross-taxon

congruence of α and β diversity among five leaf litter arthropod groups in

Colombia. Biodiversity and Conservation 21(6): 1493-1508.

Caro, T. (2010). Conservation by proxy: indicator, umbrella, keystone, flagship, and

other surrogate species. Island Press, Washington D.C.

Dauber, J., Hirsch, M., Simmering, D., Waldhardt, R., Otte, A., & Wolters, V. (2003).

Landscape structure as an indicator of biodiversity: matrix effects on species

richness. Agriculture, Ecosystems & Environment 98(1): 321-329.

Dehaan, R., Louis, J., Wilson, A., Hall, A., & Rumbachs, R. (2007). Discrimination of

blackberry (Rubus fruticosus sp. agg.) using hyperspectral imagery in Kosciuszko

53

National Park, NSW, Australia. ISPRS Journal of Photogrammetry and Remote

Sensing 62(1): 13-24.

Dormann, C.F., Schweiger, O., Augenstein, I., Bailey, D., Billeter, R., De Blust, G.,

DeFilippi, R., Frenzel, M., Hendrickx, F., Herzog, F., Klotz, S., Liira, J., Maelfait, J-

P., Schmidt, T., Speelmans, M., Van Wingerden, W. K. R. E., & Zobel, M. (2007).

Effects of landscape structure and land‐use intensity on similarity of plant and

animal communities. Global Ecology and Biogeography 16: 774-787.

Driscoll, D. A., Banks, S. C., Barton, P. S., Lindenmayer, D. B., & Smith, A. L. (2013).

Conceptual domain of the matrix in fragmented landscapes. Trends in Ecology &

Evolution 28(10): 605-613.

Fattorini, S., Dennis, R. L., & Cook, L. M. (2012). Use of cross-taxon congruence for

hotspot identification at a regional scale. PloS One 7(6): e40018.

Favreau, J. M., Drew, C. A., Hess, G. R., Rubino, M. J., Koch, F. H., & Eschelbach, K.

A. (2006). Recommendations for assessing the effectiveness of surrogate species

approaches. Biodiversity & Conservation 15(12): 3949-3969.

Fischer, J., & Lindenmayer, D. B. (2007). Landscape modification and habitat

fragmentation: a synthesis. Global Ecology and Biogeography 16(3): 265-280.

Gallardo, B., Gascón, S., Quintana, X., & Comín, F. A. (2011). How to choose a

biodiversity indicator–Redundancy and complementarity of biodiversity metrics

in a freshwater ecosystem. Ecological Indicators 11(5): 1177-1184.

Gardner, T. A., Barlow, J., Araujo, I. S., Ávila‐Pires, T. C., Bonaldo, A. B., Costa, J. E.,

Esposito, M. C., Ferreira, L. V., Hawes, J., Hernandez, M. I. M., Hoogmoed, M. S.

Leite, R. N., Lo, N. F. L. M., Malcolm, J. R., Martins, M. B., Mestre, L. A. M.,

Miranda-Santos, R., Overal, W. L., Parry, L., Peters, S. L., Ribeiro, M. A., Da Silva,

M. N. F., Motta, C. D. S. & Peres, C. (2008). The cost‐effectiveness of biodiversity

surveys in tropical forests. Ecology Letters 11(2): 139-150.

Gaspar, C., Gaston, K. J., & Borges, P. A. (2010). Arthropods as surrogates of diversity

at different spatial scales. Biological Conservation 143(5): 1287-1294.

54

Gioria, M., Bacaro, G., & Feehan, J. (2011). Evaluating and interpreting cross-taxon

congruence: potential pitfalls and solutions. Acta Oecologica 37(3): 187-194.

Grenyer, R., Orme, C. D. L., Jackson, S. F., Thomas, G. H., Davies, R. G., Davies, T. J.,

Jones, K. E., Olson, V. A., Ridgely, R. S., Rasmussen, P. C., Ding, T-S., Bennett, P.

M., Blackburn, T. M., Gaston, K. J., Gittleman, J. L., & Owens, I. P. (2006). Global

distribution and conservation of rare and threatened vertebrates. Nature 444

(7115): 93-96.

Guedo, D. D., & Lamb, E. G. (2013). Temporal changes in abundance–occupancy

relationships within and between communities after disturbance. Journal of

Vegetation Science 24(4): 607-615.

Haines-Young, R. (2009). Land use and biodiversity relationships. Land Use Policy

26: S178-S186.

Heink, U., & Kowarik, I. (2010). What are indicators? On the definition of indicators

in ecology and environmental planning. Ecological Indicators 10(3): 584-593.

Heino, J., & Soininen, J. (2007). Are higher taxa adequate surrogates for species-level

assemblage patterns and species richness in stream organisms? Biological

Conservation 137(1): 78-89.

Heino, J., Tolonen, K. T., Kotanen, J., & Paasivirta, L. (2009). Indicator groups and

congruence of assemblage similarity, species richness and environmental

relationships in littoral macroinvertebrates. Biodiversity and Conservation 18(12):

3085-3098.

Hess, G. R., Bartel, R. A., Leidner, A. K., Rosenfeld, K. M., Rubino, M. J., Snider, S. B.,

& Ricketts, T. H. (2006). Effectiveness of biodiversity indicators varies with extent,

grain, and region. Biological Conservation 132(4): 448-457.

Howeth, J. G., & Leibold, M. A. (2010). Species dispersal rates alter diversity and

ecosystem stability in pond metacommunities. Ecology 91(9): 2727-2741.

55

Hunter, M., Westgate, M., Barton, P., Calhoun, A., Pierson, J., Tulloch, A., Beger, M.,

Braunquinho, C., Caro, T., Gross, J., Heino, J., Lane, P., Longo, C., Martin, K.,

McDowell, W. H., Mellin C., Salo, H., & Lindenmayer, D.B. (2016). Two roles for

ecological surrogacy: Indicator surrogates and management surrogates. Ecological

Indicators 63: 121-125.

Jellinek, S., Driscoll, D. A., & Kirkpatrick, J. B. (2004). Environmental and vegetation

variables have a greater influence than habitat fragmentation in structuring lizard

communities in remnant urban bushland. Austral Ecology 29(3): 294-304.

Kati, V., Devillers, P., Dufrêne, M., Legakis, A., Vokou, D., & Lebrun, P. (2004).

Testing the value of six taxonomic groups as biodiversity indicators at a local

scale. Conservation biology 18(3): 667-675.

Lamoreux, J. F., Morrison, J. C., Ricketts, T. H., Olson, D. M., Dinerstein, E.,

McKnight, M. W., & Shugart, H. H. (2006). Global tests of biodiversity

concordance and the importance of endemism. Nature 440(7081): 212-214.

Larsen, F. W., Bladt, J., Balmford, A., & Rahbek, C. (2012). Birds as biodiversity

surrogates: will supplementing birds with other taxa improve effectiveness?

Journal of Applied Ecology 49(2): 349-356.

Lawton, J. H., Bignell, D. E., Bolton, B., Bloemers, G. F., Eggleton, P., Hammond, P.

M., Hodda, M., Holt, R. D., Larsen, T. B., Mawdsley, N. A., Stork, N. E., Srivastava,

D., S. & Watt, A. D. (1998). Biodiversity inventories, indicator taxa and effects of

habitat modification in tropical forest. Nature 391(6662): 72-76.

Legendre, P., & Fortin, M. J. (1989). Spatial pattern and ecological analysis. Vegetatio

80(2): 107-138.

Legendre, P., & Legendre, L. (1998). Numerical ecology. Second Edition.

Developments in Environmental Modelling 20. Elsevier BV, UK.

Leibold, M.A. (2009) Spatial and metacommunity dynamics in biodiversity. Pp 312-

319 In Levin, S. A., Carpenter, S. R., Godfray, H. C. J., Kinzig, A. P., Loreau, M.,

56

Losos, J. B., Walker, B. & Wilcove, D. S. (Eds.). The Princeton Guide to Ecology.

Princeton University Press, USA.

Lewandowski, A. S., Noss, R. F., & Parsons, D. R. (2010). The effectiveness of

surrogate taxa for the representation of biodiversity. Conservation Biology 24(5):

1367-1377.

Lichstein, J. W. (2007). Multiple regression on distance matrices: a multivariate

spatial analysis tool. Plant Ecology 188(2): 117-131.

Lindenmayer, D. B., & Likens, G. E. (2011). Direct measurement versus surrogate

indicator species for evaluating environmental change and biodiversity loss.

Ecosystems 14(1): 47-59.

Lindenmayer, D. B., Cunningham, R. B., MacGregor, C., Tribolet, C., & Donnelly, C.

F. (2001). A prospective longitudinal study of landscape matrix effects on fauna in

woodland remnants: experimental design and baseline data. Biological

Conservation 101(2): 157-169.

Lindenmayer, D. B., Fischer, J., Felton, A., Crane, M., Michael, D., Macgregor, C.,

Montague-Drake, R., Manning, A., & Hobbs, R. J. (2008). Novel ecosystems

resulting from landscape transformation create dilemmas for modern

conservation practice. Conservation Letters 1(3):129-135.

Lindenmayer, D. B., Cunningham, R. B., MacGregor, C., Crane, M., Michael, D.,

Fischer, J., Montague-Drake, R., Felton, A. & Manning, A. (2008). Temporal

changes in vertebrates during landscape transformation: a large-scale “natural

experiment”. Ecological Monographs 78(4): 567-590.

Lindenmayer, D. B., Barton, P. S., & Pierson, J. (Eds.). (2015). Indicators and

Surrogates of Biodiversity and Environmental Change. CSIRO Publishing,

Australia.

Lovell, S., Hamer, M., Slotow, R., & Herbert, D. (2007). Assessment of congruency

across invertebrate taxa and taxonomic levels to identify potential surrogates.

Biological Conservation 139(1): 113-125.

57

Magurran, A. E., Baillie, S. R., Buckland, S. T., Dick, J. M., Elston, D. A., Scott, E. M.,

Smith, R. I., Somerfield, P. J., & Watt, A. D. (2010). Long-term datasets in

biodiversity research and monitoring: assessing change in ecological communities

through time. Trends in Ecology & Evolution 25(10): 574-582.

Mattison, E. H., & Norris, K. (2005). Bridging the gaps between agricultural policy,

land-use and biodiversity. Trends in Ecology & Evolution 20(11): 610-616.

McDiarmid, R. W., Foster, M. S., Guyer, C., Gibbons, J. W., & Chernoff, N. (Eds.).

(2011). Reptile biodiversity: standard methods for inventory and monitoring.

University of California Press, USA.

McIntyre, S., & Barrett, G. W. (1992). Habitat variegation, an alternative to

fragmentation. Conservation Biology 6 (1): 146-147.

Meir, E., Andelman, S., & Possingham, H. P. (2004). Does conservation planning

matter in a dynamic and uncertain world? Ecology Letters 7(8): 615-622.

Morgan, P., Aplet, G. H., Haufler, J. B., Humphries, H. C., Moore, M. M., & Wilson,

W. D. (1994). Historical range of variability: a useful tool for evaluating ecosystem

change. Journal of Sustainable Forestry 2(1-2): 87-111.

Mortelliti A., Westgate M., & Lindenmayer D.B. (2014). Experimental evaluation

shows limited influence of pine plantations on the connectivity of highly

fragmented bird populations. Journal of Applied Ecology 51: 1179–1187.

Nally, R. M., & Fleishman, E. (2002). Using “indicator” species to model species

richness: model development and predictions. Ecological Applications 12(1): 79-92.

Negi, H. R., & Gadgil, M. (2002). Cross-taxon surrogacy of biodiversity in the Indian

Garhwal Himalaya. Biological Conservation 105(2): 143-155.

Paszkowski, C. A., & Tonn, W. M. (2000). Community concordance between the fish

and aquatic birds of lakes in northern Alberta, Canada: the relative importance of

environmental and biotic factors. Freshwater Biology 43(3): 421-437.

58

Prendergast, J. R., & Eversham, B. C. (1997). Species richness covariance in higher

taxa: empirical tests of the biodiversity indicator concept. Ecography 20(2): 210-

216.

R Development Core Team. (2013). R: A language and environment for statistical

computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-

900051-07-0, URL http://www.R-project.org/.

Reidsma, P., Tekelenburg, T., Van den Berg, M., & Alkemade, R. (2006). Impacts of

land-use change on biodiversity: an assessment of agricultural biodiversity in the

European Union. Agriculture, Ecosystems & Environment 114 (1): 86-102.

Ricketts, T. H. (2001). The matrix matters: effective isolation in fragmented

landscapes. The American Naturalist 158(1): 87-99.

Ricketts, T. H., Dinerstein, E., Olson, D. M., & Loucks, C. (1999). Who's where in

North America? Patterns of species richness and the utility of indicator taxa for

conservation. Bioscience 49(5): 369-381.

Roberge, J. M., & Angelstam, P. E. R. (2004). Usefulness of the umbrella species

concept as a conservation tool. Conservation Biology 18(1): 76-85.

Rodrigues, A. S., Gaston, K. J., & Gregory, R. D. (2000). Using presence–absence data

to establish reserve selection procedures that are robust to temporal species

turnover. Proceedings of the Royal Society of London B: Biological Sciences

267(1446): 897-902.

Rohr, J. R., Mahan, C. G., & Kim, K. C. (2007). Developing a monitoring program for

invertebrates: guidelines and a case study. Conservation Biology 21(2): 422-433.

Rondinini, C., Wilson, K. A., Boitani, L., Grantham, H., & Possingham, H. P. (2006).

Tradeoffs of different types of species occurrence data for use in systematic

conservation planning. Ecology letters 9(10): 1136-1145.

Ruijven, J. V., & Berendse, F. (2007). Contrasting effects of diversity on the temporal

stability of plant populations. Oikos 116(8): 1323-1330.

59

Sala, O. E., Chapin, F. S., Armesto, J. J., Berlow, E., Bloomfield, J., Dirzo, R., Huber-

Sanwald, E., Huenneke L. F., Jackson, R. B., Kinzig, A., Leemans, R., Lodge, D. M.,

Mooney, H. A., Oesterheld, M., Poff, N. L., Sykes, M. T., Walker, B. H., Walker,

M., & Wall, D. H. (2000). Global biodiversity scenarios for the year 2100. Science

287(5459): 1770-1774.

Santi, E., Maccherini, S., Rocchini, D., Bonini, I., Brunialti, G., Favilli, L., Perini, F.,

Pezzo, F., Piazzini, S., Rota, E., Salerni, E., & Chiarucci, A. (2010). Simple to

sample: vascular plants as surrogate group in a nature reserve. Journal for Nature

Conservation 18(1): 2-11.

Sauberer, N., Zulka, K. P., Abensperg-Traun, M., Berg, H. M., Bieringer, G.,

Milasowszky, N., Moser, D., Plutzar, C., Pollheimer, M., Storch, C., Trostl, R.,

Zechmeister, H., & Grabherr, G. (2004). Surrogate taxa for biodiversity in

agricultural landscapes of eastern Austria. Biological Conservation 117(2): 181-190.

Sayer, J., Sunderland, T., Ghazoul, J., Pfund, J. L., Sheil, D., Meijaard, E., Venter, M.,

Boedhihartono, A.K., Day, M., Garcia, C., van Oosten, C. & Buck, L. E. (2013). Ten

principles for a landscape approach to reconciling agriculture, conservation, and

other competing land uses. Proceedings of the National Academy of Sciences

110(21): 8349-8356.

Schulze, C. H., Waltert, M., Kessler, P. J., Pitopang, R., Veddeler, D., Mühlenberg,

M., Gradstein, S. R., Leuschner, C., Steffan-Dewenter, I., & Tscharntke, T. (2004).

Biodiversity indicator groups of tropical land-use systems: comparing plants,

birds, and insects. Ecological Applications 14(5): 1321-1333.

Stoch, F., Artheau, M., Brancelj, A., Galassi, D. M., & Malard, F. (2009). Biodiversity

indicators in European ground waters: towards a predictive model of stygobiotic

species richness. Freshwater Biology 54(4): 745-755.

Stow, A., Maclean, N., & Holwell, G. I. (Eds.). (2014). Austral Ark. Cambridge

University Press, UK.

60

Su, J. C., Debinski, D. M., Jakubauskas, M. E., & Kindscher, K. (2004). Beyond species

richness: Community similarity as a measure of cross‐taxon congruence for

coarse‐filter conservation. Conservation Biology 18(1): 167-173.

Suter, G. W., Efroymson, R.A., Sample, B.E. & Jones, D.S. (2000) Ecological risk

assessment for contaminated sites. CRC Press, USA.

Thomson, J. R., Fleishman, E., Nally, R. M., & Dobkin, D. S. (2007). Comparison of

predictor sets for species richness and the number of rare species of butterflies

and birds. Journal of Biogeography 34(1): 90-101.

Thomson, R. J., Hill, N. A., Leaper, R., Ellis, N., Pitcher, C. R., Barrett, N. S., & J.

Edgar, G. (2014). Congruence in demersal fish, macroinvertebrate, and macroalgal

community turnover on shallow temperate reefs. Ecological Applications 24(2):

287-299.

Vandewalle, M., De Bello, F., Berg, M. P., Bolger, T., Dolédec, S., Dubs, F., Feld, C.K.,

Harrington, R., Harrison, P. A., Lavorel, S., da Silva, P. M., Moretti, M., Niemela,

J., Santos, P., Sattler, P., Sousa, J. P., Sykes, M. T., Vanbergen, A. J. & Woodcock, B.

A. (2010). Functional traits as indicators of biodiversity response to land use

changes across ecosystems and organisms. Biodiversity and Conservation 19(10):

2921-2947.

van Weerd, M., & de Haes, H. A. U. (2010). Cross-taxon congruence in tree, bird and

bat species distributions at a moderate spatial scale across four tropical forest

types in the Philippines. Biodiversity and conservation 19(12): 3393-3411.

Weibull, A. C., Östman, Ö., & Granqvist, Å. (2003). Species richness in

agroecosystems: the effect of landscape, habitat and farm management.

Biodiversity & Conservation 12(7): 1335-1355.

Westgate, M. J., Barton, P. S., Lane, P. W., & Lindenmayer, D. B. (2014). Global meta-

analysis reveals low consistency of biodiversity congruence relationships. Nature

Communications 5: 3899.

61

Wiens, J. A., Hayward, G. D., Holthausen, R. S., & Wisdom, M. J. (2008). Using

surrogate species and groups for conservation planning and management.

Bioscience 58(3): 241-252.

Wolters, V., Bengtsson, J., & Zaitsev, A. S. (2006). Relationship among the species

richness of different taxa. Ecology 87(8): 1886-1895.

Yates, C.J. & Hobbs, R.J. (1997). Temperate eucalypt woodlands: a review of their

status, processes threatening their persistence and techniques for restoration.

Australian Journal of Botany 45: 949-973.

62

Figures

Figure 1. Plots showing variation in Spearman’s ρ (congruency of species richness)

over the four study years. Error bars represent 95% confidence intervals after 1,000

bootstraps. Diamond-shaped points represent woodland remnants in pine

(treatment) while circle-shaped points represent woodland remnants in the

agricultural matrix. Only the relationship between year and congruency of species

richness for bird-mammal congruency was found to be significant (model adjusted

R-square = 0.998, coefficient estimate = 0.0297, P = 0.0007)

63

Figure 2. Plots showing variation in partial Mantel R (congruency of species

composition) over the study period. Error bars represent the 95% confidence

intervals after 999 permutations. Diamond shaped points represent woodland

remnants in the pine matrix (treatment) while circle-shaped points represent

woodland remnants in the agricultural matrix.

64

Figure 3. Scatterplots showing shifts in mean site species richness and abundance for

birds, reptiles and mammals over the study years spanning 1999 and 2013. (Legend:

shaded diamond-shaped points represent woodland remnants in pine (treatment)

while shaded circle-shaped points represent remnants in agriculture; unshaded

diamond- and circle-shaped points represent the control sites in the respective pine

and agricultural matrix)

65

Figure 4. NMDS ordination plots for (a) mammal, (b) reptile and (c) bird

communities in woodland remnants surrounded in the pine (treatment) matrix, and

(d) mammals and (e) birds in woodland remnants in the agricultural matrix. The

ordination plot for reptiles in agricultural woodland remnants is not shown due to

its wide scatter of point clusters. Number of dimensions and stress values for all

NMDS ordinations are shown on each plot. (Legend: black diamond – 1999, purple

triangle – 2001, blue circle – 2011, red square – 2013)

66

Table 1. Glossary of selected important terms in the concept of cross-taxonomic

surrogacy and their definitions.

Term Definition

Cross-taxonomic surrogacy

The hypothesis that changes in the diversity or composition in a

defined taxon (the surrogate) reflects a similar and commensurate

change in another taxon (the target).

Congruence

The degree of concordance between measures of two defined

taxonomic units (e.g. Fattorini et al. 2012; Westgate et al. 2014).

Often measured by the level of correlation between diversity

metrics of the defined taxonomic groups (e.g. Su et al. 2004), and is

an important requisite in identifying cross-taxonomic surrogates

(Gioria et al. 2011)

Indicator species A species that can be used as a surrogate or proxy measure for the

distribution and occurrence of other species, species groups

(Ricketts et al 1999) and environmental conditions.

Species-based surrogate A surrogate approach based on data of individual species, defined

groups of species or measures of species diversity.

Species richness The total number of species in a defined biotic community; also a

commonly used metric in measures of biodiversity.

Species composition A metric of a biodiversity that considers the identity and relative

abundance of species in a defined biotic community.

Species temporal turnover Change in species composition in a biotic community over time.

67

Table 2. Results of partial Mantel correlations of species composition for three taxa

over the study period. Results for pine control (PIN) sites are not presented as there

was only adequate species data for one site.

Taxa correlated Mantel R n Mantel R n Mantel R n Mantel R n

Agriculture control

(AGR)

1999 2001 2011 2013

Bird vs Reptile - - -0.403 4 0.424* 7 0.0531 9

Bird vs Mammal - - - - - - - -

Reptile vs Mammal - - - - - - - -

Woodland remnants in agricultural matrix

Bird vs Reptile -0.134 17 0.0576 20 -0.0271 45 -0.102 50

Bird vs Mammal -0.154 7 0.0350 11 0.306** 25 0.168* 50

Reptile vs Mammal 0.173 25 -0.0314 25 -0.0779 32 0.0548 33

Woodland remnants in pine matrix

Bird vs Reptile 0.191* 17 0.114 21 -0.0780 41 0.0739 42

Bird vs Mammal 0.0880 13 0.279** 9 -0.0996 28 0.0791 29

Reptile vs Mammal - - -0.0521 17 -0.109 32 0.145 32

Significance P < 0.001 **, P <≤ 0.05 *, 0.05 < P ≤0.1 •

68

Table 3. Multiple regression in matrix (MRM) models and summary statistics for

predictor variables. Predictor variables included bird, mammal and reptile

composition, and geographic space.

Predictor

variable

2013 2011

Coefficient P Coefficient P

Woodland remnants in pine matrix

Mammal ~ Bird + Reptile + Space R2 = 0.0309 R2 = 0.0254

Bird 0.312 0.157 -0.385 0.0911

Reptile 0.150 0.0650 -0.144 0.0731

Space -0.0500 0.915 0.475 0.184

Reptile ~ Bird + Mammal + Space R2 = 0.0309 R2 = 0.0285

Bird 0.312 0.167 -0.321 0.285

Mammal 0.150 0.0771 -0.100 0.098

Space -0.0500 0.924 0.555 0.165

Woodland remnants in agricultural matrix

Mammal ~ Bird + Reptile + Space R2 = 0.0565* R2= 0.182*

Bird 0.425 0.00400* 0.859 0.00100*

Reptile 0.0979 0.258 -0.0870 0.376

Space 0.195 0.429 -0.0420 0.881

Reptile ~ Bird + Mammal + Space R2 =0.0287 R2= 0.0176

Bird -0.225 0.131 0.00942 0.955

Mammal 0.0582 0.280 -0.0620 0.343

Space 0.559 0.0130* 0.483 0.0450*

Significance P < 0.001 **, P <≤ 0.05 *, 0.05 < P ≤0.1 ·

69

Table 4. Significant habitat structure correlates of bird, reptile and mammals in two

different landscape contexts, identified with non-metric multidimensional scaling.

Variable NMDS1 NMDS2 R2 Variable NMDS1 NMDS2 R2

Woodland remnants in pine matrix

Bird species composition

Woodland remnants in agricultural matrix

Bird species composition

% crown affected 0.581 -0.814 0.163* Blackberry 0.463 -0.886 0.104·

Basal count 0.907 -0.420 0.272** Dead trees 0.777 0.629 0.265**

Blackberry -0.997 -0.0793 0.149· Exposed rock 0.900 -0.436 0.162**

Logs 10-20cm 0.961 -0.276 0.114· Ground cover 0.377 0.926 0.137*

Number of strata -0.762 -0.647 0.129· Number of strata 0.801 0.599 0.190**

Reptile species composition Number of trees 0.543 0.840 0.214**

Blackberry 0.989 -0.149 0.292** Shrub cover 0.941 0.339 0.236**

Dominant cover -0.487 0.874 0.226* Stand height -0.566 0.825 0.119*

Number of strata 0.999 0.0545 0.145· Subdominant cover 0.522 0.853 0.119*

Stem count 11-

20cm

0.846 -0.534 0.154* Reptile species composition

Mammal species composition Foliage depth 0.486 -0.874 0.116·

Logs >50cm 0.336 0.942 0.181· Mammal species composition

Blackberry 0.957 -0.290 1.000*

Exposed rock 0.00146 1.000 0.241*

Foliage depth -0.00110 1.000 0.190*

Stand height -0.00122 1.000 0.168*

Significance P < 0.001 **, P <≤ 0.05 *, 0.05 < P ≤0.1 ·

70

Supplementary material

Table A1. Definition of each landscape class and total number of study sites identified at the

commencement of the study.

Landscape context Definition Number of

plantation/property

Number of

sites

Eucalypt woodland

remnants in

agricultural matrix

Strips and patches of grassy

Eucalypt woodland

surrounded in part or in

whole by grassy grazing

pastures

6 farm properties 56

Eucalypt woodland

remnants in pine

plantation matrix

Strips and patches of grassy

Eucalypt woodland

surrounded in part or in

whole by plantations of

exotic pines

2 plantations (Cotway

& Nanangroe

plantations)

52

Agriculture control Randomly selected sites

within grassy grazing

pastures

- 10

Pine control Randomly selected sites

embedded within plantations

of exotic pine

- 10

71

Table A2. List and description of habitat structure variables.

Habitat structural

variable

Description (plot size: 20m by 20m)

Dominant Cover % cover of plants which height are taller than 10m

Sub-dominant Cover % cover of plants which height is between 2-10m

Shrub Cover % vegetation cover of <2m in height

Grass Cover % cover of native and exotic annual and perennial grasses

Grass Height Estimate of average height of grasses

Ground Cover % cover of native broad leaves, herbs and forbs

Weed Cover % cover of weeds (exotic ground cover such as broad leaves, herbs and forbs)

Litter Layer % cover of leaf litter

% Blackberry % cover of blackberry

% Exposed Rock % cover of exposed rock

Logs 20-30cm Number of logs > 10cm in diameter, and length >20cm and <30cm

Logs 30-40cm Number of logs >10cm in diameter, and length >30cm and <40cm

Logs 40-50cm Number of logs >10cm in diameter, and length >40cm and <50cm

Logs >50cm Number of logs >50cm in length, and >10cm in diameter

Number Of Strata Number of strata. Maximum of four strata (ground cover, under-storey, mid-storey

and over-storey)

Number Of Trees Number of trees

Regrowth % cover of regrowth

Stem diameter 1-5cm Number of stems <5cm in diameter

Stem diameter 6-10cm Number of stems >5cm, and <10cm in diameter

Stem diameter 11-20cm Number of stems >10cm, and <20cm in diameter

Stem diameter 21-30cm Number of stems >20cm, and <30cm in diameter

Stem diameter 31-40cm Number of stems >30cm and <40cm in diameter

Stem diameter 41-50cm Number of stems that >40cm, and <50cm in diameter

Stem diameter >51cm Number of stems >50cm in diameter

Stumps Number of stumps

Basal Count Measure of the number and the size of trees in a stand. obtained by holding a 1cm

wide gauge at 50cm away from eye

Canopy Height Height of the biggest trees

Canopy Depth Distance (m) from top of foliage to the bottom of foliage of the biggest tree

Hollow Count Number of tree hollows

Mistletoe Count Number of clumps of mistletoe

Dead Trees Number of dead trees

Dead Shrubs Number of dead shrubs

% Crown Affected Estimated extent of dieback in terms of percentageof unhealthy canopy cover

(affected by dry weather or insect attack)

72

Table A3. Spearman’s correlations of species richness for three taxa over the study period.

Due to low species richness, mammal and reptile data were pooled for one set of analysis

over all four study years.

Taxa correlated Spearman’s

ρ

n Spearman’s

ρ

n Spearman’s

ρ

n Spearman’s

ρ

n

Agriculture control (AGR) 1999 2001 2011 2013

Bird vs Reptile 0.226 10 - - 0.0481 10 0.699* 10

Bird vs Mammal - - - 0.441 10 0.322 10

Bird vs Mammal +Reptile 0.226 10 - - 0.168 10 0.657* 10

Reptile vs Mammal - - - 0.545 10 0.557 10

Pine control (PIN)

Bird vs Reptile -0.243 9 0.216 8 0.386 10 0 10

Bird vs Mammal - - - - - 0.155 10

Bird vs Mammal +Reptile -0.243 9 0.216 8 0.386 10 0.102 10

Reptile vs Mammal - - - - - -0.167 10

Woodland remnants in agricultural matrix

Bird vs Reptile -0.0345 54 -0.00133 54 0.152 55 0.00910 53

Bird vs Mammal -0.0644 54 0.00179 49 0.306* 55 0.347* 53

Bird vs Mammal +Reptile -0.0640 54 -0.0525 49 0.308** 55 0.170 53

Reptile vs Mammal 0.182 54 -0.0204 50 -0.00807 55 0.184 53

Woodland remnants in pine matrix

(TREAT)

Bird vs Reptile -0.0699 37 -0.339* 39 -0.169 45 0.142 44

Bird vs Mammal 0.261 37 0.0723 44 0.220 45 0.277· 44

Bird vs Mammal +Reptile 0.224 37 0.0472 38 -0.0762 45 0.300* 44

Reptile vs Mammal -0.332* 38 -0.205 37 -0.153 45 -0.0719 44

Significance P < 0.001 **, P <≤ 0.05 *, 0.05 < P ≤0.1 ·

73

Table A4. Candidate models based on multiple regression on distance matrices (MRM)

using species data in 1999 and 2001, and summary statistics for predictor variables. Predictor

variables included bird, mammal and reptile composition, and geographic space.

Year

Variable

1999 2001

Coefficient P Coefficient P

Woodland remnants in pine matrix (TREAT)

Mammal ~ Bird + Reptile + Space R2 = 0.193

Bird 1.268 0.0210 - -

Reptile -0.143 0.583 - -

Space 0.109 0.906 - -

Reptile ~ Bird + Mammal + Space R2 = 0.0476

Bird 0.559 0.291 - -

Mammal -0.111 0.529 - -

Space -0.675 0.474 - -

Woodland remnants in agricultural matrix

Mammal ~ Bird + Reptile + Space R2 = 0.147

R2 = 0.147

Bird 1.362 0.0960 -0.236 0.154

Reptile -0.0310 0.876 0.0207 0.595

Space -1.476 0.157 -0.0691 0.887

Reptile ~ Bird + Mammal + Space R2 = 0.0344 R2 = 0.121

Bird -0.0404 0.940 -0.689 0.582

Mammal -0.0202 0.897 0.411 0.555

Space 0.669 0.338 -1.060 0.3253

74

Table A5. Habitat structural variables evaluated in global MRM models for all taxa in

woodland remnant in pine, and agriculture sites. (All values to three significant figures)

MRM Model Global model (bird) Global model (reptile) Global model (mammal)

Woodland remnants in pine matrix R2 0.164 0.148 0.0945 F 7.870 5.638 2.053

Habitat variables

Intercept 0.598 0.748 0.557 % crown affected 0.000203 -0.000874 0.00180• Basal count 0.00333* -0.00113 -0.0000390 Blackberry 0.000123 0.00216 0.000385 Dead trees 0.00239 -0.00139 -0.00188 Dominant cover -0.0000383 0.00313 0.000157 Exposed rock -0.00143 0.00114 0.000402 Grass cover -0.0000491 -0.000786 -0.0000769 Grass height -0.000964 -0.000802 0.000556 Hollow count 0.00522 -0.0322* 0.00577 Litter layer 0.00187 0.0224 0.00900 Logs (>50cm) 0.0109 0.0438 -0.00287 Logs (10-20cm) 0.00115 -0.00558 -0.00704• Logs (30-40cm) -0.00161 -0.0195 -0.0126 Logs (40-50cm) 0.00693 0.0153 -0.0311 Mistletoe count 0.000201 -0.0107 0.0132* Number of strata 0.00183 0.0838* -0.0208 Number of trees -0.00226 0.000106 -0.00148 Stem diameter >51cm -0.000421 -0.0298 0.0148 Stem diameter 41-50cm -0.00737 -0.00109 -0.00734 Subdominant cover -0.00216 0.000546 0.00398• Weed cover -0.000445 0.000113 -0.000260

Woodland remnants in agricultural matrix R2 0.183* 0.117* 0.185• F 15.65* 6.080* 5.133

Intercept 0.621 0.795 0.435 % crown affected 0.000288 0.00173• 0.00268 Basal count -0.000796 -0.00158 0.00131 Blackberry 0.0151 0.000356 0.0572• Dead trees 0.00160 -0.0112 0.0554* Dominant cover -0.000686 0.00301* -0.00217 Exposed rock 0.000248* 0.00114 0.00265• Foliage depth 0.00379 0.000242 -0.000368 Grass cover -0.000651 0.0000 -0.000378 Ground cover 0.000262 0.00142* 0.000418 Hollow count 0.00185 -0.0116 -0.00968 Litter layer -0.0223 0.0610* 0.0496 Logs (>50cm) -0.0220 0.0122 0.0778 Logs (10-20cm) -0.00544• -0.00430 -0.00422 Logs (30-40cm) 0.00932 -0.0403• 0.0169 Logs (40-50cm) 0.0115 0.00866 0.0904• Mistletoe count 0.000633 -0.0146• 0.00356 Number of strata 0.0300 -0.0251 0.0413 Number of trees 0.00571 0.00258 0.00450 Stumps 0.0170 -0.0236 -0.00653 Subdominant cover 0.000657 0.00175 -0.00969 Weed cover 0.000150 0.000254 0.00195

Significance P < 0.001 **, P <≤ 0.05 *, 0.05 < P ≤0.1 •

75

Table A6. List of birds recorded in the Nanangroe landscape from 1999-2013

Species Scientific Name 1999 2001 2011 2013

Painted Buttonquail Turnix varia + Brown Quail Coturnix ypsilophora + Stubble Quail Coturnix pectoralis + + + Little Black Cormorant Phalacrocorax sulcirostris + Australasian Grebe Tachybaptus novaehollandiae + + + Australasian Wood Duck Chenonetta jubata + + + + Pacific Black Duck Anas superciliosa + + + Chestnut Teal Anas castanea + Grey Teal Anas gracilis + White-faced Heron Egretta novaehollandiae + + + + Nankeen Night Heron Nycticorax caledonicus + Straw-necked Ibis Threskiornis spinicollis + Black-fronted Dotterel Elseyornis melanops + Black-shouldered Kite Elanus axillaris + + Brown Goshawk Accipiter fasciatus + + + Collared Sparrowhawk Accipiter cirrocephalus + Wedge-tailed Eagle Aquila audax + + + + Brown Falcon Falco berigora + + + Peregrine Falcon Falco peregrinus + Nankeen Kestrel Falco cenchroides + + + + Peaceful Dove Geopelia placida + + + + Crested Pigeon Ocyphaps lophotes + + + + Common Bronzewing Phaps chalcoptera + + + + Gang-gang Cockatoo Callocephalon fimbriatum + + + + Sulphur-crested Cockatoo Cacatua galerita + + + + Yellow-tailed Black-Cockatoo Calyptorhynchus funereus + + Galah Eolophus roseicapilla + + + + Little Corella Cacatua sanguinea + Red-rumped Parrot Psephotus haematonotus + + + + Australian King-Parrot Alisterus scapularis + + + + Little Lorikeet Glossopsitta pusilla + + Crimson Rosella Platycercus elegans + + + + Eastern Rosella Platycercus eximius + + + + Shining Bronze-Cuckoo Chrysococcyx lucidus + + Horsfield’s Bronze Cuckoo Chrysococcyx basalis + + + Brush Cuckoo Cacomantis variolosus + Fan-tailed Cuckoo Cacomantis flabelliformis + + + + Pallid Cuckoo Cacomantis pallidus + Laughing Kookaburra Dacelo novaeguineae + + + + Sacred Kingfisher Todiramphus sanctus + + + + Rainbow Bee-eater Merops ornatus + + + + Dollarbird Eurystomus orientalis + Superb Lyrebird Menura novaehollandiae + Varied Sittella Daphoenositta chrysoptera + + + + White-throated Treecreeper Climacteris affinis + + + + Brown Treecreeper Climacteris picumnus + + + + Superb Fairy-wren Malurus cyaneus + + + + Spotted Pardalote Pardalotus punctatus + + + + Striated Pardalote Pardalotus striatus + + + + Mistletoebird Dicaeum hirundinaceum + + + + White-browed Scrubwren Sericornis frontalis + + + + Speckled Warbler Pyrrholaemus sagittatus + Western Gerygone Gerygone fusca + + + + White-throated Gerygone Gerygone olivacea + + + + Yellow-rumped Thornbill Acanthiza chrysorrhoa + + + + Yellow Thornbill Acanthiza nana + +

76

Brown Thornbill Acanthiza pusilla + + + + Striated Thornbill Acanthiza lineata + + + + Buff-rumped Thornbill Acanthiza reguloides + + + + Weebill Smicrornis brevirostris + + + + Red Wattlebird Anthochaera carunculata + + + + White-eared Honeyeater Nesoptilotis leucotis + + + + White-plumed Honeyeater Ptilotula penicillata + + + + Yellow-faced Honeyeater Caligavis chrysops + + + + White-naped Honeyeater Melithreptus lunatus + + + + Black-chinned Honeyeater Melithreptus gularis + +

Brown-headed Honeyeater Melithreptus brevirostris + + + + Blue-faced Honeyeater Entomyzon cyanotis + Crescent Honeyeater Phylidonyris pyrrhopterus + + + New Holland Honeyeater Phylidonyris novaehollandiae + + + + Fuscous Honeyeater Ptilotula fusca + + + Noisy Friarbird Philemon corniculatus + + + + Little Friarbird Philemon citreogularis + + + + Noisy Miner Manorina melanocephala + + + + Eastern Spinebill Acanthorhynchus tenuirostris + + + + Spotted Quail-thrush Cinclosoma punctatum + Eastern Yellow Robin Eopsaltria australis + + + + Flame Robin Petroica phoenicea + + + Red-capped Robin Petroica goodenovii + + Rose Robin Petroica rosea + + Scarlet Robin Petroica boodang + + + + Jacky Winter Microeca fascinans + + + + Leaden Flycatcher Myiagra rubecula + + + + Restless Flycatcher Myiagra inquieta + + + + Satin Flycatcher Myiagra cyanoleuca + Crested Shriketit Falcunculus frontatus + + + + Grey Shrike-thrush Colluricincla harmonica + + + + Rufous Whistler Pachycephala rufiventris + + + + Australian Golden Whistler Pachycephala pectoralis + + + Grey Fantail Rhipidura albiscapa + + + + Willie Wagtail Rhipidura leucophrys + + + + Magpie-lark Grallina cyanoleuca + + + + Olive-backed Oriole Oriolus sagittatus + + + + Satin Bowerbird Ptilonorhynchus violaceus + + + + Black-faced Cuckooshrike Coracina novaehollandiae + + + + White-winged Triller Lalage tricolor + + + + Dusky Woodswallow Artamus cyanopterus + + + + White-browed Woodswallow Artamus superciliosus + + Grey Butcherbird Cracticus torquatus + + + + Pied Butcherbird Cracticus nigrogularis + + Australian Magpie Gymnorhina tibicen + + + + Pied Currawong Strepera graculina + + + + Australian Raven Corvus coronoides + + + + Little Raven Corvus mellori + + + + White-winged Chough Corcorax melanorhamphos + + + + Welcome Swallow Hirundo neoxena + + + + Tree Martin Petrochelidon nigricans + + + + Fairy Martin Petrochelidon ariel + + + + Australasian Pipit Anthus australis + + + + Rufous Songlark Megalurus mathewsi + + + + Brown Songlark Megalurus cruralis + + + Australian Reed Warbler Acrocephalus australis + Golden-headed Cisticola Cisticola exilis + Red-browed Finch Neochmia temporalis + + + + Diamond Firetail Stagonopleura guttata + + + + Mistletoebird Dicaeum hirundinaceum + + + + Silvereye Zosterops lateralis + + + +

77

Table A7. List of mammals detected in Nanangroe landscape during spotlighting surveys

from 1999-2013

Species Scientific Name 1999 2001 2011 2013

Common Wombat Vombatus ursinus + + +

Common Brushtail Possum Trichosurus vulpecula + + + +

Sugar Glider Petaurus breviceps + + + +

Squirrel Glider Petaurus norfolcensis +

Greater Glider Petauroides volans + +

Common Ringtail Possum Pseudocheirus peregrinus + + + +

Eastern Grey Kangaroo Macropus giganteus + + +

Red-necked Wallaby Macropus rufogriseus + + +

Common Wallaroo Macropus robustus + +

Black Wallaby Wallabia bicolor + + +

78

Table A8. List of reptiles recorded in the Nanangroe landscape from 1999-2013

Species Scientific Name 1999 2001 2011 2013

Marbled Gecko Christinus marmoratus + + + +

Stone Gecko Diplodactylus vittatus + +

Burton's Legless Lizard Lialis burtonis +

Olive Legless Lizard Delma inornata + + +

Eastern Three-lined Skink Acritoscincus duperreyi + +

Red-throated Skink Acritoscincus platynotum + +

Four-fingered Skink Carlia tetradactyla + + + +

Ragged Red-eyed Skink Cryptoblepharus pannosus + +

Straight-browed Skink Ctenotus spaldingi + + +

Copper-tailed Skink Ctenotus taeniolatus + + +

Cunningham's skink Egernia cunninghami + +

Tree Skink Egernia striolata + + +

Southern water Skink Eulamprus heatwolei + + +

Eastern Three-toed Earless

Skink

Hemiergis talbingoensis + + + +

Delicate Skink Lampropholis delicata + + + +

Common Garden Skink Lampropholis guichenoti + + + +

White's Skink Liopholis whitii +

Boulenger's skink Morethia boulengeri + + + +

Eastern Blue-tongue Tiliqua scincoides + + +

Eastern Bearded Dragon Pogona barbata + +

Dwyer's Snake Parasuta dwyeri +

Eastern Brown Snake Pseudonaja textilis +

79

Supplementary material (figures)

Figure A1. NMDS ordination plots showing shifts in habitat structural variables in woodland

remnants in pine plantation (a) and agricultural matrix (b) over the study period (1999–

2010). (Legend: black circles – 1999, red triangles – 2005, blue diamonds – 2010). (Note:

habitat structural variables are unavailable in 2005 for study sites in the agricultural matrix

and thus not analysed)

80

Figure A2. Cluster dendrogram plot showing species compositional differences over the four

study years based on within-group and between-group Bray-Curtis dissimilarities for all

three animal taxa, in woodland remnants in pine (treatment) and agricultural matrix

(remnants).

81

II. Yong, D. L., Barton, P. S., Okada, S., Crane, M., Cunningham, S. A., &

Lindenmayer, D. B. (In review). Conserving bee and beetle assemblages in

farming landscapes: the role of landscape context and structure on cross-

taxonomic congruence. Biodiversity and Conservation.

82

Conserving bee and beetle assemblages in farming landscapes: the

role of landscape context and structure on cross-taxonomic

congruence

Abstract

Identifying shared responses of different insect taxa to landscape modification is

necessary for understanding how conservation of key insect groups might benefit

other groups. Yet, little is understood about the processes driving cross-taxonomic

patterns or how this knowledge can be used to better manage biodiversity in

farming landscapes. We investigated how wild bee and ground-active beetle

assemblages respond to different landscape contexts and habitat structural

components, and how this influenced cross-taxonomic congruence of the two

groups. Wild bee and ground-active beetle assemblages were sampled in woodland

patches in two landscape contexts: surrounded by pine plantation (the pine matrix)

or by cleared grazing land (the agricultural matrix), and in pine monoculture. Total

bee species richness, and richness of functionally-defined groups were not different

across landscape contexts. However, total beetle species richness, and richness of

functionally-defined groups differed significantly across landscape contexts. We

found that landscape context had a stronger effect on species composition than

species richness for both groups. Although some landscape and habitat variables

were useful in predicting the diversity of either insect group, few were shared. Our

findings showed that wild bee and beetles are poor surrogates for each other in

farming landscapes. Our study highlights the need to consider: (1) taxon-specific

responses to landscape context, (2) different metrics of cross-taxonomic surrogacy

and (3) differences in ecological attributes across insect taxa when managing

landscapes for their conservation. It should not be assumed that agricultural

landscapes managed to conserve charismatic insects (e.g. bees) will necessarily

benefit other major insect groups.

83

Introduction

Transformation of the world’s landscapes for human use is a major driver of

biodiversity decline (Foley et al. 2005; Souza et al. 2015). Currently, land used for

grazing and animal fodder constitutes half of the world’s agricultural land area, and

well over 10% of the world’s terrestrial surface (FAO 2011). Further expansion and

intensification of agriculture is expected to impact the world’s biodiversity at various

levels and scales (Flynn et al. 2009; Le Féon et al. 2010; Tscharntke et al. 2012).

Therefore, there is an urgent need to identify better ways to conserve and manage

biodiversity in anthropogenic landscapes (e.g. Henle et al. 2008; Batáry et al. 2010).

While information on biodiversity is needed to guide conservation, not every

component of biodiversity can be cost-effectively measured (Lindenmayer and

Likens 2011). Furthermore, data on many animal taxa are scarce and remain

logistically difficult to obtain due to the high sampling effort needed (Favreau et al.

2006; Caro 2012). Conservation practitioners therefore rely on surrogate measures to

quantify these biodiversity components (Prendergast and Eversham 1997; Caro 2010;

Lindenmayer et al. 2015), especially cross-taxonomic surrogates (Westgate et al.

2014). Cross-taxonomic surrogate approaches are underpinned by the assumption

that patterns shown by one taxon (i.e. ‘the surrogate’) consistently predict changes

in another taxon of interest (i.e. ‘the target’) or broader components of biodiversity

(Rodrigues and Brooks 2007; Barton et al. 2015). Determining whether different taxa

show similar responses to habitat disturbance and modification (Schulze et al. 2004)

or associations in diversity (Kati et al. 2004; Westgate et al. 2014) is fundamental to

identifying cross-taxonomic surrogates.

Insects are a challenging component of biodiversity to document and conserve given

their immense diversity, the poor state of knowledge for many species and the

limited taxonomic expertise available (e.g., Stork et al. 2015; Hochkirch 2016).

84

However, the broader conservation of insects has received far less attention than

vertebrates (Dunn 2005; Samways 2005; Guiney & Oberhauser 2008). This is despite

the important ecological roles played by insects as pollinators, herbivores, ecosystem

engineers and as prey for other species (e.g. Losey and Vaughan 2006; Nichols et al.

2008). Anthropogenic modification of the landscape affects insect communities at

many levels and scales (e.g. Samways 2005; Kennedy et al. 2013). For example, land-

use change can modify habitats, alter landscape configuration, or change plant

diversity and environmental conditions important to insects (Fischer and

Lindenmayer 2007; Tscharntke et al. 2008; Woltz et al. 2012; Rösch et al. 2013). These

anthropogenic impacts have led to the decline of key insect groups and are expected

to have a significant impact on ecosystem functioning and human well-being (Dirzo

et al. 2014).

Two insect groups of great interest worldwide are bees (Order Hymenoptera,

superfamily Apoidea,) and beetles (Order Coleoptera). Bees are recognised as the

most important group of pollinating insects and are crucial for maintaining

functioning ecosystems (Klein et al. 2007; Hopwood 2008). Given their importance

to agricultural systems and the threat of a ‘global pollinator crisis’ (Potts et al. 2010),

the responses of pollinating insects to habitat modification have been reasonably

well studied (Le Féon et al. 2010; Kleijn et al. 2015). Additionally, bees are targeted by

many initiatives aimed at conserving pollinator diversity (NAPPC 2015; The World

Bee Project 2016). By contrast, other important insect groups such as beetles have

received commensurably less attention from conservationists even though they

constitute a third of all described insect species and perform equally important

ecological roles (New 2007; Barton et al. 2009; Hangay and Zborowski 2010; Stork et

al. 2015). Since relatively few studies have examined the response of multiple insect

groups to landscape modification (e.g. Tscharntke et al. 2002; Gardner et al. 2009), it

remains unclear if agricultural landscapes managed to conserve some groups of

insects (e.g. bees) will benefit other important insects such as beetles.

85

In this study, we compared responses of wild bees and ground-active beetles to

habitat modification in a heavily-transformed landscape in south-eastern Australia

widely representative of the plantation and pasture landscapes across the region

(e.g. Lindenmayer and Hobbs 2004). In surveying these two taxa, we investigated if

either group can be used as a surrogate for the other by assessing: (1) their responses

to different landscape contexts (we subsequently refer to the plantation and grazing

land matrix as “landscape contexts”) and, (2) cross-taxonomic congruency between

the two groups across the whole landscape. We also evaluated landscape and habitat

structure as surrogates for insect diversity since a comparative approach can lead to

better surrogate selection for management (Lindenmayer et al. 2014; Barton et al.

2015). We structured our questions into a conceptual framework that represents the

links between these two insect groups, habitat structure and landscape context

(Figure 1). This framework formed the basis to our approach in investigating the

different types of surrogacy (e.g. habitat and cross-taxonomic surrogacy) with

respect to landscape modification.

First, we were interested in determining similarities in responses of bees and beetles

to landscape context. We therefore asked: (1) How does species richness and species

composition of each insect group respond to different landscape contexts? We then

asked: (2) How do groups with shared functional attributes respond to different

landscape contexts? We predicted that the response in species richness would be

similar, but responses at the species composition level, and between defined

functional groups could differ strongly across landscape contexts. This is because

studies of insect assemblages have revealed stronger responses to habitat structural

differences at the functional-group level (Ribera et al. 2001; Purtauf et al. 2005).

Next, we were interested in identifying the different components of the landscape

and vegetation structure that can be used as surrogates of species richness and

composition of both insect groups. We therefore asked: (3) What are the landscape

and habitat structure variables that best predict bee and beetle species richness and

composition? Identification of important habitat components means that easily

86

measured habitat structure variables can be considered independently as surrogates

for insect assemblages. Finally, to test if either insect group could be used to predict

the diversity of the other group, we asked: (4) Are patterns of bee and beetle species

diversity congruent across the study landscape?

Methods

Study sites

The Nanangroe landscape consists of nearly 30,000 ha of agricultural (i.e. grazing)

land and exotic Monterey pine Pinus radiata plantations (See map: Figure 2). Most of

the original vegetation (box-gum grassy woodlands) has been cleared in the past two

centuries for agriculture and grazing land (Lindenmayer et al. 2008). The present

landscape is ‘highly variegated’ (McIntyre and Barrett 1992) and consists of distinct

patches of remnant woodland of varying tree densities and scattered eucalypt trees.

These woodlands are enveloped by either a matrix of pastures actively grazed by

livestock, and monoculture plantations.

52 woodland remnants in four size classes were identified based on their constituent

vegetation in 1999. In 1998, the landscape matrix surrounding these remnants was

transformed with the establishment of dense plantations of the Monterey Pine Pinus

radiata (Lindenmayer et al. 2008), which now exceed 251,000 ha in size in New

South Wales (Forestry Corporation NSW 2016). A further 56 patches of woodland

remnants of matching vegetation classes and sizes were identified in surrounding

agricultural land (thereafter referred to as ‘woodland remnants in agricultural

matrix’). Permanent transects were marked and established at study sites prior to

the commencement of the study. For this study a subset of 20-23 remnant woodland

patches each in the pine plantation and agricultural matrices were randomly chosen

87

to represent the full range of patch size classes (Table 1). Additionally, five sites

within the pine plantation monoculture were selected as habitat contrasts.

Insect sampling

We sampled bees using coloured vane traps. This survey method has been

increasingly used in open, temperate landscapes in Australia and North America

(e.g. Hogendoorn 2011; Lentini et al. 2012; Joshi et al. 2015). We sampled all 48 sites at

the midpoint of each line transect with two traps at each site, located in trees

approximately 20m apart from which they were suspended at about 1.5-2.0m above

ground. Each trap consisted of blue coloured vanes attached to a bright yellow

plastic jar. Bee sampling was conducted from November to December 2014 during

the austral spring when bee activity is at its peak. At the end of the sampling period,

the traps (81 traps from 43 sites) were retrieved and all insects caught were preserved

in 70% ethanol before species-level sorting. Bees that were difficult to identify were

assembled into a reference collection for subsequent species-level identification,

following the protocol in Droeges (2015). Bees were identified to the species-level

using the Pest and Diseases Image Library (PaDIL 2016) and identification keys (e.g.

Walker 1995; Michener 2000). Identified bee species were then validated by a

taxonomist (Michael Batley, Australian Museum). Some genera of bees (i.e.

Exoneura sp.) were classified only to the morphospecies level due to their unstable

taxonomy (M, Schwarz. pers comm. 2015).

To sample ground-dwelling beetles, we used non-baited pitfall traps placed in four

rows, with each row located about 1.0m apart. Each pitfall trap consisted of a plastic

container of 5.0cm diameter and 7.5cm depth. Traps were filled with 100ml of

ethylene glycol. To increase catch rates we mounted a 1.0m x 0.2m plastic drift fence

along each pair of traps. Beetle trapping was conducted from November to

December 2014. In total, 384 pitfall traps were set up across 48 study sites. 330 traps

88

were recovered from 44 sites at the end of the sampling period while all traps at four

sites were damaged by livestock. All beetle specimens were preserved in 70% ethanol

before being sorted to morphospecies level using identification keys (e.g. Matthews

1992; Hangay and Zborowski 2010). Voucher specimens for each morphospecies were

assembled into a reference collection for comparison. Highly similar morphospecies

from speciose families such as Staphylinidae were further checked by an expert

familiar with beetle assemblages in similar landscapes (M. John Evans) for accuracy.

Vegetation sampling

To determine the vegetation structure across our study landscape, a total of 34

vegetation and habitat structural variables were measured at each sampling site (see

Supplementary Information Table S3 for full list of variables). Woodland remnants

were classified by their constituent tree genera into Eucalyptus-dominated and

Casuarina-dominated remnants. In addition, we conducted observational surveys to

categorically estimate flowering activity within a 50m radius of the insect traps at

the ground, shrub and canopy level. To quantify native tree cover in each woodland

remnant, we used tree cover area as measured in a circle (with a 250m and 500m

radius) centred at each sampling site. We defined native tree cover to include

clusters of trees within habitat patches, as well as the single scattered trees in the

landscape. Native tree cover area was measured using digitised aerial photographs in

ArcGIS version 10.1 (see Mortelliti and Lindenmayer 2015).

Statistical analysis

We calculated site-level species richness for both insect groups (Question 1).

Because some of the pitfall traps were damaged by livestock, we included only sites

that retained the majority of the pitfall traps (at least four out of eight traps) for

89

analysis. Similarly, we used only bee data from traps that were not damaged by bad

weather. We computed species richness estimates using one of three non-parametric

estimators (Chao1) (see Walther and Moore 2005). We then plotted sample-based

rarefaction curves using 999 random permutations to assess sampling completeness,

and compare species richness for both groups in different landscape contexts

(Gotelli and Colwell 2001). We performed this analysis using the function ‘specpool’

in the ‘vegan’ package (Oksanen et al. 2016).

We calculated Moran’s I to assess the effects of spatial autocorrelation (Legendre

1993) on observed species richness for both insect groups. Moran’s I test assesses the

relationship of the dependant variable against a matrix of weights (i.e. geographic

distances) and was implemented on the ‘ape’ package (Paradis et al. 2004). We then

we fitted a series of generalised linear models to test if mean site-level species

richness can be explained by the different landscape contexts. Because the

dependant variable (i.e. observed species richness) involved count data, models were

fitted with a Poisson-error distribution and a logarithmic link function. If species

richness was found to be spatially correlated, we accounted for this in the

generalised linear models by fitting ‘site’ as a random effect while ‘landscape context’

was retained as a factor with three levels. This analysis was carried out using the

‘glmer’ function available in the package ‘lme4’ (Bates et al. 2014).

To compare site-level species composition across landscape contexts (Question 1),

we performed non-metric multi-dimensional scaling (NMDS) analysis to ordinate

site-level counts in species space for both insect groups. Raw abundances of all

species were first square root-transformed to reduce the influence of highly

abundant species. We then used the multiple response permutation procedure

(MRPP) as a non-parametric test to assess for significant differences in species

composition across the two landscape contexts and pine contrasts, and between

pairs of landscape contexts (McCune and Grace 2002). Each MRPP analysis yields

90

the effect size statistic A which measures within-class heterogeneity with the Bray-

Curtis dissimilarity, and which is assessed using 1,000 permutations.

To test if functional attributes influenced responses of both bee and beetle species

assemblages to landscape context (Question 2), we compared species richness and

composition of groups defined by their shared functional attributes. We partitioned

the bee dataset into two groups based on data on species-specific nesting

requirements available in Dollin et al. (2000) and Michener (2000), and further

validated by an expert (M. Batley). Past studies found that nesting strata

significantly influenced responses to various types of habitat change (e.g. Cane et al.

2006; Williams et al. 2010). Consequently, life-history traits that influenced how a

species responds to environmental conditions could prove useful for understanding

fine-scale community responses (Larsen et al. 2005; Greenleaf et al. 2007; Barton et

al. 2013). We grouped bee species that nest in cavities or plant parts (e.g. Exoneura

sp.) and defined them as ‘above-ground’ nesting bees (sensu Williams et al. 2010).

The remaining bees were ground-nesting species (e.g. Amegilla, Lasioglossum sp.).

Each individual beetle morphospecies from our reference collection were carefully

checked under a stereo microscope and considered as flightless if the elytra were

fused or if wings were absent. All other beetle species were recorded as flight-

capable.

Again, we fitted a series of generalised linear models with a Poisson-error

distribution to compare species richness of each functionally-defined species group

across the landscape contexts. Because of spatial autocorrelation in the bee dataset

detected earlier in the Moran’s I test, we fitted ‘site’ as a random effect in models for

both functionally-defined groups. To compare species composition for each group

across landscape contexts, we used MRPP tests with 1,000 permutations. Lastly, to

test if each group defined by a shared functional attribute were useful in predicting

species richness of the other three groups, we performed Spearman’s rank

91

correlation. Spearman’s rank correlation was used as the sample size (number of

study sites) was low and did not meet the parametric assumptions needed for

Pearson’s correlation tests.

We evaluated the influence of habitat and landscape variables on species richness

and species composition (Figure 1; Question 3). For species richness, we fitted a

series of generalised linear models that related species richness to six explanatory

landscape variables. Highly correlated variables (Pearson’s r > 0.5) were excluded

from the analysis after inspecting a correlogram matrix of all variables. Three

categorical variables, ‘forest type’, ‘topography’ and ‘water body’ were transformed

into factors. We applied a Poisson error distribution and logarithmic link function in

the models rather than log-transforming our data (O’Hara and Kotze 2010). Two sets

of candidate models were fitted and assessed using Bayesian Model Averaging

(BMA) (Wintle et al. 2003). We implemented BMA to account for model uncertainty

in the model selection process by taking the average of the best candidate models

based on their posterior model probability. The best five models in each candidate

set were ranked by the Bayesian Information Criterion (BIC) and their posterior

probability. Model selection was implemented using the package ‘BMA’ (Raftery et

al. 2015).

To compare the effects of different habitat structure on bee and beetle species

composition, we fitted vectors for selected habitat structures variables into our

NMDS ordination results. The full set of variables were first assessed using a

correlogram matrix and retained for analysis only if found to be not strongly

correlated (Pearson’s r < 0.5). The function ‘envfit’ available in the ‘vegan’ package

computes factor averages or vectors for each habitat structural variable fitted to the

ordination matrix (Oksanen et al. 2016). Excluding highly correlated variables

(Pearson’s r > 0.5), we compared 15 habitat variables. The significance of the fitted

vectors for each habitat structure variable was assessed using 999 permutations.

92

In our earlier questions, we focussed on the responses of the two insect assemblages

to different landscape contexts. Here, we aimed to assess how congruent the

diversity patterns of these two groups are across the study landscape (Figure 1;

Question 4). To assess for congruency of bee and beetle species richness, we

performed Spearman’s rank correlation using site-level species richness. We used

the Spearman’s correlation coefficient as a measure of correlation strength because

our dataset found to be non-normally distributed using the Shapiro-Wilk test. To

assess for congruency in species composition, we used partial Mantel tests

implemented on the package ‘ecodist’ (Goslee and Urban 2013). Partial mantel tests

were used as the strength of correlation between two matrices conditioned on a

third matrix of geographic distances allowed the effects of space to be accounted for

(Goslee and Urban 2007), thus partitioning out the variation due to space. The

statistical significance of each partial Mantel tests was assessed using 999

permutations. All statistical analyses were conducted using R v. 2.15.1 (R Project for

Statistical Computing, http://www.r-project.org).

Results

We collected a total of 3,717 beetles representing 274 morphospecies in 36 families,

and 1,714 bees representing 33 species in four families. Bee species were best

represented by the families Halictidae (18 species), followed by Apidae (8 species)

and Megachilidae (5 species) (Table S1). Among beetle families, the most species-

rich were the rove beetles (family Staphylinidae, 35 morphospecies), scarabs (family

Scarabeidae, 33 morphospecies) and ground beetles (family Carabidae, 30

morphospecies) (Table S2). Our sampling effort was consistent for both insect

groups, detecting 64.0-77.8% of predicted bee diversity, and 61.5-71.1% of predicted

beetle diversity (see Supplementary Figure S1). We found that bee species richness

was weakly, spatially correlated between woodland remnants across the landscape

93

(Moran’s I = 0.0998, P = 0.004) but not beetle species richness (Moran’s I = -0.00185,

P = 0.649).

How do bee and beetle assemblages respond to different landscape contexts?

(Question 1)

We found bee species richness to be highest in woodland remnants in the

agricultural matrix (Chao1 estimate: 43 species), followed by remnants in the pine

plantation matrix (Figure 3a). However, predicted beetle species richness was higher

in woodland remnants in the pine plantation matrix (Chao1 estimate: 281 species)

than woodland remnants in the agricultural matrix (Figure 3b). Predicted species

richness was lowest in the pine plantation sites for both bees (Chao1 estimate: 12

species) and beetles (Chao1 estimate: 75 species) (Figure S1). However, bee species

richness was not significantly different between woodland remnants in either the

pine plantation or agricultural matrix (mean difference = -0.0905, Z = -0.395, P =

0.693) after accounting for variation due to random effects (variance = 0.0298).

Similarly, there were no difference in bee species richness between woodland

remnants in either the pine (mean difference = -0.236, Z = -0.950, P = 0.342) or

agricultural matrix (mean difference = -0.326, Z = 1.111, P = 0.267), and the pine

plantation sites. On the other hand, beetle species richness was significantly

different between pairwise comparisons of all landscape contexts. Beetle species

richness was significantly different between woodland remnants in both landscape

contexts (mean difference = -0.215, Z = 3.370, P < 0.001). Pine plantation sites were

significantly poorer in beetle species richness than both kinds of woodland remnants

(mean difference = -0.274, Z = -2.337, P = 0.0194) (Table S5).

We found that species composition of both bee and beetle assemblages were

significantly different between comparisons of all landscape contexts (Figure 4,

Table 2). However, species composition of the beetle assemblages was generally

94

more dissimilar across the landscape contexts than the bee assemblages (Table S5).

Pairwise comparisons of beetle species composition betweel landscape contexts were

always stronger when compared with similar pairwise comparisons for bees. For

instance, beetle species composition differed more strongly between woodland

remnants in the agricultural matrix and pine plantation sites (A = 0.0794, P = 0.001),

than that for bees (A = 0.0430, P = 0.010).

How do bee and beetle groups with similar functional attributes respond to the

landscape contexts? (Question 2)

We found that neither landscape contexts had a significant effect on species richness

of ground-nesting (mean difference = -0.0264, Z = -0.139, P = 0.890) or above-ground

nesting bees (mean difference = -0.159, Z = -0.340, P = 0.734) after accounting for

variation due to random effects (Figure 5b, Table S4). However, no species of above-

ground nesting bee occurred in the pine contrast sites even though ground-nesting

species persisted (mean richness = 4.4 species). We found that species richness of

flightless beetles differed across landscape contexts (mean difference = -0.515, Z = -

2.714, P < 0.01) (Figure 5b), but not between remnants in either landscape context

and pine plantation. Similarly, woodland remnants in the pine matrix supported a

significantly higher richness of flight-capable beetles than remnants in the

agriculture matrix (mean difference = 0.318, Z = 4.637, P < 0.001). Species richness of

flight-capable beetles between the pine contrast sites and woodland remnants in

pine were significantly different (mean difference = 0.553, Z = 4.415, P < 0.001) but

not between woodland remnants in agriculture.

When species composition of bee groups with similar functional attributes were

compared (Table 2), we found that above-ground nesting (A = 0.0677, P < 0.01) and

ground-nesting bee assemblages (A = 0.0411, P = 0.001) between woodland remnants

in the pine and agricultural matrix were significantly different. However, ground-

95

nesting bee assemblages in the pine plantation sites were not significantly different

when compared with woodland remnants in either landscape contexts. Species

composition of flightless beetles was significantly different for all pairwise

comparisons except for that between woodland remnants in the pine matrix and the

pine plantation sites (A = 0.0079, P = 0.200). However, species composition of flight-

capable beetles was significantly different between all pairwise comparisons of sites.

What are the landscape variables that best predict bee and beetle species richness

and composition? (Question 3)

We selected the best five of a series of candidate models (bees: 57 models, beetles: 25

models for beetles) based on their Bayesian Information Criterion (BIC) values and

posterior probabilities. The candidate model that best explained bee species

richness incorporated only the intercept (Table 3, Posterior probability = 0.159).

Among the landscape variables, we found that native tree cover was the most

important covariate even though it was only weakly correlated with bee species

richness (Supplementary Figure S2), occurring in only 45.2% of the candidate

models. Landscape context and topography were the next most important landscape

variables, occurring in 25.5% and 26.7% of the models. Generally, woodland

remnants in the pine plantation matrix and on slopes were characterised by lower

bee species richness.

The candidate model that best explained beetle species richness (Posterior

probability = 0.293) contained elevation, distance to water, and topography as the

explanatory variables. We found that elevation was the most important predictor of

beetle species richness, occurring in 100% of the candidate models (Table 3;

Supplementary Figure S2). Distance to water (58.8% of models), the location of a site

on a slope (85.2% of models), and/or a ridge (14.2% of models) were the next most

important predictor variables. Distance to water was only weakly associated with

96

higher species richness, while slope and ridge topographies of sites were associated

with low species richness. Unlike bees, native tree cover was found to be

unimportant for beetles (16.6% of models), and was in fact negatively correlated to

beetle species richness.

We identified four habitat structure variables that were significantly correlated with

the NMDS ordination axes describing bee species composition (Table 4): canopy

depth (R2 = 0.189), blackberry cover (R2 = 0.160), tree crown (R2 = 0.253) and the

extent of exposed rocks (R2 = 0.279). For beetle species composition, we identified

five significantly correlated variables. Basal stem count was the mostly strongly

correlated variable (R2 = 0.473). The other significant variables included tree crown

structure, extent of exposed rocks, litter layer and weed cover. Only two of these

variables were shared with bees (i.e. crown structure and extent of exposed rocks).

Are bee and beetle species richness and composition congruent across the study

landscape? (Question 4)

We found that site-level species richness for bees was not significantly correlated

with the species richness of beetles (Spearman’s ρ = 0.290, P = 0.063) across the

study landscape, within each landscape context and among functionally-defined

sub-groups (Figure S3). However, bee and beetle species composition were weakly

correlated across the study landscape (partial Mantel R = 0.108, P = 0.024). When

correlations of species composition were considered for each landscape context, all

relationships were weak and insignificant.

Discussion

97

Our study revealed that wild bee and ground-active beetle assemblages responded to

human transformation of the landscape in different ways. Our findings also provided

evidence of how functionally-defined attributes can influence a taxonomic group’s

response to different landscape contexts (e.g. Ribera et al. 2001; Williams et al. 2010).

Given better understanding of the role of native bee species in crop pollination,

there is an increasing interest in the conservation of wild bee assemblages in

Australia (Batley and Hogendoorn 2009; Dollin et al. 2016) and agricultural

landscapes globally (Klein et al. 2007; Jauker et al. 2009). However, wild bee

assemblages on their own may have limited use as surrogates for beetle and other

insect groups. Nonetheless, there will remain a perennial need to consider insects in

conservation planning and identify better surrogates to capture their diversity given

their overwhelming ecological importance (New 1999; Samways 2005; Barton et al.

2009).

How do bee and beetle assemblages respond to different landscape contexts?

Our findings showed that bee and beetle species richness were affected differently

by landscape context even though both groups were consistently depauperate in the

pine plantation sites (Question 1). Our data suggested that the landscape matrix had

a more pronounced effect on beetle than wild bee assemblages, and is consistent

with similar studies on insect assemblages in heavily transformed landscapes (e.g.

Öckinger et al. 2012). This may be explained by an increase in habitat complexity

arising from changes to the landscape matrix. Such changes in habitat complexity

may result in bees ranging over larger distances while foraging (Steffan-Dewenter

and Kuhn 2003). Second, the microclimate of the woodland remnants embedded in

the pine plantation may be different from remnants in the agricultural matrix (e.g.

Driscoll et al. 2013). The pine plantation matrix can alter microclimate in adjacent

woodland remnants through the reduction of wind and light penetration (e.g. Fahy

and Gormally 1998; Jukes et al. 2001), resulting in cascading effects on soil

98

conditions. These changes may create favourable microhabitats for ground-dwelling

beetles and drive the higher compositional heterogeneity observed in beetle

assemblages. While the landscape matrix can affect the dispersal of bees (Steffan-

Dewenter et al. 2002; Jauker et al. 2009), its transformation may concentrate bees

into ‘islands’ of woodland remnants embedded in a resource (i.e. nectar, pollen)-

scarce plantation matrix. Collectively, our findings suggest that differences in the

landscape matrix in human-modified landscapes should not be assumed to exert

similar effects across different insect assemblages (e.g. Jauker et al. 2009; Driscoll et

al. 2013).

How do bee and beetle groups with similar functional attributes respond to the

landscape contexts?

We found that bee species assemblages defined by shared functional attributes

exhibited different responses to landscape context (Question 2). Many ground-

nesting species (e.g. Lasioglossum sp.) were abundant and widespread across the

landscape, including the pine plantation monoculture. Moreover, ground-nesting

bee species richness and composition did not differ between woodland remnants in

the pine matrix and pine plantation sites. In contrast, species richness of above-

ground nesting bees was reduced in woodland remnants in the pine matrix and no

species occurred in pine plantation sites. Such patterns may be attributed to the

changes in fine-scale vegetation structure arising from the transformation of the

matrix, which reduced nesting resources. For instance, the dense pine stands in the

plantations may limit growth of hollow-bearing shrubs at the interface of woodland

remnants and the pine matrix, which are important to Exoneura sp. and other

above-ground nesting bees (Dollin et al. 2000).

We found that flight-capable beetles responded to landscape context more strongly

than flightless species. Such a response was inconsistent with our expectation that

99

flightless beetles should be more dispersal-limited (e.g. Assmann 1999) since

dispersal ability has major consequences on species persistence in fragmented

landscapes (Driscoll et al. 2013). However, as apterous or brachypterous beetles

constitute only a small proportion of the total species pool (43 of 274 species), the

effect of landscape context on their diversity may be diminished by their lower

species richness. Additionally, many ground (carabid) beetle species were large-

bodied and long-legged, and can disperse well across the landscape (Horák et al.

2013) or respond quickly to changes in habitat structure despite their flightlessness.

By comparison, many flight-capable beetles (e.g. Corylophidae, Mordellidae) were

small-bodied and may have smaller spatial requirements given their limited foraging

area (e.g. Jetz et al. 2004). Difference in a species’ spatial requirements may also be

more strongly influenced other factors such as body size, dietary guild and foraging

habits (e.g. Lassau et al. 2005).

What landscape and habitat structure variables best predict bee and beetle richness

and composition?

We found that the landscape variables that best predicted the species richness of

bees and beetles were different (Question 3). For instance, native tree cover was

found to be a relatively important predictor of bee species richness (Figure 5),

consistent with studies in similar landscapes (e.g. Lentini et al. 2012; Threlfall et al.

2015). However, beetle species richness was more strongly influenced by elevation,

distance to water and the topography of the remnants. One explanation for this

difference is that foraging bees tended to be limited by floral resources. In contrast,

topography and proximity to water can interact to influence habitat structural

components on the ground that are important to beetles, such as the amount of

accumulated organic material (e.g. plant debris). Second, being better dispersers

(Francis & Chadwick 2013), bees can respond to changes in the landscape and its

different structural components more rapidly compared to beetles. Third, while

100

beetles are less vagile, they are far more speciose than bees. Beetles would therefore

exhibit a greater spectrum of microhabitat preferences and respond more strongly to

environmental heterogeneity at finer spatial scales (Weibull et al. 2003; Lassau et al.

2005; Barton et al. 2009).

We found that beetle species composition was most correlated to basal stem count,

leaf litter, weed cover and the extent of exposed rocks. This contrasted strongly with

the habitat structural variables most strongly associated with bee composition.

These findings suggest that habitat variables useful as surrogates of diversity for one

insect group should not be extrapolated onto others (Question 3). For example, the

number of trees in each plot, as determined by basal stem counts may affect the

ground layer by contributing fallen leaves and deadwood. This, in turn, creates a

diversity of habitat types for ground-dwelling, saproxylic beetles (e.g. Barton et al.

2009), but not necessarily so for bees (e.g. Roulston and Goodell 2011).

Are bee and beetle species assemblages congruent across the study landscape?

We found that congruency of bee and beetle species richness was limited across the

study landscape (Question 4). The low level of species richness congruence did not

improve even when functional attributes and different landscape contexts were

considered. However, congruency in species compositional similarity performed

better, consistent with other studies (e.g. Su et al. 2004). These findings are

predictable since our analyses have revealed the importance of different landscape

and habitat structure variables to each group. Given that determining the extent of

cross-taxonomic congruency is often a starting point in identifying surrogates of

biodiversity (Su et al. 2004; Caro 2010; Westgate et al. 2014), both bees and beetles

will have limited use as surrogates for each other. Our findings also highlight the

problems of using specific insect groups as surrogates for others (Dauber et al. 2003),

101

especially bees and other pollinators which are often singled out for conservation

prioritisation in agricultural landscapes (Hopwood 2008; Jauker et al. 2009).

Implications for insect conservation in agricultural landscapes

Our study demonstrates that insects are not necessarily good surrogates for each

other in agricultural landscapes. Even though it is recognised that vertebrates can be

poor surrogates for invertebrates (e.g. Oliver et al. 1998; Moritz et al. 2001), there is

equally limited consensus on whether insects can offer better alternatives as

surrogates for other invertebrate groups (e.g. Ricketts et al. 2002; Lovell et al. 2007).

This problem is accentuated by the varied differences in the spatial and ecological

requirements among insect families, as well as the influence of biotic and bionomic

factors on species acting at far smaller scales (Hortal et al. 2010). Such considerations

may be overlooked by conservation planning approaches using occurrence data at

large spatial scales (e.g. Fattorini et al. 2011). Our study revealed that wild bee

diversity has limited use as a surrogate for beetle assemblages, especially when

species richness is used as the metric of diversity. However, if cross-taxonomic

surrogates are to be considered for managing agricultural landscapes to conserve

insects, then measures of compositional (dis)similarity could be more informative

(e.g. Su et al. 2004), especially when comparing assemblages across habitats or

landscapes. Additionally, sets of landscape and habitat variables can be used as

surrogates of specific insect groups. For instance, native tree cover was a relatively

important predictor of bee species richness and thus retaining tree cover in

agricultural landscapes can directly support the conservation of wild bee

assemblages.

Second, our study draws attention to the role of landscape context and its effect on

taxon-specific responses across insect assemblages at the species richness and

compositional level. Changes in landscape context arising from the transformation

102

of the matrix surrounding woodland patches may alter finer-scale aspects of habitat

structure important to different insect assemblages. Such changes in the matrix

could impact the ecology of bee and beetle assemblages differently, especially in

relation to dispersal and foraging resources (e.g. Holzschuh et al. 2006; Jauker et al.

2009; Driscoll et al. 2013). It is therefore important for land managers planning for

biodiversity conservation goals to consider how changes in agricultural land use in

relation to remnant, semi-natural habitats in the wider landscape can lead to

substantially different effects on differing taxa (e.g. Woltz et al. 2012).

Lastly, our findings revealed low congruency of wild bee and beetle assemblages,

and drew attention to the fact that cross-taxonomic patterns of diversity are limited

even among the best-studied insect groups (e.g. Ricketts et al. 2002). Against this

backdrop, we recognise that the diversity of many less charismatic yet ecologically

important insects such as flies (order Diptera) and springtails (order Collembola)

remain poorly understood in the conservation planning context. Therefore, there is a

need to investigate how diversity and abundance patterns of better known insect

groups co-vary with other insects (Lovell et al. 2007) to improve the broader

conservation of invertebrate assemblages (New 1999; Barton et al. 2009). Insights

drawn from such studies will be important in identifying more effective surrogates

of insect diversity (Samways 2007), which can in turn strengthen insect conservation

outcomes in different agricultural landscapes.

Acknowledgements

This study was funded by a Lesslie Foundation research grant and an Australian

postgraduate scholarship to DLY. DBL is supported by an Australian Research

Council Laureate Fellowship. We are grateful to the all the landowners who granted

us access to their properties during the field work. We thank Michael Batley,

Michael Schwarz, Michael Neave and John Ascher for invaluable advice on the

103

identification of bee species, taxonomy and sampling techniques. We thank M. John

Evans for helping with the sorting of beetles. Clive Hilliker assisted with the

preparation of the maps.

References

Assmann T (1999) The ground beetle fauna of ancient and recent woodlands in the

lowlands of north-west Germany (Coleoptera, Carabidae). Biodivers Conserv 8(11):

1499-1517

Barton PS, Manning AD, Gibb H, Lindenmayer DB, Cunningham SA (2009)

Conserving ground-dwelling beetles in an endangered woodland community:

multi-scale habitat effects on assemblage diversity. Biol Conserv 142(8): 1701-1709

Barton PS, Cunningham SA, Macdonald BC, McIntyre S, Lindenmayer DB, Manning

AD (2013) Species traits predict assemblage dynamics at ephemeral resource

patches created by carrion. PloS ONE 8(1): e53961

Barton PS, Pierson JC, Westgate MJ, Lane PW, Lindenmayer DB (2015) Learning

from clinical medicine to improve the use of surrogates in ecology. Oikos 124(4):

391-398

Batáry P, Báldi A, Kleijn D, Tscharntke T (2010) Landscape-moderated biodiversity

effects of agri-environmental management: a meta-analysis. Proc Royal Soc of

London B: Biological Sciences: rspb20101923

Bates D, Maechler M, Bolker B, Walker S (2014) lme4: Linear mixed-effects models

using Eigen and S4_. R package version 1.1-7, <URL: http://CRAN.R-

project.org/package=lme4>.

Batley M, Hogendoorn K (2009) Diversity and conservation status of native

Australian bees. Apidologie, 40(3): 347-354

104

Cane JH, Minckley RL, Kervin LJ, Williams NM (2006). Complex responses within a

desert bee guild (Hymenoptera: Apiformes) to urban habitat fragmentation. Ecol

Appl 16(2): 632-644

Caro T (2010) Conservation by proxy: indicator, umbrella, keystone, flagship and

other surrogate species. Island Press, USA

Dauber J, Hirsch M, Simmering D, Waldhardt R, Otte A, Wolters V (2003)

Landscape structure as an indicator of biodiversity: matrix effects on species

richness. Agr Ecosys Environ 98(1): 321-329

Dirzo R, Young HS, Galetti M, Ceballos G, Isaac NJ, Collen B (2014) Defaunation in

the Anthropocene. Science 345(6195): 401-406

Dollin A, Batley M, Robinson M, Faulkner B (2000) Native Bees of the Sydney

Region: A Field Guide. Australian Native Bee Research Centre

Dollin A, Hogendoorn A, Heard T, Cunningham SA, Rader R, Saunders M, Latty T,

Threlfall C, Smith T, Halcroft M, Lloyd-Prichard D (2016) A practical handbook:

Australian native bees. Department of Primary Industries, New South Wales

Government

Driscoll DA, Banks SC, Barton PS, Lindenmayer DB, Smith AL (2013) Conceptual

domain of the matrix in fragmented landscapes. Trends Ecol Evol 28(10): 605-613

Droeges S (2015) The very handy manual: how to catch and identify bees and

manage a collection. USGS Bee Inventory and Monitoring Lab (BIML), Patuxent

Wildlife Research Center, Beltsville, Maryland, U.S.A.

Dunn RR (2005) Modern insect extinctions, the neglected majority. Conserv Biol

19(4): 1030-1036

FAO (2011) The state of the world’s land and water resources for food and agriculture

(SOLAW) – Managing systems at risk. Food and Agriculture Organization of the

United Nations, Rome and Earthscan, London

105

Fahy O, Gormally M (1998). A comparison of plant and carabid beetle communities

in an Irish oak woodland with a nearby conifer plantation and clearfelled site.

Forest Ecol Manag 110(1): 263-273

Fattorini S, Dennis RL, Cook LM (2011) Conserving organisms over large regions

requires multi-taxa indicators: one taxon’s diversity-vacant area is another taxon’s

diversity zone. Biol Conserv 144(5): 1690-1701

Favreau JM, Drew CA, Hess GR, Rubino MJ, Koch FH, Eschelbach KA (2006)

Recommendations for assessing the effectiveness of surrogate species approaches.

Biodivers Conserv 15(12): 3949-3969

Fischer J, Lindenmayer DB (2007) Landscape modification and habitat

fragmentation: a synthesis. Glob Ecol Biogeogr 16(3): 265-280

Foley JA, DeFries R, Asner GP, Barford C, Bonan G, Carpenter SR, Chapin FS, Coe

MT, Daily GC, Gibbs HK, Helkowski JH, Holloway T, Howard EA, Kucharik CJ,

Monfreda C, Patz JA, Prentice IC, Ramankutty N, Snyder PK (2005) Global

consequences of land use. Science 309(5734): 570-574

Flynn DF, Gogol‐Prokurat M, Nogeire T, Molinari N, Richers BT, Lin BB, Simpson N,

Mayfield M, DeClerck F (2009) Loss of functional diversity under land use

intensification across multiple taxa. Ecol. Lett. 12(1), 22-33

Forestry Corporation NSW (2016) Radiata pine ‘the remarkable pine’. Forest Facts

4(3): 1-3

Francis RA, Chadwick MA (2013) Urban ecosystems: understanding the human

environment. Routledge, U.K.

Greenleaf SS, Williams NM, Winfree R, Kremen C (2007) Bee foraging ranges and

their relationship to body size. Oecologia 153(3): 589-596

Gotelli NJ, Colwell RK (2001) Quantifying biodiversity: procedures and pitfalls in the

measurement and comparison of species richness. Ecol Lett 4(4): 379-391

106

Guiney MS, Oberhauser KS (2008) Insects as flagship conservation species.

Terrestrial Arthropod Rev 1: 111-123

Goslee SC, Urban DL (2007) The ecodist package for dissimilarity-based analysis of

ecological data. J Stat Software 22(7): 1-19

Goslee SC, Urban DL (2015) Dissimilarity-based functions for ecological analysis.

URL: https://cran.r-project.org/web/packages/ecodist/ecodist.pdf\

Hangay G, Zborowski P (2010) A guide to the beetles of Australia. CSIRO Publishing,

Australia

Henle K, Alard D, Clitherow J, Cobb P, Firbank L, Kull T, McCrackeng D, Moritzh

RFA, Niemeläi J, Rebanej M, Wascherk D, Watt A, Young J (2008) Identifying and

managing the conflicts between agriculture and biodiversity conservation in

Europe– A review. Agr Ecosys Env 124(1): 60-71

Hogendoorn K (2011) Development of native bees as pollinators of vegetable seed

crops. Horticulture Australia Ltd, Sydney

Holzschuh A, Steffan‐Dewenter I, Kleijn D, Tscharntke T (2007) Diversity of flower‐

visiting bees in cereal fields: effects of farming system, landscape composition and

regional context. J Appl Ecol 44(1):41-49

Hopwood JL (2008) The contribution of roadside grassland restorations to native

bee conservation. Biol Conserv 141(10): 2632-2640

Horák J, Vodka Š, Pavlíček J, Boža P (2013) Unexpected visitors: flightless beetles in

window traps. J Insect Conserv 17(3): 441-449

Hortal J, Roura-Pascual N, Sanders N, Rahbek C (2010) Understanding (insect)

species distributions across spatial scales. Ecography 33(1):51-53.

Jauker F, Diekötter T, Schwarzbach F, Wolters, V (2009) Pollinator dispersal in an

agricultural matrix: opposing responses of wild bees and hoverflies to landscape

structure and distance from main habitat. Land Ecol 24(4): 547-555

107

Jetz W, Carbone C, Fulford J, Brown JH (2004) The scaling of animal space use.

Science 306(5694): 266-268

Joshi NK, Leslie T, Rajotte EG, Kammerer MA, Otieno M, Biddinger DJ (2015)

Comparative trapping efficiency to characterize bee abundance, diversity, and

community composition in apple orchards. Ann Entomol Soc America 108(5):

785-799

Jukes MR, Peace AJ, Ferris R (2001) Carabid beetle communities associated with

coniferous plantations in Britain: the influence of site, ground vegetation and

stand structure. Forest Ecol Manag 148(1): 271-286

Kati V, Devillers P, Dufrêne M, Legakis A, Vokou D, Lebrun P (2004) Testing the

value of six taxonomic groups as biodiversity indicators at a local scale. Conserv

Biol 18(3): 667-675

Kennedy CM, Lonsdorf E, Neel MC, Williams NM, Ricketts TH, Winfree R,

Bommarco R, Brittain C, Burley AL, Cariveau D, Carvalheiro LG (2013) A global

quantitative synthesis of local and landscape effects on wild bee pollinators in

agroecosystems. Ecol Lett 16(5): 584-599

Kleijn D, Winfree R, Bartomeus I, Carvalheiro LG, Henry M, Isaacs R, Klein AM,

Kremen C, M'gonigle LK, Rader R, Ricketts TH (2015) Delivery of crop pollination

services is an insufficient argument for wild pollinator conservation. Nature

Comm 6: 7414

Klein AM, Vaissiere BE, Cane JH, Steffan-Dewenter I, Cunningham SA, Kremen C,

Tscharntke T (2007) Importance of pollinators in changing landscapes for world

crops. Proc Roy Soc of London B: Biol Sci 274(1608): 303-313

Kremen C, Colwell RK, Erwin TL, Murphy DD, Noss RA, Sanjayan MA (1993)

Terrestrial arthropod assemblages: their use in conservation planning. Conserv

Biol 7(4): 796-808

Larsen, TH, Williams NM, Kremen C (2005) Extinction order and altered community

structure rapidly disrupt ecosystem functioning. Ecol Letters, 8(5): 538-547.

108

Lassau, SA, Hochuli DF, Cassis G, Reid CA (2005) Effects of habitat complexity on

forest beetle diversity: do functional groups respond consistently? Divers Distrib

11(1): 73-82

Le Féon V, Schermann-Legionnet A, Delettre Y, Aviron S, Billeter R, Bugter R,

Hendrickx F, Burel F (2010) Intensification of agriculture, landscape composition

and wild bee communities: a large scale study in four European countries. Agr

Ecosys Env 137(1): 143-150

Legendre P (1993) Spatial autocorrelation: trouble or new paradigm. Ecology, 74(6),

1659-1673.

Lentini PE, Martin TG, Gibbons P, Fischer J, Cunningham SA (2012) Supporting wild

pollinators in a temperate agricultural landscape: maintaining mosaics of natural

features and production. Biol Conserv 149(1): 84-92

Lindenmayer DB, Hobbs RJ (2004) Biodiversity conservation in plantation forests - a

review with special reference to Australia. Biol Conserv, 119: 151-168.

Lindenmayer DB, Likens GE (2011) Direct measurement versus surrogate indicator

species for evaluating environmental change and biodiversity loss. Ecosystems

14(1): 47-59

Lindenmayer DB, Fischer J, Felton A, Crane M, Michael D, Macgregor C, Montague-

Drake R, Manning A, Hobbs RJ (2008) Novel ecosystems resulting from landscape

transformation create dilemmas for modern conservation practice. Conserv Lett

1(3): 129-135

Lindenmayer DB, Barton PS, Lane PW, Westgate MJ, McBurney L, Blair D, Gibbons

P, Likens, GE. (2014) An empirical assessment and comparison of species-based

and habitat-based surrogates: a case study of forest vertebrates and large old

trees. PloS One, 9(2): e89807

Lindenmayer DB, Barton PS, Pierson J (Eds) (2015) Indicators and Surrogates of

Biodiversity and Environmental Change. CSIRO Publishing, Australia

109

Losey JE, Vaughan M (2006) The economic value of ecological services provided by

insects. Bioscience 56(4): 311-323

Lovell S, Hamer M, Slotow R, Herbert D (2007) Assessment of congruency across

invertebrate taxa and taxonomic levels to identify potential surrogates. Biol

Conserv 139(1): 113-125

Matthews EG (1992) A guide to the beetles of South Australia. South Australian

Museum, Adelaide, Australia

McCune B, Grace JB (2002) Analysis of ecological communities. MjM Software

Design, Gleneden Beach, Oregon, USA

McIntyre S, Barrett GW (1992) Habitat variegation, an alternative to fragmentation.

Conserv Biol 6: 146-147

Michener CD (2000) The Bees of the World. The John Hopkins University Press,

Baltimore, U.S.A. & London, U.K.

Morandin LA, Winston ML, Abbott VA, Franklin MT (2007) Can pastureland

increase wild bee abundance in agriculturally intense areas? Basic Appl Ecol 8(2):

117-124

Moritz C, Richardson KS, Ferrier S, Monteith GB, Stanisic J, Williams SE, Whiffin T

(2001) Biogeographical concordance and efficiency of taxon indicators for

establishing conservation priority in a tropical rainforest biota. Proc Roy Soc

London B: Biol Sci 268(1479): 1875-1881

Mortelliti A, Lindenmayer DB (2015) Effects of landscape transformation on bird

colonization and extinction patterns in a large‐scale, long‐term natural

experiment. Conserv Biol 29(5): 1314-1326

NAPPC (2015) North America Pollinator Protection Campaign. URL:

http://pollinator.org/nappc/

New TR (eds.) (2007) Beetle conservation. Springer, Dortrecht, The Netherlands

110

New TR (1999) Untangling the web: spiders and the challenges of invertebrate

conservation. J Insect Conserv 3(4): 251-256

Nichols E, Spector S, Louzada J, Larsen T, Amezquita S, Favila ME (2008) Ecological

functions and ecosystem services provided by Scarabaeinae dung beetles. Biol

Conserv 141(6): 1461-1474

Öckinger E, Lindborg R, Sjödin NE, Bommarco R (2012) Landscape matrix modifies

richness of plants and insects in grassland fragments. Ecography 35(3): 259-267

O’Hara RB, Kotze DJ (2010) Do not log‐transform count data. Methods Ecol Evol

1(2): 118-122

Oksanen J, Blanchet FG, Friendly M, Kindt R, Legendre P, McGlinn D, Minchin PR,

O’Hara RB, Simpson GL, Solymos P, Stevens MHH, Szoecs E, Wagner H (2016)

Community ecology package. URL: https://cran.r-project.org,

https://github.com/vegandevs/vegan

Oliver I, Beattie AJ, York A (1998) Spatial fidelity of plant, vertebrate, and

invertebrate assemblages in multiple‐use forest in eastern Australia. Conserv Biol

12(4): 822-835

PADiL (2016) Pest and disease image library. URL:

http://www.padil.gov.au/pollinators/search?queryType=all

Paradis E, Claude J, Strimmer K (2004) APE: analyses of phylogenetics and evolution

in R language. Bioinformatics 20: 289-290

Potts SG, Biesmeijer JC, Kremen C, Neumann P, Schweiger O, Kunin WE (2010)

Global pollinator declines: trends, impacts and drivers. Trends Ecol Evol, 25(6):

345-353

Prendergast JR, Eversham BC (1997) Species richness covariance in higher taxa:

empirical tests of the biodiversity indicator concept. Ecography 20(2): 210-216

Purtauf T, Dauber J, Wolters V (2005) The response of carabids to landscape

simplification differs between trophic groups. Oecologia 142(3): 458-464

111

Raftery A, Hoeting J, Volinsky C, Painter I, Yeung KY (2015) Bayesian model

averaging. URL: ftp://cran.r-project.org/pub/R/web/packages/BMA/BMA.pdf

Ribera I, Dolédec S, Downie IS, Foster GN (2001) Effect of land disturbance and

stress on species traits of ground beetle assemblages. Ecology 82(4): 1112-1129

Ricketts TH, Daily GC, Ehrlich PR (2002) Does butterfly diversity predict moth

diversity? Testing a popular indicator taxon at local scales. Biol Conserv 103(3):

361-370

Rodrigues AS, Brooks TM. (2007) Shortcuts for biodiversity conservation planning:

the effectiveness of surrogates. Annu Rev Ecol Evol Syst 38: 713-737

Rösch V, Tscharntke T, Scherber C, Batary P (2013) Landscape composition,

connectivity and fragment size drive effects of grassland fragmentation on insect

communities. J Appl Ecol 50(2): 387-394

Roulston TAH, Goodell K (2011) The role of resources and risks in regulating wild

bee populations. Annu Rev Ento 56: 293-312

Rykken JJ, Capen DE, Mahabir SP (1997) Ground beetles as indicators of land type

diversity in the Green Mountains of Vermont. Conserv Biol 11(2): 522-530

Samways MJ (2005) Insect diversity conservation. Cambridge University Press, U.K.

Samways MJ (2007) Insect conservation: a synthetic management approach. Annu

Rev Entomol 52: 465-487

Schulze CH, Waltert M, Kessler PJ, Pitopang R, Veddeler D, Mühlenberg M,

Gradstein SR, Leuschner C, Steffan-Dewenter I, Tscharntke T, 2004. Biodiversity

indicator groups of tropical land-use systems: comparing plants birds and insects.

Ecol Appl 14(5): 1321-1333

Souza DM, Teixeira RF, Ostermann OP (2015) Assessing biodiversity loss due to land

use with Life Cycle Assessment: are we there yet? Glob Change Biol 21(1): 32-47

112

Stanley DA, Gunning D, Stout JC (2013) Pollinators and pollination of oilseed rape

crops (Brassica napus L.) in Ireland: ecological and economic incentives for

pollinator conservation. J Insect Conserv 17(6): 1181-1189

Steffan-Dewenter I, Münzenberg U, Bürger C, Thies C, Tscharntke T (2002). Scale‐

dependent effects of landscape context on three pollinator guilds. Ecology 83(5),

1421-1432

Steffan-Dewenter I, Kuhn A (2003) Honeybee foraging in differentially structured

landscapes. Proc Royal Soc of London B: Biol Sci 270(1515): 569-575

Stork NE, McBroom J, Gely C, Hamilton AJ (2015) New approaches narrow global

species estimates for beetles, insects, and terrestrial arthropods. Proc Nat Acad

Sci 112(24): 7519-7523

Su JC, Debinski DM, Jakubauskas ME, Kindscher K (2004) Beyond species richness:

Community similarity as a measure of cross‐taxon congruence for coarse‐filter

conservation. Conserv Biol 18(1): 167-173

The World Bee Project (2016) The World Bee Project. URL:

http://worldbeeproject.org/

Threlfall CG, Walker K, Williams NS, Hahs AK, Mata L, Stork N, Livesley SJ (2015)

The conservation value of urban green space habitats for Australian native bee

communities. Biol Conserv 187: 240-248

Tscharntke T, Steffan-Dewenter I, Kruess A, Thies C (2002) Contribution of small

habitat fragments to conservation of insect communities of grassland–cropland

landscapes. Ecol Appl 12(2): 354-363

Tscharntke T, Sekercioglu CH, Dietsch TV, Sodhi NS, Hoehn P, Tylianakis JM (2008)

Landscape constraints on functional diversity of birds and insects in tropical

agroecosystems. Ecology 89(4): 944-951

113

Tscharntke T, Clough Y, Wanger TC, Jackson L, Motzke I, Perfecto I, Vandermeer J,

Whitbread A (2012) Global food security, biodiversity conservation and the future

of agricultural intensification. Biol Conserv 151(1): 53-59

Walker KL (1995) Revision of the Australian native bee subgenus Lasioglossum

(Chilalictus) (Hymenoptera: Halictidae). Memoirs of the Museum of Victoria

55(1): 1-423

Walther BA, Moore JL (2005) The concepts of bias, precision and accuracy, and their

use in testing the performance of species richness estimators, with a literature

review of estimator performance. Ecography 28(6): 815-829

Westgate MJ, Barton PS, Lane PW, Lindenmayer DB (2014) Global meta-analysis

reveals low consistency of biodiversity congruence relationships. Nature Comm 5:

3899.

Williams NM, Crone EE, T’ai HR, Minckley RL, Packer L, Potts SG (2010) Ecological

and life-history traits predict bee species responses to environmental

disturbances. Biol Conserv 143(10): 2280-2291

Wintle BA, McCarthy MA, Volinsky CT, Kavanagh RP (2003) The use of Bayesian

model averaging to better represent uncertainty in ecological models. Conserv

Biol 17(6): 1579-1590

Woltz JM, Isaacs R, Landis DA (2012) Landscape structure and habitat management

differentially influence insect natural enemies in an agricultural landscape. Agr

Ecosys Environ 152: 40-49

114

Figures

Figure 1. Simplified conceptual framework showing the linkages between the

different components of our study landscape measured, and how this relates to

habitat and cross-taxonomic surrogacy (see inset).

115

Figure 2. Map of the Nanangroe experimental landscape, with inset map of Australia

showing locations of the woodland remnants studied and pine plantation sites.

116

Figure 3. Sample-based rarefaction curves for (a) wild bee and (b) ground-active

beetle based on 999 random permutations. Black squares represent woodland

remnants in the pine plantation matrix; red triangles represent woodland remnants

in the agricultural matrix; blue circles represent pine plantation sites.

117

Figure 4. NMDS ordination plots of (a) wild bee and (b) ground-active beetle

species composition across the different landscape contexts. (Black square –

woodland remnants in pine, red triangle – woodland remnants in agriculture, blue

circle – pine plantation)

118

Figure 5. (a) Mean site species richness (with standard error bars) for all bees and

beetles, (b) for functionally-defined bee groups classified by nesting requirement

across different landscape contexts and, (c) Mean site morphospecies richness for

functionally-defined beetle groups.

119

Tables

Table 1. Summary information on study site attributes, and mean site-level species

richness for wild bee and beetle in each landscape context

Landscape

context

N Mean area

(hectares ) (± se)

Mean perimeter

(km) (± se)

Mean bee

richness (± se)

Mean beetle

richness (± se)

Woodland

remnant in pine

20 3.880 ± 1.165

0.976 ± 0.117 6.30 ± 0.493 28.06 ± 2.17

Woodland

remnant in

agriculture

23 2.097 ± 0.234 0.809 ± 0.0446 7.28 ± 0.795 22.62 ± 1.98

Pine plantation 5 - - 4.40 ± 0.872 17.20 ± 3.10

120

Table 2. Pairwise MRPP values denoting differences between site-level, wild bee and beetle species composition in the different

landscape contexts. The A statistic is the measure of effect size for each MRPP analysis.

Landscape context

comparison

All wild bees All beetles Ground-

nesting bees

(N = 23)

Above-ground-

nesting bees

(N = 10)

Flightless

beetles (N =

43)

Flight-capable

beetles

(N = 231)

A P A P A P A P A P A P

All landscape contexts 0.0590 0.001 0.0610 0.001 0.0451 0.003 - - 0.0533 0.001 0.0486 0.001

Pine

plantation

Remnants in

pine

0.0406 0.010 0.0511 0.001 0.0207 0.070 - - 0.0079 0.205 0.0419 0.001

Pine

plantation

Remnants in

agriculture

0.0430 0.010 0.0794 0.001 0.0224 0.100 - - 0.0431 0.002 0.0691 0.001

Remnants in

pine

Remnants in

agriculture

0.0442 0.002

0.0266 0.001

0.0411 0.001 0.0677 0.002 0.0463 0.001 0.0179 0.003

121

Table 3. Model parameters for the best five candidate models relating wild bee (57 models)

and ground-active beetle (25 models) site species richness to a set of landscape variables.

Models were selected and ranked by Bayesian Model Averaging (BMA) and their posterior

probability.

Model parameters Post.

Prob.

Slope

estimate

Best Candidate Models

1 2 3 4 5

Wild bee species richness

Intercept 1.000 1.754 1.911 1.721 1.985 1.986 1.797

Native tree cover 0.452 0.0173 - 0.0365 - - 0.0392

Landscape context:

pine

0.255 -0.0479 - - -0.1443 - -

Topography: slope 0.267 -0.0553 - - - -0.1205 -0.1451

BIC value - - -89.38 -88.96 -87.07 -86.62 -86.58

Posterior probability - - 0.159 0.129 0.050 0.040 0.039

Beetle species richness

Intercept 1.000 2.490 2.479 2.462 2.477 2.528 2.513

Native tree cover 0.166 -0.00154 - - - - -0.00804

Elevation 1.000 0.00171 0.00173 0.00182 0.00159 0.00165 0.00174

Distance to water 0.588 0.000224 0.000382 - - 0.000442 0.000366

Topography: slope 0.852 -0.1851 -0.2451 -0.1529 - -0.2748 -0.2438

Topography: ridge 0.142 -0.01118 - - - -0.1374 -

BIC value - - -18.65 -17.22 -15.73 -15.67 -15.48

Posterior probability - - 0.293 0.143 0.068 0.066 0.060

122

Table 4. Significant habitat structure variables for wild bee and ground-active beetle

assemblages across the Nanangroe landscape, identified by fitted vectors on NMDS

ordination axes.

Habitat structure

variable

Wild bee Beetle

NMDS1 NMDS2 R2 NMDS1 NMDS2 R2

Basal count 0.266 0.964 0.130· -0.989 0.149 0.473***

Canopy depth 0.298 0.955 0.189** -0.409 -0.913 0.0286

% Crown affected -0.525 -0.851 0.253** 0.961 -0.277 0.137*

% Blackberry cover -0.249 0.968 0.160* -0.720 -0.694 0.0802

% Exposed rock 0.0507 -0.999 0.279** 0.404 0.915 0.184*

Litter layer 0.387 0.922 0.128· -0.998 0.0683 0.144*

Weed cover -0.0633 0.998 0.0343 -0.992 0.126 0.143*

P ≤ 0.001 ***, P ≤ 0.01**, P ≤ 0.05* , P ≤ 0.1·

123

Supplementary tables and figures

Table S1. List of wild bee species sampled from the Nanangroe landscape

Species name (Subgenus in parenthesis)

Nesting strata Abundance (across landscape)

Family Apidae Amegilla (Notamegilla) chlorocyanea Ground 63 Amegilla (Zonamegilla) asserta Ground 14 Apis mellifera Above ground 22 Thyreus waroonensis Ground 3 Ceratina (Neoceratina) australensis Ground 11 Exoneura sp. Above ground 11 Exoneurella sp. Above ground 2 Brevineura sp. Above ground 44

Family Megachilidae Megachile apicata Above ground 3 Megachile heriadiformis Above ground 51 Megachile (Hackeriapis) oblonga Above ground 2 Megachile (Eutricharaea) serricauda Above ground 1 Megachile semiluctuosa Above ground 3

Family Halictidae Lasioglossum (Chilalictus) lanarium Ground 935 Lasioglossum (Chilalictus) clelandi Ground 166 Lassioglossum (Chilalictus) orbatum Ground 156 Lasioglossum (Chilalictus) cognatum Ground 52 Lasioglossum (Chilalictus) helichrysi Ground 7 Lasioglossum (Chilalictus) hemichalceum Ground 12 Lasioglossum (Chilalictus) imitans Ground 2 Lasioglossum (Chilalictus) instabilis Ground 2 Lasioglossum (Chilalictus) mediopolitum Ground 1 Lasioglossum (Chilalictus) mundulum Ground 34 Lasioglossum (Chilalictus) willsi Ground 2 Lasioglossum (Austevylaeus) sp.1 Ground 1 Lasioglossum (Parasphecodes) hiltacum Ground 3 Lipotriches (Austronomia) australica Ground 1 Lipotriches (Austronomia) flavoviridis Ground 1 Lipotriches (Austronomia) sp.1 Ground 1 Homalictus (Homalictus) urbanus Ground 2 Homalictus (Homalictus) sphecodoides Ground 98

Family Colletidae Euhesma (Euhesma) sp.1 Ground 1

124

Table S2. Full list of beetle families sampled and corresponding morphospecies

richness and abundance for each family, as represented in the Nanangroe landscape.

Family Number of morphospecies Abundance

Aderidae 4 13

Anthicidae 5 299

Anobiidae 3 13

Anthribidae 1 1

Archeocryptidae 2 2

Buprestidae 1 1

Byrrhidae 2 6

Cantharidae 1 1

Carabidae 30 1236

Cerambycidae 2 3

Chrysomelidae 17 29

Clambidae 2 20

Coccinelidae 8 13

Corylophidae 6 119

Curculionidae 26 142

Dermestidae 3 72

Elateridae 9 57

Histeridae 2 53

Latriidae 5 38

Leiodidae 6 11

Limnichidae 1 1

Lycidae 2 2

Melyridae 2 39

Mordellidae 8 141

Nitidulidae 4 89

Phalacridae 1 1

Psephalidae 11 74

Ptilidae 3 5

Scarabaeidae 33 318

Scraptiidae 2 9

Scydmaenidae 5 68

Silphidae 1 3

Silvanidae 1 1

Staphylinidae 35 588

Tenebrionidae 24 158

Trogidae 1 90

125

Table S3. List of vegetation and habitat structural variables with their accompanying

descriptions.

Habitat structural

variable

Description (plot size: 20m by 20m)

Dominant Cover % cover of plants which height are taller than 10m

Sub-dominant Cover % cover of plants which height is between 2-10m

Shrub Cover % vegetation cover of <2m in height

Grass Cover % cover of native and exotic annual and perennial grasses

Grass Height Estimate of average height of grasses

Ground Cover % cover of native broad leaves, herbs and forbs

Weed Cover % cover of weeds (exotic ground cover such as broad leaves, herbs and

forbs)

Litter Layer % cover of leaf litter

% Blackberry % cover of blackberry

% Exposed Rock % cover of exposed rock

Logs 10-20cm Number of logs >10cm in diameter, and length is >10cm and <20cm

Logs 20-30cm Number of logs > 10cm in diameter, and length >20cm and <30cm

Logs 30-40cm Number of logs >10cm in diameter, and length >30cm and <40cm

Logs 40-50cm Number of logs >10cm in diameter, and length >40cm and <50cm

Logs >50cm Number of logs >50cm in length, and >10cm in diameter

Number Of Strata Number of strata. Maximum of four strata (ground cover, under-storey,

mid-storey and over-storey)

Number Of Trees Number of trees

Regrowth % cover of regrowth

Stem diameter 1-5cm Number of stems <5cm in diameter

Stem diameter 6-10cm Number of stems >5cm, and <10cm in diameter

Stem diameter 11-20cm Number of stems >10cm, and <20cm in diameter

Stem diameter 21-30cm Number of stems >20cm, and <30cm in diameter

Stem diameter 31-40cm Number of stems >30cm and <40cm in diameter

Stem diameter 41-50cm Number of stems that >40cm, and <50cm in diameter

Stem diameter >51cm Number of stems >50cm in diameter

Stumps Number of stumps

Basal Count Measure of the number and the size of trees in a stand. obtained by

holding a 1cm wide gauge at 50cm away from eye

Canopy Height Height of the biggest trees

Canopy Depth Distance (m) from top of foliage to the bottom of foliage of the biggest tree

Hollow Count Number of tree hollows

Mistletoe Count Number of clumps of mistletoe

Dead Trees Number of dead trees

Dead Shrubs Number of dead shrubs

% Crown Affected Estimated extent of dieback in terms of percentage of unhealthy canopy

cover (affected by dry weather or insect attack)

126

Table S4. Summary of mixed- and fixed-effects models relating landscape context to

insect species richness for each defined insect group (to 3 s.f.).

Parameter Model parameter estimates

Mean difference (+ s.e.) Z-value P-value

Total bee species richness

Random-effects variance = 0.0668 residual (d.f.) = 39

Intercept 1.860 ± 0.163 11.416 <0.001

Remnants in agriculture Pine plantation -0.326 ± 0.294 -1.111 0.267

Remnants in pine Pine plantation -0.236 ± 0.248 0.950 0.342

Remnants in pine Remnants in

agriculture

-0.0905 ± 0.229 -0.395 0.693

Ground-nesting bee species richness

Random-effects variance = 0.0232 residual (d.f.) = 39

Intercept 1.635 ± 0.137 11.900 <0.001

Remnants in agriculture Pine plantation -0.0978 ± 0.274 -0.357 0.721

Remnants in pine Pine plantation -0.0714 ± 0.255 0.281 0.779

Remnants in pine Remnants in

agriculture

-0.0264 ± 0.190 -0.139 0.890

Above ground-nesting bee species richness

Random effects variance = 0.1774 residual (d.f.) = 35

Intercept 0.358 ± 0.336 1.066 0.286

Remnants in agriculture Remnants in pine -0.159 ± 0.467 -0.340 0.734

Total beetle species richness

Intercept 3.119 ± 0.0459 67.973 <0.001

Remnants in agriculture Pine plantation -0.274 ± 0.117 -2.337 0.0194

Remnants in pine Pine plantation -0.489 ± 0.117 4.194 <0.001

Remnants in pine Remnants in

agriculture

0.215 ± 0.0639 3.370 <0.001

Flight-capable beetle species richness

Intercept 2.929 ± 0.0504 58.071 <0.001

Remnants in agriculture Pine plantation -0.235 ± 0.127 -1.852 0.0641

Remnants in pine Pine plantation 0.553 ± 0.125 4.415 <0.001

Remnants in pine Remnants in

agriculture

0.318 ± 0.0686 4.637 <0.001

Flightless beetle species richness

Intercept 1.362 ± 0.110 12.335 <0.001

Remnants in agriculture Pine plantation -0.487 ± 0.309 -1.575 0.115

Remnants in pine Pine plantation -0.0282 ± 0.327 -0.0860 0.931

Remnants in pine Remnants in

agriculture

-0.515 ± 0.190 -2.714 <0.01

127

Table S5. Within-group (landscape context) dissimilarity for wild bee and ground-

active beetle assemblages.

Landscape context Wild bee Beetle

N Group-level

dissimilarity

N Group-level

dissimilarity

Pine plantation 5 0.451 5 0.580

Woodland remnants in pine 20 0.519 18 0.801

Woodland remnants in agriculture 20 0.611 21 0.807

128

Figure S1. Bar chart comparing observed species richness and three non-parametric

estimators (Chao1, Jackknife 1, Bootstrap) of species richness for (a) bees and (b)

beetles.

129

Figure S2. Relationship (linear regression line and 95% confidence region) of three continuous landscape variables with respect to

species richness of wild bees (a, b, c) and ground-active beetles (d, e, f).

130

Figure S3. Correlations of species richness for functionally-defined bee and beetle

groups based on Spearman’s rho. No correlation were statistically significant.

131

III. Yong, D. L., Barton, P. S., Ikin, K., Evans, M. J., Crane, M., Okada, S.,

Cunningham, S. A. & Lindenmayer, D. B. (In review). Validating cross-

taxonomic surrogates for biodiversity conservation in farming landscapes - a

multi-taxa approach. Biological Conservation.

132

Validating cross-taxonomic surrogates for biodiversity

conservation in farming landscapes - a multi-taxa approach

Abstract

Cross-taxonomic surrogates are often used in conservation planning given the

unfeasibility of inventorying large suites of taxa. However they are seldom tested

rigorously at the spatial scales at which conservation management occurs. We

evaluated the effectiveness of five ecologically diverse taxa as cross-taxonomic

surrogates in woodland patches within an agricultural landscape. We first compared

species richness and compositional heterogeneity across taxa before testing for

cross-taxonomic congruence using a correlative approach. To validate surrogate taxa

identified this way, we quantified how well each taxon incidentally represented

other taxa in sets of woodland patches selected using a complementarity-based

approach. We found that the beetle assemblage exhibited the highest compositional

heterogeneity among the taxa. Significant pairwise associations were found between

some taxa (bird-bee), but no single taxon was strongly correlated with all other taxa.

Patch sets prioritised for beetles represented other taxa best, but were the costliest

and required conserving the largest amount of woodland. This was followed by

birds. Patch sets prioritised for wild bees and trees incidentally represented other

taxa poorly, but were far less costly. Our study highlights the importance of taxon-

specific patterns of compositional turnover to how remnant box-gum grassy

woodland should be prioritised for farmland conservation, a consideration not

immediately apparent in correlative analyses of surrogacy. Taxa with high

compositional heterogeneity will demand the conservation of more, and larger

woodland patches. Therefore, while species-rich taxa such as beetles are ideal as

surrogates, practical considerations on the extent of land spared for biodiversity in

landscapes prioritised for agricultural production cannot be overlooked.

133

1. Introduction

Land use change driven by agricultural expansion and intensification are among the

leading causes of biodiversity loss worldwide (Foley et al. 2011; Alexander et al. 2015).

Presently, a large proportion of the world’s agricultural land is already used for

grazing livestock, with permanent pastures covering nearly a quarter of the world’s

land surface (Wirsenius et al. 2010; FAOSTAT 2014). Intensification of agricultural

production in existing farming landscapes is expected to exacerbate biodiversity

declines (Benton et al. 2003; Donald et al. 2006; Cunningham et al. 2013). Therefore,

effective conservation of biodiversity will necessitate extension of conservation

initiatives into agricultural landscapes, and underpinned by robust ecological

research (Norris 2008; Kay et al. 2016).

Knowledge of biodiversity patterns is essential to informing on the consequences of

land use change, and guiding conservation management decisions (Margules &

Pressey 2000; Ferrier 2002; Phalan et al. 201; Guisan et al. 2013). Given that it is

neither cost-effective nor practical to exhaustively inventory large suites of taxa,

there is a need to adopt surrogate approaches drawing on more easily gathered data

to guide biodiversity conservation (Rodrigues & Brooks 2007; Caro 2010;

Lindenmayer et al. 2015). Such surrogate approaches are usually grounded on the

presumption that a measured subset of biodiversity components in a landscape can

provide useful information on other components of biodiversity, therefore allowing

variation in other components of biodiversity to be predicted (Heino 2010; Larsen et

al. 2012; Barton et al. 2015). Many surrogate approaches adopted in conservation

management and monitoring use species data, often in combination with habitat

and environmental data (e.g. Grantham et al. 2010; Lindenmayer et al. 2014). Over

time, the interest in using species-based surrogates to support conservation

decision-making has fuelled a large amount of research to evaluate their

effectiveness.

134

Because cross-taxonomic surrogates offer an expedient means to assess biodiversity

broadly for conservation planning, easily surveyed taxa such as birds have been

widely proposed as surrogates (e.g. Carrascal et al. 2012; Eglington et al 2012; Di

Minin & Moilanen 2014; Ikin et al. 2016). However, while some studies endorse the

use of cross-taxonomic surrogates (e.g. Larsen et al. 2012), others have highlighted

problems (e.g. Andelman & Fagan 2000; Paavola et al. 2006). First, there is

increasing evidence of how spatial scale, grain and resolution can shape the extent of

correlation between different taxa, thus compromising their effectiveness as

surrogates for other groups (e.g. Hess et al. 2006; Paavola et al. 2006; Westgate et al.

2014). Second, differences in ecology and responses to environmental variables can

be expected to drive taxon-specific turnover patterns in space and time (e.g.

Turtureanu et al. 2014), reducing the strong cross-taxonomic congruence expected of

a good surrogate (Yong et al. 2016). Third, the diversity of criteria, concepts and

approaches used to evaluate the effectiveness of biodiversity surrogates across

different studies has rendered it difficult to draw a consensus on what constitute a

good surrogate (e.g. Favreau et al. 2006; Hunter et al. 2015). Put together, these

problems draw attention to the need to identify better sets of biodiversity

surrogates, and cross-validate their effectiveness through different analytical

approaches (Favreau et al. 2006).

In this study, we tested the effectiveness of a cross-taxonomic surrogate approach in

guiding conserving planning for woodland biodiversity in a compact, agricultural

landscape. The conceptual framework for our study was underpinned by four

questions (Figure 1) and grounded systematically on field inventorying, initial

identification of surrogate taxa, and cross-validation of these surrogate groups under

a systematic conservation planning scenario. First, we asked: (1) which pairs of taxa

show strong cross-taxonomic congruence? To identify them, we inventoried two

vertebrate groups (birds, herpetofauna), two insect groups (bees, beetles) and one

plant group (trees). We then applied a correlative approach to assess the degree of

135

cross-taxonomic association (i.e. cross-taxonomic congruence) in species richness

and composition among groups (Sauberer et al. 2004; Su et al. 2004; Rooney &

Azeria 2015). We hypothesized that taxa showing stronger cross-taxonomic

congruency at the species richness and composition level could perform better as

surrogates for other taxa.

Second, we asked: (2) How effective are cross-taxonomic surrogates in incidentally

representing the occurrences of the other taxa in sets of woodland patches

prioritised for the surrogate? This question is important because it allows taxonomic

surrogates initially identified from a correlative approach to be validated under a

systematic conservation planning scenario. To do so, we used a complementarity-

based, site-selection approach (see Table 1 for definitions of terms) to identify near-

optimal sets of remnant habitat patches using representation targets set for each

taxa a priori. We then determined how well other taxa were represented in the patch

sets selected for the surrogate taxon (e.g. Sætersdal et al. 2004; Albuquerque & Beier

2016). A desirable surrogate taxon could be expected to capture a high proportion of

the representation targets for other taxa (Larsen et al. 2012; Di Minin & Moilanen

2014), without requiring the conservation of large numbers of woodland patches.

Third, we asked: (3) How similar are these near-optimal sets of woodland patches

selected for each taxon, and at each defined representation target. To do so, we

compared the sets of habitat patches selected for each taxon at each target by

assessing the degree of spatial correspondence (as measured with dissimilarity

distances) in patch sets between taxa (e.g. Ikin et al. 2016). Since many species in

farming (disturbed) landscapes can be expected to be wide-ranging generalists given

the effects of biotic homogenisation (Ekroos et al. 2010), we predicted that

differences between sets of habitat patches selected for each taxon to be low.

136

Finally, we asked: (4) How does the relative proportion of remnant woodland

retained in the landscape to conserve each taxon compare with other taxa across a

range of representation targets? Determining the area of woodland patches retained

for each taxon provides a proxy of relative cost and compromised production

opportunities in conservation planning scenarios using different taxa as surrogates.

Tackling this question is essential because it could provide insights on the trade-offs

in opportunity costs of using different taxonomic surrogates for conserving

biodiversity in agro-ecosystems by informing on the extent of remnant habitat in

agricultural landscapes to be conserved by land-sparing (e.g. Fischer et al. 2008).

2. Methods and materials

2.1. Study area and design

The Nanangroe landscape (34°58'S, 148°28'E) consists of approximately 30,000 ha of

agricultural (i.e. grazing) land and exotic Monterey pine Pinus radiata plantations

(See map: Figure 2). Most of the original box-gum grassy woodlands have been

cleared in the past two centuries for agriculture, leaving numerous scattered

remnant patches across the landscape (Lindenmayer et al. 2008). In 1998, the

landscape matrix surrounding many of these remnants was transformed by the

establishment of extensive plantations of pine (Lindenmayer et al. 2008). As a result,

the remnant woodland patches in our study landscape are embedded either within a

matrix of pasture actively grazed by livestock or a matrix of dense pine plantations.

Permanent transects were marked and established at all study patches prior to the

commencement of the study. In woodland patches exceeding one hectare, a 200m

long transect was established while 100m long transects were established for patches

smaller than one hectare. For this study, a total of 42 remnant woodland patches in

both kinds of matrix were chosen to represent the full range of patch size classes.

137

2.2. Biodiversity sampling

We conducted field surveys of four different animal taxa in our study landscape.

These taxa included birds, reptiles and amphibians (collectively considered in our

study as ‘herpetofauna’), wild bees and ground-dwelling beetles. The species

composition of the dominant trees (i.e. tree assemblage) in each of our study

woodland patches were identified and recorded as part of a detailed vegetation

survey last carried out in spring 2015.

2.2.1. Bird sampling

We sampled bird occurrence and diversity at each woodland patch using three, five-

minute point counts along an established transect. Point counts were conducted

between 0500–1000hrs during the middle of the Austral spring (October-November

2014). At each point, observers recorded the numbers of individual bird species

heard or seen within a 50m radius. Species observed in flight at the sampling point

was excluded from the survey. Each point was re-sampled by a different observer on

another day during the survey period to minimise detection biases resulting from

weather and variation in identification skills between different observers.

2.2.2. Herpetofauna sampling

As both amphibians and reptiles were surveyed using the same method, we pooled

these two taxa together into one group, and defined them collectively as

“herpetofauna”. To survey amphibian and reptiles, we conducted standardised, area-

constrained searches at three points along each transect once a year in early spring

(October – November 2014). During the establishment of the transects, artificial

substrates consisting of corrugated iron sheets, hardwood timber sleepers and roof

tiles were placed at the 0m, 100m and 200m points to simulate microhabitats for

small ground reptiles like skinks and geckos. Active searches for amphibians and

138

reptiles were carried out by turning over logs, rocks and the artificial substrates

placed at each point along the transect during a survey. Reptiles observed while

walking along the transect between each point were also recorded as having

occurred in the patch.

2.2.3. Wild bee sampling

To sample wild bee occurrence and diversity, we used blue vane traps (e.g. Lentini et

al. 2012; Joshi et al. 2015). We sampled all 42 woodland patches at the midpoint of

each line transect with two traps at each site. Traps were set in trees and placed

approximately 20m apart from where they were suspended at about 1.5-2.0m above

ground. Bee sampling was conducted for 14 days from November to December 2014

and in tandem with beetle sampling. At the end of the sampling period, the traps

were retrieved and all bees were preserved in 70% ethanol before species-level

sorting. Bees that were difficult to identify were: (1) carefully separated from the

other insects, (2) washed in detergent and, (3) blown-dry before being prepared in a

reference collection for subsequent species-level identification based on the

methodology recommended in Droeges (2015). Most bees were identified to the

species-level using the online database, Pest and Diseases Image Library (PaDIL

2016) and major bee identification keys (e.g. Walker 1995; Michener 2000). All bees

were then checked by a bee taxonomist (Michael Batley, Australian Museum) for

accuracy.

2.2.4. Ground-dwelling beetle sampling

To survey ground-dwelling beetles, we used non-baited pitfall traps. Pitfall traps

were placed in pairs along four rows at the centre of the transect within each

woodland patch. Each pitfall trap consisted of a plastic container of 5.0cm diameter

and 7.5cm depth. Traps were filled with 100ml of ethylene glycol, which functioned

both as a killing agent and a preservative. To increase the invertebrate catch rate, we

mounted plastic drift fences (1.0m x 0.2m) along each pair of traps. Trapping was

139

conducted from November to December 2014, with every trap opened for 14 days.

Beetle specimens were preserved in 70% ethanol before being sorted to family and

morphospecies level using a stereo microscope and relevant beetle identification

keys (e.g. Matthews 1980; Hangay & Zborowski 2010). Vouchered specimens of each

morphospecies were assembled into a reference collection for identification. Highly

similar morphospecies from species-rich families such as Staphylinidae and

Carabidae were further checked for the accuracy of identifications. The beetle

dataset was then partitioned into two for analysis: (1) the full dataset which

contained every beetle morphospecies, and (2) a subset of the dataset which

excluded very rare species represented by singletons.

2.3. Statistical analysis

Due to the differences in ecological attributes and sampling approach across the

taxa, we explored the potential of each taxon as a surrogate by comparing their

species diversity patterns across the study landscape. First, we pooled data for

sampling points within each site to calculate species richness at the site level (α-

diversity) for each woodland patch. We then pooled species richness from all

woodland patches across the landscape to estimate γ-diversity. To estimate the

completeness of our sampling effort, we calculated and plotted smoothed

accumulation curves based on observed species richness for each taxon. To quantify

detection rates, we defined observed species richness as a proportion of the mean of

four non-parametric estimator of species richness. To estimate the degree of species

compositional heterogeneity for each taxon, we calculated mean site species

richness (α-diversity) as a proportion of the landscape-level species richness (γ-

diversity).

2.3.1. Correlations of species richness and composition (Question 1)

140

To identify which pairs of taxa showed strong cross-taxonomic congruence at the

landscape scale, we used a combination of Spearman’s rank correlations and partial

Mantel tests. The correlation strength in species richness between two taxa is

frequently used as a metric of cross-taxonomic congruence (e.g. Hess et al. 2006;

Wolters et al. 2006; Duan et al. 2016). To assess the level of correlation between

patch-level species richness for each pairwise combination of taxa, we calculated the

Spearman’s correlation coefficient (e.g. Similä et al. 2006). Spearman’s correlation

was used instead of Pearson’s correlation as the sample size for number of sites was

small (N = 42) and did not meet parametric assumptions based on the Shapiro-Wilk

test (e.g. Royston 1983).

We used partial Mantel tests to assess the strength of cross-taxonomic congruence

in compositional dissimilarity between each pairwise combination of taxa (Landeiro

et al. 2012). Partial Mantel tests were used because they measure the correlation

between two matrices (Su et al. 2004) after accounting for spatial variation

associated with a third matrix of Euclidean distances, thus addressing the potential

issue of spatial autocorrelation. We first quantified species compositional

dissimilarity for each taxon across the full set of habitat patches using the Jaccard

similarity index, which is based on absence-presence data (Magurran 2004). We

then performed the partial Mantel test for all pairwise combinations of taxa. The

significance of each Mantel test was assessed using 999 permutations.

2.3.2. Analysis of cross-taxonomic surrogate representation among patches (Question

2, 4)

We used a complementarity-based, site-selection approach to prioritise sets of

woodland patches that best represented the species richness and occurrence of each

taxon. In such an approach, woodland patches are added to the patch set based on

the level of complementarity in species composition (absence/presence) with

respect to subsequent patches until the representation target for this taxon is met.

141

To quantify incidental representation of other taxa in the patch set selected for the

surrogate taxon, we calculated the total occurrences of other (target) taxa

represented in these patch sets as a proportion of their total occurrences in the

whole landscape (Question 2) and the corresponding amount of woodland area

needed to best conserve each taxa (Question 4).

A total of 399 conservation features was established from the five taxa compared.

We defined the presence of each species in each woodland patch as a ‘conservation

feature’ (Game & Grantham 2008; Ardron et al. 2010) (See Table 1 for definitions).

We then set representation targets for each taxon at 10% intervals, and ranging from

10% to 80%. We chose not to include representation targets of 90–100% as it is not

realistic to retain the majority of habitat patches in active, production landscapes

such as ours (Ikin et al. 2016). For each taxon and at each representation target, we

ran 100 iterations using the simulated annealing algorithm implemented in Marxan

to identify the best patch set for each taxon (Game & Grantham 2008; Ball et al.

2009). We did not set any constrains on the number of patches (sites), therefore

allowing as many patches as required to be included in each solution to meet pre-

defined representation target. The ‘species penalty factor’, a Marxan numerical

parameter that measures the costs for failing to meet targets for each conservation

feature (Game & Grantham 2008) was kept constant for all conservation features,

thereby not giving added weightage to any single species in the landscape.

To compare incidental representation of other taxa by the surrogate taxon, we used

the patch set in the best solution (hereafter as ‘best patch set’) out of a total of 100

solutions identified by Marxan to meet representation targets for the surrogate

taxon. Based on this solution, we then calculated the total number of conservation

features for all other taxa represented in the set, while recording if a priori

representation targets set for the surrogate taxon have been correspondingly met for

the target taxa. To quantify representation of each target taxa in the best patch sets

142

selected for the surrogate taxon, we calculated species occurrences of each taxon

captured in the patch set as a proportion of the total species occurrences in all

patches. Lastly, to compare the amount of woodland area needed to best conserve

each taxon, we calculated the sum of woodland area (in hectares) in the best patch

sets for each taxon, and at every representation target.

2.3.3. Similarity in best patch sets across taxa (Question 3)

To compare the level of similarity or spatial correspondence among the best patch

sets selected for all taxa, we performed cluster analysis using an agglomerative

hierarchical clustering approach (Question 3). First, we created a distance matrix

based on the Jaccard similarity coefficient for each taxon, and across four

representation targets (20%, 40%, 60%, and 80%) using the function ‘distance’.

Jaccard similarity was calculated based on the presence or absence of a patch (site)

in the best solution for each representation target. We then implemented the

function ‘hclust’ using the complete-linkage clustering method. Pairs of clusters

separated by the shortest distances were thus combined in the cluster dendrogram,

allowing the extent of similarity between all patch sets for the five taxa to be

visualised. Cluster analysis and partial Mantel correlations were carried out using

the ‘ecodist’ package (Goslee & Urban 2007), available on the R platform (R

Development Core Team 2013).

3. Results

3.1. Assemblage diversity for taxa sampled

We recorded a total of 77 bird, 21 herpetofauna, 31 bee, 258 beetle and 9 tree species

or morphospecies in the Nanangroe landscape. Based on smoothed species

143

accumulation curves (Figure 3; Supplementary Table S1), our sampling effort was

fairly complete for all taxa. Patch-level species richness (α-diversity) as a proportion

of landscape-level species richness (γ-diversity) was lowest for beetles at 9.98%

(Figure 4; Supplementary Table S1). If (rare) beetles represented by singletons were

excluded, this rose to 13.43%. By comparison, each woodland patch supported

16.98% of the total bird, 20.22% of the total bee, and 22.49% of the total tree species

pool (see Supplementary Table S2).

3.2. Correlations of species richness and composition (Question 1)

We found that only bird species richness was correlated with bee species richness at

the patch level (Spearman’s ρ = 0.309, P < 0.05), but not with species richness of any

other taxa. Herpetofauna, beetle and tree species richness was not correlated with

that of other taxa we compared (Figure 5a). As with species richness correlations, we

found that bird species composition was significantly correlated with bee species

composition (partial Mantel R = 0.207, P < 0.01) but not with any other taxa (Figure

5b). Similarly, herpetofaunal species composition was only significantly correlated

with that of beetles (partial Mantel R = 0.137, P < 0.05) and no other taxon. Exclusion

of rare beetles weakened this association slightly (partial Mantel R = 0.127, P < 0.05).

Bee species composition was significantly correlated with that of beetles (partial

Mantel R = 0.128, P < 0.05). This correlation was strengthened if rare beetles were

excluded (partial Mantel R = 0.140, P < 0.05). Compared with the four animal groups

inventoried, tree species composition at the patch level was not significantly

correlated with that of any other taxa.

3.3. Incidental representation of target taxa by the surrogate taxon (Question 3)

144

We found that representation targets for other taxa were rarely met in the best patch

sets selected for the surrogate taxon (Figure 6). An exception to this is that the patch

sets selected to meet representation targets for the beetle assemblage (Figure 6d)

were also able to incidentally represent targets of trees from 10%–70%. Patch sets

selected to meet the full range of representation targets for the beetle assemblage

also met representation targets for the subset of the beetle assemblage excluding

rare species (Figure 6f).

We found that the best patch sets identified to meet representation targets for both

sets of beetle data incidentally represented other taxa better than equivalent patch

sets for all other taxa. For example, the best patch sets selected for the full beetle

assemblage were able to incidentally represent over 70% of all bird occurrences

across the landscape for the full range of targets, reaching 91.23% in the best patch

set that captured 80% of beetle targets (Figure 6a). Likewise, the best patch sets

selected for the beetle assemblage excluding rare species were able to incidentally

represent 44.3% of bird occurrences at a 20% target for beetles, and as high as 97.0%

for birds when the beetle target was 80%.

After beetles, birds were the next best group in incidental representation of other

taxa. Although representation targets for other taxa were seldom met in bird patch

sets (except for bees at a 60% target for birds), the best patch sets selected to meet

representation targets for the birds were able to achieve consistently high incidental

representation of other taxa. Moreover, incidental representation of the

herpetofauna in patch sets selected for birds were as high as patch sets selected for

the beetle assemblage excluding rare species. For example, the best patch set

selected to meet a 10% representation target for birds was able to incidentally

represent 55.46% of herpetofauna occurrences (Figure 6b). Similarly, the best patch

sets selected to meet representation targets for the bird assemblage were also able to

achieve high incidental representation for bees (Figure 6c). Across a range of bird

145

representation targets from 10% to 80%, incidental representation for bees in the

patch set ranged from 47.72% to as high as 89.0% (Figure 6c).

Compared to beetles and birds, wild bees, trees and herpetofauna were less efficient

in incidentally representing other taxa. Incidental representation of other taxa based

on the best patch sets for bees, trees or the herpetofauna was far lower than birds or

beetles at equivalent target representation thresholds. For instance, the best patch

sets selected to meet bee representation targets consistently achieved lower

incidental representation of other taxa than the patch sets selected for birds (Figure

6a), beetles (Figures 6e, 6f), herpetofauna (Figure 5b) and trees (Figure 6d) at similar

representation targets.

3.4. Similarity in patch sets for each taxon (Question 4)

We found that the best patch sets selected for each taxon were very dissimilar at low

representation targets ranging from 20-40%, but became increasingly similar at

higher representation targets (Figure 7). For instance, the best patch sets selected to

meet representation targets for wild bees from 20-60% were more similar to each

other than with that of any other taxon. Likewise, patch sets selected to meet

representation targets for the herpetofauna from 20-60% were more similar to each

other. At high representation targets, the best patch sets for the five taxa become

increasingly overlapping. For instance, the best patch sets selected to meet a range

of representation targets from 60%–80% for beetles were most closely clustered with

patch sets selected to meet similar targets for both the bird and tree assemblages.

3.5. Relative costs in retaining habitat patches in agricultural landscapes (Question

4)

146

We found that the best patch sets needed to conserve beetle assemblages in our

landscape were the most costly among the five taxa. Conserving 10%–60% of the

representation targets for the full beetle assemblage required maintaining more than

80% of the total area of woodland patches regardless of representation targets

(Figure 8). However, if rare beetles were excluded, we found that the relative cost to

capture the range of representation targets from 10%–50% became substantially

lower. While the relative cost to conserve this subset of the beetle assemblage was

similar to that for birds for targets ranging from 10%–50%, it became increasingly

different from birds at representation targets exceeding 50%.

The relative costs of conserving woodland patches for wild bees and herpetofauna

were lower than beetles and birds, as well as trees at higher representation targets.

For instance, only 60% of the total area of habitat patches are needed to meet

representation targets of 80% for both taxa. While the cost of conserving habitat

patches for trees were low when the representation target is low (10-20%), the cost

increases progressively with higher representation targets and exceed that for bees

and herpetofauna at representation targets above 40%.

4. Discussion

4.1. Species richness and composition

Our study revealed that relatively high species richness of vertebrate and

invertebrate assemblages can persist in remnant patches of box-gum grassy

woodland embedded within a wider matrix of agriculture-monoculture landscapes.

Mean patch species richness was highest for the beetles, but its proportion of total

147

landscape species richness (γ-diversity) was the lowest among the taxa (9.98% per

patch). The low proportion of total species richness of beetles at each patch suggests

that compositional heterogeneity of the beetle assemblage is the highest relative to

other taxa (e.g. Soininen et al. 2007). By contrast, each patch supported lower

compositional turnover of birds and bees, given the high proportion of species in the

landscape occurring in each patch. Such differences in compositional heterogeneity

underscore taxon-specific responses to habitat or other abiotic gradients at the

landscape scale (e.g. Benton et al. 2003; Lovell et al. 2007). These patterns may in

turn be driven by the dispersal ability of different taxa or other ecological traits

specific to each group (Cadotte & Fukami 2005; Cadotte 2006; Soininen 2010) and in

relation to the landscape (Janssen et al. 2016). Good dispersers like birds and bees

can be expected to show lower compositional turnover across equivalent spatial

scales when compared to less vagile groups such as beetles, reptiles and amphibians

(Baselga et al. 2012; Qian & Ricklefs 2012).

When framed in the context of cross-taxonomic surrogacy, important implications

can be inferred from species compositional turnover and other measures of beta-

diversity (Beier et al. 2016) for individual taxa. Species groups and taxa with different

spatial patterns of diversity (e.g. composition turnover) will require different degrees

of comprehensiveness in spatial prioritisation to plan for their conservation (Ferrier

2002; Si et al. 2015). Inevitably, taxa with high species richness and compositional

heterogeneity will require more sites in any reserve network to conserve them

effectively (Ryti 1992; Ikin et al. 2016). On the one hand, prioritising such taxa as

surrogates to conserve other groups could translate into a need for higher allocation

of land (e.g. through land sparing) in agricultural landscapes (e.g. Fischer et al.

2008) to meet conservation targets. On the other, it is important to recognise how

differences in diversity patterns specific to each taxon could compromise the

effectiveness of certain taxa as surrogates (e.g. Part & Soderstrom 1999). For

example, conservation planning and prioritisation based on taxa with highly

148

dissimilar turnover patterns could lead to high representation of some taxa in a

reserve network, but inadequate or uneven representation of other taxa.

4.2. Correlations of species richness and composition (Question 1)

Cross-taxonomic surrogates have an important role to play in supporting

conservation planning, especially in identifying priority areas for conservation.

Species richness and composition correlations have been used widely as a first step

to guide the identification of cross-taxonomic surrogates (Sætersdal et al. 2004;

Gardner et al. 2008; Landeiro et al. 2012). As with other studies (e.g. Billeter et al.

2008), we found that no one taxon was a good surrogate for all other taxa. Of five

groups assessed, only birds showed strong congruence in species richness with wild

bees, underscoring the potential of either taxon as a cross-taxonomic surrogate (e.g.

Sauberer et al. 2004). This association between bird and wild bee diversity was

further corroborated by strong correlation at the species composition level. Such a

cross-taxonomic association in diversity may arise as a result of the similar (high)

dispersal ability of both groups. At the same time, wild bee and bird assemblages

may respond to similar biotic gradients such as shared food resources (e.g. flowers)

(Lovell et al. 2007). Other significant pairwise associations between taxa were also

revealed through species composition correlations (e.g. Su et al. 2004; Heino 2010).

Significant associations between species composition of herpetofauna and ground-

dwelling beetles further underlined the potential of these two terrestrial groups as

surrogates for each other.

Generally, species-rich groups like beetles and birds exhibited stronger cross-

taxonomic associations with other taxa, while pairs of species-poor groups tend to

be weakly correlated. One explanation that could account for such a pattern is the

similar responses in compositional turnover of different groups to habitat and

environmental gradients (e.g. Lovell et al. 2007; Duan et al. 2016) at the landscape

149

scale. Additionally, despite differences in species turnover across the groups, it is

recognised that species-rich groups can capture a greater range of environmental

conditions and habitat variation (Larsen et al. 2009; Larsen et al. 2012; Ikin et al.

2016). Therefore, woodland patches selected for species-rich surrogate groups in

systematic conservation planning scenarios can be expected to represent other taxa

more effectively (Larsen et al. 2009). Such a prediction was found to be further

corroborated with our complementary patch-selection analysis which revealed that

patch sets selected for groups like beetles were able to achieve high representation

of other taxa.

4.3. Representation of targets in patch sets selected for different taxa (Question 2, 4)

We found that representation targets of other taxa were generally unmet in the best

patch sets selected to meet targets of the surrogate taxon, beetles being an

exception. Patch sets selected to meet representation targets for the diverse beetle

assemblage were able to capture a high proportion of most other taxa. This validated

parts of our earlier analysis on cross-taxonomic congruency which revealed stronger

associations between beetles with other taxa (Figure 6). While this might imply that

beetle assemblages could be used as surrogates to prioritise remnant habitats in

agricultural landscapes for conservation, two practical issues immediately arise.

First, the use of beetle assemblages as surrogates to guide conservation planning

would demand significant investment into sampling, sorting and identification effort

given the high beetle diversity in many landscapes (New 2007). Second, sparing land

in farming landscapes to conserve beetles would be exceedingly costly as it demands

more sites and larger areas of woodland to capture the high compositional

heterogeneity of beetle assemblages. Conservation planning to target species rich

taxa could thus directly conflict with economic output of farming landscapes

prioritised for agricultural production.

150

After beetles, birds were the next best group in representing other taxa incidentally

(e.g. Sauberer et al. 2004). The best patch sets selected to meet bird targets showed

correspondingly higher representation of bees, beetles, trees and herpetofauna than

patch sets prioritised for representing these other taxa per se. This finding was

consistent with expectation since we earlier found bird diversity to be well

correlated with other taxa (e.g. wild bees), underscoring its relative effectiveness as a

cross-taxonomic surrogate. Additionally, birds are easy to identify and can be readily

surveyed in biodiversity inventories (Ikin et al. 2016). High species richness of birds

also indicate that sites prioritised to conserve them could represent a broader range

of environments, which would then incidentally “capture” a greater diversity of

other taxa (Ryti 1992; Lund & Rahbek 2002; Larsen et al. 2012).

4.4. Similarity in patch sets for different taxa (Question 3)

Although the best patch sets selected for each taxon were generally dissimilar to

other taxa, they became increasingly similar at high representation (> 40%) targets.

Increasing similarity of patch sets at high representation targets for each taxon arises

because a larger proportion (and number) of woodland patches with high species

richness of each taxon increasingly overlap in these patch sets. A clear conservation

implication is that for cross-taxonomic surrogates of biodiversity to be most

effective, representation targets set for the surrogate taxon of interest for

conservation prioritisation and planning will need to be reasonably high.

4.6. Conclusions

The use of surrogate taxa to prioritise landscapes for biodiversity conservation is a

well-established idea in conservation biology (Ryti 1992; Margules & Pressey 2000;

Larsen et al. 2009). While many studies of cross-taxonomic surrogates have been

151

completed at large spatial scales (e.g. Howard et al. 2006; Billeter et al. 2008), our

study provides one of few examples where cross-taxonomic surrogacy is down-scaled

into a complex agricultural landscape (c. 300km2) and based on simultaneous

sampling of five ecologically-important taxonomic groups. Our study offers three

key insights for biodiversity conservation in farming landscapes.

First, spatial turnover patterns in species composition tend to be taxon-specific,

being highest for species-rich taxa such as beetles. Less speciose groups containing

many good dispersers like bees tend to show lower spatial turnover. Therefore,

conservation planning in highly heterogenous agricultural landscapes based on a

limited set of surrogate taxa should not overlook taxon-specific turnover patterns

(e.g. Ferrier 2002; Si et al. 2015), especially if sparing land for biodiversity is the end

goal of the project. Conservation planning and prioritisation shaped by assessments

of species-poor and/or non-vagile groups like herpetofauna may result in a smaller

number and area of habitat patches being conserved, to the detriment of other taxa.

On the contrary, prioritising woodland conservation with taxa that are species-rich

and show high compositional turnover such as beetles could become costly as it will

necessitate the sparing of more land to meet conservation goals.

Second, sets of woodland remnant patches selected to prioritise conservation of

different taxa become increasingly similar at higher representation targets. This

indicates that conservation planning in landscapes based on low representation

targets of one or few surrogate taxa will be ineffective and highly uneven in

representing broad suites of other biota. While it ideal to spare as much remnant

natural woodland as possible for biodiversity conservation, trade-offs and

opportunity costs will need to be better defined to meet biodiversity conservation

targets in farming landscapes without overly compromising agricultural production.

Third, species-rich taxa like beetles and birds were able to achieve high incidental

representation of other taxa in systematic conservation planning scenarios. Species-

152

rich groups have been recognised as being more effective surrogates (e.g. Larsen et

al. 2012) and should therefore be the focus and yardsticks of initiatives to manage

and prioritise remnant woodland in farming and other heavily-disturbed landscapes

for biodiversity conservation. While birds are already well known surrogates in

agricultural landscapes (e.g. Part & Soderstrom 1999; Eglington et al. 2012; Ikin et al.

2016), biodiversity assessments based on bird data can be complemented with that

of highly diverse beetle assemblages to identify remnant habitat important to a

broad suite of species-rich groups, especially where different species-rich groups

converge.

Acknowledgments

We are grateful to the landowners who granted us access to their private land during

field work. DLY is supported by the Lesslie Foundation and an Australian National

University Postgraduate Scholarship. DBL is supported by an Australian Research

Council Laureate Fellowship. We also wish to thank Michael Batley and Michael

Schwarz for helping with the identification of bee specimens, and Mick Neave, John

Ascher for advice in sampling bees.

References cited

Albuquerque, F. S., Beier, P., 2016. Downscaling patterns of complementarity to a

finer resolution and its implications for conservation prioritization. Ecol. Evol.

6(12), 4032-4040.

Alexander, P., Rounsevell, M. D., Dislich, C., Dodson, J. R., Engström, K., Moran, D.,

2015. Drivers for global agricultural land use change: the nexus of diet,

population, yield and bioenergy. Glob. Environ. Chang. 35, 138-147.

153

Andelman, S. J., Fagan, W. F., 2000. Umbrellas and flagships: efficient conservation

surrogates or expensive mistakes? Proc. Natl. Acad. Sc. U.S.A. 97(11), 5954-5959.

Anderson, R. S., Ashe, J. S., 2000. Leaf litter inhabiting beetles as surrogates for

establishing priorities for conservation of selected tropical montane cloud forests

in Honduras, Central America (Coleoptera; Staphylinidae, Curculionidae).

Biodivers. Conserv. 9(5), 617-653.

Ardron, J. A., Possingham, H. P., Klein, C. J., eds. 2010. Marxan Good Practices

Handbook, Version 2. Pacific Marine Analysis and Research Association, Victoria,

British Columbia, Canada.

Ball, I. R., Possingham, H. P., Watts. M., 2009. Marxan and relatives: Software for

spatial conservation prioritisation. In Moilanen, A., K. A. Wilson, H. P.

Possingham. eds. Spatial conservation prioritisation: Quantitative methods and

computational tools. Pp. 185-195 in Oxford University Press, Oxford, UK.

Barton, P. S., Westgate, M. J., Lane, P. W., MacGregor, C., Lindenmayer, D. B., 2014.

Robustness of habitat‐based surrogates of animal diversity: a multitaxa

comparison over time. J. Appl. Ecol. 51(5), 1434-1443.

Barton, P. S., Pierson, J. C., Westgate, M. J., Lane, P. W., Lindenmayer, D. B., 2015.

Learning from clinical medicine to improve the use of surrogates in ecology.

Oikos 124(4), 391-398.

Baselga, A., Lobo, J. M., Svenning, J. C., Aragón, P., Araújo, M. B., 2012. Dispersal

ability modulates the strength of the latitudinal richness gradient in European

beetles. Glob. Ecol. Biogeogr. 21(11), 1106-1113.

Basset, Y., Kitching, R.L., 1991. Species number, species abundance and body length

of arboreal arthropods associated with a rainforest tree. Ecol. Entomol. 16, 391–

402.

Benton, T. G., Vickery, J. A., Wilson, J. D., 2003. Farmland biodiversity: is habitat

heterogeneity the key? Trends Ecol. Evol. 18(4), 182-188.

154

Billeter, R., Liira, J., Bailey, D., Bugter, R., Arens, P., Augenstein, I., Aviron, S.,

Baudry, J., Bukacek, R., Burel, F., Cerny, M., De Blust, G., De Cock, R., Diekotter,

T., Dietz, H., Dirksen, J., Dormann, C., Durka, W., Frenzel, M., Hamersky, R.,

Hendrickx, F., Herzog, F., Klotz, D., Koolstra, B., Lausch, A., Le Couer, D.,

Maelfait, J. P., Opdam, P., Roubalova, M., Schermann, A., Schermann, N.,

Schmidt, T., Schweiger, O., Smulders, M. J. M., Speelmans, M., Simova, P.,

Verboom, J., van Wingerden, W. K. R. E., Zobel, M., Edwards, P.J., 2008.

Indicators for biodiversity in agricultural landscapes: a pan‐European study. J.

Appl. Ecol. 45(1), 141-150.

Cadotte, M. W., 2006. Dispersal and species diversity: a meta-analysis. Am. Nat.

167(6), 913-924.

Cadotte, M. W., & Fukami, T., 2005. Dispersal, spatial scale, and species diversity in a

hierarchically structured experimental landscape. Ecol. Lett, 8(5), 548-557.

Caro, T., 2010. Conservation by proxy: indicator, umbrella, keystone, flagship, and

other surrogate species. Island Press, U.S.A.

Carrascal, L. M., Cayuela, L., Palomino, D., Seoane, J., 2012. What species-specific

traits make a bird a better surrogate of native species richness? A test with insular

avifauna. Biol. Conserv. 152, 204-211.

Cunningham, S. A., Attwood, S. J., Bawa, K. S., Benton, T. G., Broadhurst, L. M.,

Didham, R. K., Didham, R. K., McIntyre, S., Perfecto, I., Samways, M. J.,

Tscharntke, T., Vandermeer, J., Villard, M., Young, A. G., Lindenmayer, D. B.,

2013. To close the yield-gap while saving biodiversity will require multiple locally

relevant strategies. Agri. Ecosys. Environ. 173, 20-27.

Di Minin, E., Moilanen, A., 2014. Improving the surrogacy effectiveness of

charismatic megafauna with well‐surveyed taxonomic groups and habitat types. J.

Appl. Ecol. 51(2), 281-288.

155

Donald, P. F., Sanderson, F. J., Burfield, I. J., Van Bommel, F. P., 2006. Further

evidence of continent-wide impacts of agricultural intensification on European

farmland birds, 1990–2000. Agri. Ecosys. Environ. 116(3), 189-196.

Duan, M., Liu, Y., Yu, Z., Baudry, J., Li, L., Wang, C., Axmacher, J. C., 2016.

Disentangling effects of abiotic factors and biotic interactions on cross-taxon

congruence in species turnover patterns of plants, moths and beetles. Sci. Rep. 6,

23411.

Eglington, S. M., Noble, D. G., Fuller, R. J., 2012. A meta-analysis of spatial

relationships in species richness across taxa: birds as indicators of wider

biodiversity in temperate regions. J. Nat. Conserv. 20(5), 301-309.

Ekroos, J., Heliölä, J., Kuussaari, M., 2010. Homogenization of lepidopteran

communities in intensively cultivated agricultural landscapes. J. Appl. Ecol. 47(2),

459-467.

FAOSTAT. 2014. Food and agriculture data. Food and Agricultural Organisation of

the United Nations. URL: http://www.fao.org/faostat/

Favreau, J. M., Drew, C. A., Hess, G. R., Rubino, M. J., Koch, F. H., Eschelbach, K. A.,

2006. Recommendations for assessing the effectiveness of surrogate species

approaches. Biodivers. Conserv. 15(12), 3949-3969.

Ferrier, S., 2002. Mapping spatial pattern in biodiversity for regional conservation

planning: where to from here? Syst. Biol. 51(2), 331-363.

Fischer, J., Brosi, B., Daily, G. C., Ehrlich, P. R., Goldman, R., Goldstein, J.,

Lindenmayer, D. B., Manning, A. D., Mooney, H. A., Pejchar, L., Ranganathan, J.,

Tallis, H., 2008. Should agricultural policies encourage land sparing or wildlife‐

friendly farming? Front. Ecol. Environ. 6(7), 380-385.

Foley, J. A., Ramankutty, N., Brauman, K. A., Cassidy, E. S., Gerber, J. S., Johnston,

M., Mueller, N. D., O’Connell, C., Ray, D. K., West, P. C., Balzer, C., Bennett, E.

M., Carpenter, S. R., Hill, J., Monfreda, C., Polasky, S., Rockstrom, J., Sheehan, J.,

156

Siebert, S., Tilman, D., Zaks, D. P. M., 2011. Solutions for a cultivated planet.

Nature 478(7369), 337-342.

Game, E. T., Grantham, H. S., 2008. Marxan user manual: for Marxan version 1.8.10.

St. Lucia, Queensland, Australia and Vancouver, British Columbia, Canada:

University of Queensland and Pacific Marine Analysis and Research Association.

Gardner, T. A., Barlow, J., Araujo, I. S., Ávila‐Pires, T. C., Bonaldo, A. B., Costa, J. E.,

Esposito, M. C., Ferreira, L. V., Hawes, J., Hernandez, M. I. M., Hoogmoed, M. S.,

Leite, R. N., Lo, N. F., Malcolm, J. R., Martins, M. B., Mestre, L. A. M., Miranda-

Santos, R., Overal, W. L., Parry, L., Peters, S. L., Ribeiro, M. A., Da Silva, M. N. F.,

Motta, C. D., Peres, C. A., 2008. The cost‐effectiveness of biodiversity surveys in

tropical forests. Ecol. Lett. 11(2), 139-150.

Gaston, K. J., Blackburn, T. M., Goldewijk, K. K., 2003. Habitat conversion and global

avian biodiversity loss. Proc. R. Soc. Lond. B: Biol. Sci. 270(1521), 1293-1300.

Goslee, S.C., Urban, D.L., 2007. The ecodist package for dissimilarity-based analysis

of ecological data. J. Stat. Soft. 22(7):1-19.

Grantham, H. S., Pressey, R. L., Wells, J. A., Beattie, A. J., 2010. Effectiveness of

biodiversity surrogates for conservation planning: different measures of

effectiveness generate a kaleidoscope of variation. PLoS One, 5(7), e11430.

Guisan, A., Tingley, R., Baumgartner, J. B., Naujokaitis‐Lewis, I., Sutcliffe, P. R.,

Tulloch, A. I., Regan, T. J., Brotons, L., McDonald-Madden, E., Mantyka-Pringle,

C., Martin, T. G., Rhodes, J. R., Maggini, R., Setterfield, S. A., Elith, J. A., Schwartz,

M. W., Wintle, B., Broennimann, O., Austin, M., Ferrier, S., Kearney, M. R.,

Possingham, H. P., Buckley, Y. M., 2013. Predicting species distributions for

conservation decisions. Ecol. Lett. 16(12), 1424-1435.

Hangay, G., Zborowski, P., 2010. A guide to the beetles of Australia. CSIRO

Publishing, Australia

Heino, J., 2010. Are indicator groups and cross-taxon congruence useful for

predicting biodiversity in aquatic ecosystems? Ecol. Indic. 10(2), 112-117.

157

Hess, G. R., Bartel, R. A., Leidner, A. K., Rosenfeld, K. M., Rubino, M. J., Snider, S. B.,

& Ricketts, T. H., 2006. Effectiveness of biodiversity indicators varies with extent,

grain, and region. Biol. Conserv. 132(4), 448-457.

Hunter, M., Westgate, M., Barton, P., Calhoun, A., Pierson, J., Tulloch, A., Beger, M.,

Branquinho, C., Caro, T., Gross, J., Heino, J., Lane, P., Longo, C., McDowell, W.

H., Mellin, C., Salo, H., Lindenmayer, D., 2016. Two roles for ecological surrogacy:

Indicator surrogates and management surrogates. Ecol. Indic. 63, 121-125.

Ikin, K., Yong, D. L., Lindenmayer, D. B., 2016. Effectiveness of woodland birds as

taxonomic surrogates in conservation planning for biodiversity on farms. Biol.

Conserv. 204, 411-416.

Janssen, P., Cateau, E., Fuhr, M., Nusillard, B., Brustel, H., Bouget, C. 2016. Are

biodiversity patterns of saproxylic beetles shaped by habitat limitation or

dispersal limitation? A case study in unfragmented montane forests. Biodivers.

Conserv. 25(6), 1167-1185.

Joshi, N. K., Leslie, T., Rajotte, E. G., Kammerer, M. A., Otieno, M., Biddinger, D. J.,

2015. Comparative trapping efficiency to characterize bee abundance, diversity,

and community composition in apple orchards. Ann. Entomol. Soc. Am. 108(5),

785-799.

Justus, J., Sarkar, S. 2002. The principle of complementarity in the design of reserve

networks to conserve biodiversity: a preliminary history. J. Biosci. 27(4), 421-435.

Kay, G. M., Barton, P. S., Driscoll, D. A., Cunningham, S. A., Blanchard, W.,

McIntyre, S., Lindenmayer, D. B. 2016. Incorporating regional‐scale ecological

knowledge to improve the effectiveness of large‐scale conservation programmes.

Anim. Conserv. 19(6), 515-525.

Landeiro, V. L., Bini, L. M., Costa, F. R., Franklin, E., Nogueira, A., de Souza, J. L.,

Moraes, J., Magnusson, W. E., 2012. How far can we go in simplifying

biomonitoring assessments? An integrated analysis of taxonomic surrogacy,

158

taxonomic sufficiency and numerical resolution in a megadiverse region. Ecol.

Indic. 23, 366-373.

Larsen, F.W., Bladt, J. Rahbek, C., 2009. Indicator taxa revisited: useful for

conservation planning? Divers. Distrib. 15, 70–79.

Larsen, F. W., Bladt, J., Balmford, A., Rahbek, C., 2012. Birds as biodiversity

surrogates: will supplementing birds with other taxa improve effectiveness? J.

Appl. Ecol. 49(2), 349-356.

Lentini, P. E., Martin, T. G., Gibbons, P., Fischer, J., Cunningham, S. A., 2012.

Supporting wild pollinators in a temperate agricultural landscape: maintaining

mosaics of natural features and production. Biol. Conserv. 149(1), 84-92.

Lindenmayer, D. B., Cunningham, R. B., MacGregor, C., Crane, M., Michael, D.,

Fischer, J., Montague-Drake, R., Felton, A., Manning, A., 2008. Temporal changes

in vertebrates during landscape transformation: a large‐scale “natural

experiment”. Ecol. Mono. 78(4), 567-590.

Lindenmayer, D. B., Barton, P. S., Lane, P. W., Westgate, M. J., McBurney, L., Blair,

D., Gibbons, P., Likens, G. E., 2014. An empirical assessment and comparison of

species-based and habitat-based surrogates: a case study of forest vertebrates and

large old trees. PloS One, 9(2), e89807.

Lindenmayer, D. B., Barton, P., Pierson, J., eds. 2015. Indicators and Surrogates of

Biodiversity and Environmental Change. Csiro Publishing, Australia.

Lovell, S., Hamer, M., Slotow, R., Herbert, D., 2007. Assessment of congruency

across invertebrate taxa and taxonomic levels to identify potential surrogates.

Biol. Conserv. 139(1), 113-125.

Lund, M. P., Rahbek, C., 2002. Cross‐taxon congruence in complementarity and

conservation of temperate biodiversity. Anim. Conserv. 5(2), 163-171.

Magurran, A. E., 2004. Measuring Biological Diversity. Blackwell, Oxford, U.K.

159

Margules C. R., Pressey, R. L., 2000. Systematic conservation planning. Nature

405:243-253.

Matthews, E.G., 1992. A guide to the beetles of South Australia. South Australian

Museum, Adelaide, Australia.

Michener, C.D., 2000. The Bees of the World. The John Hopkins University Press,

Baltimore, U.S.A. & London, U.K.

New, T.R., eds. 2007. Beetle conservation. Springer, Dortrecht, The Netherlands

Norris, K. 2008. Agriculture and biodiversity conservation: opportunity knocks.

Conserv. Lett. 1(1), 2-11.

Paavola, R., Muotka, T., Virtanen, R., Heino, J., Jackson, D., Mäki-Petäys, A., 2006.

Spatial scale affects community concordance among fishes, benthic

macroinvertebrates, and bryophytes in streams. Ecol. Appl. 16(1), 368-379.

PADiL., 2016. Pest and disease image library. URL:

http://www.padil.gov.au/pollinators/search?queryType=all

Pärt, T., Söderström, B., 1999. Conservation value of semi‐natural pastures in

Sweden: contrasting botanical and avian measures. Conserv. Biol. 13(4), 755-765.

Phalan, B., Onial, M., Balmford, A., Green, R. E., 2011. Reconciling food production

and biodiversity conservation: land sharing and land sparing compared. Science

333(6047), 1289-1291.

Qian, H., & Ricklefs, R. E., 2012. Disentangling the effects of geographic distance and

environmental dissimilarity on global patterns of species turnover. Glob. Ecol.

Biogeogr. 21(3), 341-351.

R Development Core Team., 2013. R: A language and environment for statistical

computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-

900051-07-0, URL http://www.R-project.org/.

Rodrigues, A. S., Brooks, T. M., 2007. Shortcuts for biodiversity conservation

planning: the effectiveness of surrogates. Annu. Rev. Ecol. Evol. Syst., 38, 713-737.

160

Rooney, R. C., Azeria, E. T., 2015. The strength of cross‐taxon congruence in species

composition varies with the size of regional species pools and the intensity of

human disturbance. J Biogeog. 42(3), 439-451.

Royston, J. P., 1983. Some techniques for assessing multivarate normality based on

the Shapiro-Wilk W. Appl. Stat. 32, 121-133.

Ryti, R. T., 1992. Effect of the focal taxon on the selection of nature reserves. Ecol.

Appl. 2(4), 404-410.

Sætersdal, M., Gjerde, I., Blom, H. H., Ihlen, P. G., Myrseth, E. W., Pommeresche, R.,

John Skartveit, J., Solhøy, T., Aas, O., 2004. Vascular plants as a surrogate species

group in complementary site selection for bryophytes, macrolichens, spiders,

carabids, staphylinids, snails, and wood living polypore fungi in a northern forest.

Biol. Conserv. 115(1), 21-31.

Sauberer, N., Zulka, K. P., Abensperg-Traun, M., Berg, H. M., Bieringer, G.,

Milasowszky, N., Moser, D., Plutzar, C., Pollheimer, M., Tröstl, R., Zechmeister,

H., Grabherr, G., 2004. Surrogate taxa for biodiversity in agricultural landscapes

of eastern Austria. Biol. Conserv. 117(2), 181-190.

Si, X., Baselga, A., Ding, P., 2015. Revealing beta-diversity patterns of breeding bird

and lizard communities on inundated land-bridge islands by separating the

turnover and nestedness components. PloS One, 10(5), e0127692.

Similä, M., Kouki, J., Mönkkönen, M., Sippola, A. L., Huhta, E., 2006. Co-variation

and indicators of species diversity: Can richness of forest-dwelling species be

predicted in northern boreal forests? Ecol. Indic. 6(4), 686-700.

Soininen, J., 2010. Species turnover along abiotic and biotic gradients: patterns in

space equal patterns in time? BioScience 60(6), 433-439.

Soininen, J., Lennon, J. J., Hillebrand, H., 2007. A multivariate analysis of beta

diversity across organisms and environments. Ecology 88(11), 2830-2838.

161

Su, J. C., Debinski, D. M., Jakubauskas, M. E., Kindscher, K., 2004. Beyond species

richness: Community similarity as a measure of cross‐taxon congruence for

coarse‐filter conservation. Conserv. Biol. 18(1), 167-173.

Tscharntke, T., Klein, A. M., Kruess, A., Steffan‐Dewenter, I., Thies, C., 2005.

Landscape perspectives on agricultural intensification and biodiversity–ecosystem

service management. Ecol. Lett. 8(8), 857-874.

Tulloch, A. I., Chadès, I., Dujardin, Y., Westgate, M. J., Lane, P. W., Lindenmayer, D.

B., 2016. Dynamic species co–occurrence networks require dynamic biodiversity

surrogates. Ecography 39, 1185-1196.

Turtureanu, P. D., Palpurina, S., Becker, T., Dolnik, C., Ruprecht, E., Sutcliffe, L. M.,

Szabo, A., Dengler, J., 2014. Scale-and taxon-dependent biodiversity patterns of

dry grassland vegetation in Transylvania. Agri. Ecosys. Environ. 182, 15-24.

Vane-Wright, R. I., Humphries, C. J., Williams, P. H. 1991. What to protect?—

Systematics and the agony of choice. Biol. Conserv. 55(3), 235-254.

Walker, K.L., 1995. Revision of the Australian native bee subgenus Lasioglossum

(Chilalictus) (Hymenoptera: Halictidae). Mem. Mus. Vict. 55(1): 1-423

Westgate, M. J., Barton, P. S., Lane, P. W., Lindenmayer, D. B., 2014. Global meta-

analysis reveals low consistency of biodiversity congruence relationships. Nat.

Comm. 5, 3899.

Wirsenius, S., Azar, C., Berndes, G., 2010. How much land is needed for global food

production under scenarios of dietary changes and livestock productivity

increases in 2030? Agri. Sys. 103(9), 621-63.

Wolters, V., Bengtsson, J., Zaitsev, A. S., 2006. Relationship among the species

richness of different taxa. Ecology, 87(8), 1886-1895.

Yong, D. L., Barton, P. S., Okada, S., Crane, M., Lindenmayer, D. B., 2016. Birds as

surrogates for mammals and reptiles: Are patterns of cross-taxonomic

162

associations stable over time in a human-modified landscape? Ecol. Indic. 69, 152-

164.

Figure 1. Conceptual framework showing our analytical approach in relation to the

four research questions asked.

163

Figure 2. Map of the Nanangroe landscape showing the location and distribution of

our 42 study sites in relation to the Murrumbidgee River. Grey-shaded areas are

monoculture plantations of Pinus radiata while unshaded areas are open grazing

land.

164

Figure 3. Smoothed species accumulation curves for each taxon in the Nanangroe

landscape showing observed species richness and sampling effort for each taxon

relative to other taxa.

165

Figure 4. Bar plots (with error bars representing standard deviation) showing mean

species richness at the patch-level as a percentage of the total species pool in the

landscape (gamma-diversity).

166

Figure 5 (a). Network diagram showing the degree of correlation in species richness between four taxa based on the Spearman’s

rho. Significant correlations are presented as black arrows while non-significant correlations are presented as grey arrows. (b).

Network diagram showing the degree of correlation in species composition between five taxa conditioned on geographic space,

based on the partial Mantel R. Significant correlations are presented as black arrows while non-significant correlations are

presented as grey arrows.

167

Figure 6. Accumulation plots showing incidental representation of each taxon

(indicated by taxon name on y-axis): (a) bird, (b) herpetofauna, (c) bee, (d) tree, (e)

all beetle, (f) beetle excluding species represented by singletons) in the best patch

sets selected for the surrogate taxa (in inset legend). Enlarged points in dark grey

indicate that the representational target for that taxon has been incidentally met.

168

Figure 7. Cluster dendrogram showing best patch sets selected for every taxa at four

target threshold (20%, 40%, 60%, 80%) based on the complete linkage method. The

taxon and its corresponding representation target is shown at each node of the

dendrogram

169

Figure 8. Plot showing relative cost of retaining remnant woodland patches (using

total area of patches as a proxy for cost) to conserve each taxon based on

representational targets ranging from 10-80%. Relative to other taxa, retaining

woodland in the landscape to prioritise beetle conservation was the most costly.

170

Table 1. Glossary of key terms in text and their definitions

Term Definition

Biodiversity surrogate A defined taxonomic group (e.g. birds) or group of

species whose occurrence or diversity predicts that

of another, usually less well-known group of species.

Cross-taxonomic

congruency

Degree of association or co-variation in the diversity

pattern of a defined group of species with respect to

another group. Common metrics include measures

of correlation strength such as Spearman’s ρ and

Pearson’s r.

Compositional turnover Variation in the composition of species assemblages

across space; an approach to quantifying beta

diversity in a landscape.

Conservation feature A unit to be represented in a solution of reserve sites

in systematic conservation planning scenarios.

Usually quantified as the absence of a species in a

defined site.

Incidental representation Representation of a species or taxa in a set of

identified sites/reserves that was not targeted a

priori, in the context of systematic conservation

planning scenarios.

Representation target Defined numerical thresholds in the representation

of selected conservation features (e.g. occurrence

and distribution of a surrogate taxa) in a systematic

conservation planning context.

Complementarity A principle in designing networks of reserve sites in

conservation whereby the selection of sites

iteratively adds sites that complement those already

selected (Vane-Wright et al. 1991; Justus & Sarkar

2002).

Simulated annealing An algorithm implemented in Marxan to identify

near-optimal solutions in selecting networks of

reserve sites in conservation (Game & Grantham

2008).

171

Supplementary tables

Table S1. Non-parametric estimators of species richness for four taxa and proportion

of each taxa detected based on the mean of four estimators.

Species

richness

Taxa sampled

Bird Wild bee Herpetofauna Ground-

dwelling Beetle

Tree

Observed 77 32 21 258 9

Chao1 99.05 58.00 51.25 331.30 9.00

Jack1 97.51 42.76 31.74 348.68 9.00

Jack2 108.23 50.41 40.37 382.41 -

Bootstrap 86.39 37.17 25.27 300.66 9.32

Mean 97.80 47.09 37.15 340.76 9.12

Proportion (%) 78.73 67.95 56.52 75.71 98.69

172

Table S2. Species richness of each taxa at the site- and landscape-level.

Species

richness

Taxa

Bird Herpetofauna Wild bee Beetle

(all)

Beetle

(excl.

singletons)

Tree

Mean (site) 13.07 ±

3.46

2.90 ±

1.64

6.27 ±

2.91

25.85 ±

9.72

23.63 ±

8.66

2.02 ±

1.23

Observed

77 21 31 259 176 9

Proportion

of total per

site (%)

16.98 13.82 20.22 9.98 13.43 22.49

173

Table S3. List of reptiles and amphibians sampled in the Nanangroe landscape.

Common Name Latin name Number of

occurrences

Reptiles

Eastern Long-necked Turtle Chelodina longicollis 1

Marbled Gecko Christinus marmoratus 14

Stone Gecko Diplodactylus vittatus 2

Olive Legless Lizard Delma inornata 1

Red-throated Skink Acritoscincus platynotum 1

Four-fingered Skink Carlia tetradactyla 16

Straight-browed Skink Ctenotus spaldingi 7

Copper-tailed Skink Ctenotus taeniolatus 6

Cunningham's skink Egernia cunninghami 1

Tree Skink Egernia striolata 7

Southern Water Skink Eulamprus heatwolei 1

Eastern Three-toed Earless Skink Hemiergis talbingoensis 33

Delicate Skink Lampropholis delicata 1

Boulenger's Skink Morethia boulengeri 17

Eastern Bearded Dragon Pogona barbata 2

Dwyer's Snake Parasuta dwyeri 1

Woodland Blind Snake Ramphotyphlops proximus 1

Amphibians

Common Eastern Froglet Crinia signifera 4

Spotted Marsh Frog Limnodynastes tasmaniensis 1

Booroolong Frog Litoria booroolongensis 1

Broad-palmed Frog Litoria latopalmata 1

174

Table S4. List of dominant trees sampled in the Nanangroe landscape.

Common Name Latin name Number of

occurrences

Family Myrtaceae

Apple Box Eucalyptus bridgesiana 2

Blakely’s Red Gum Eucalyptus blakelyi 17

Long-leaved Box Eucalyptus goniocalyx 3

Red Box Eucalyptus polyanthemos 3

Red Stringybark Eucalyptus macrorhyncha 6

White Box Eucalyptus albens 13

Yellow Box Eucalyptus melliodora 14

Family Malvaceae

Kurrajong Brachychiton populneus 8

Family Casuarinaceae

River She-oak Casuarina cunninghamiana 2

175

IV. Yong, D. L., Barton, P. S., Cunningham, S. A. & Lindenmayer, D. B.

(Submitted). Effects of anthropogenic disturbance on cross-taxonomic

congruence patterns in terrestrial ecosystems. Diversity and Distributions.

176

Effects of anthropogenic disturbance on cross-taxonomic

congruence patterns in terrestrial ecosystems

Abstract

Aim Assessments of how biodiversity responds to environmental change often

employ surrogate-based approaches underpinned by patterns of cross-taxonomic

congruence. While it is well known that anthropogenic disturbance can shift biotic

assemblages in different ways, little is understood about how disturbance gradients

can promote or diminish cross-taxonomic congruence.

Methods We reviewed the global literature on studies of cross-taxonomic

surrogates in terrestrial ecosystems. Using meta-regression models, we then

quantified the effect of (1) anthropogenic disturbance, (2) spatial scale and, and (3)

duration of study on the strength of cross-taxonomic congruence. We focused on

studies that examined cross-taxonomic congruence using correlations of both

species richness and composition between pairs of taxa.

Results We identified 146 studies which yielded 1,633 measures of correlations of

species richness and 1,030 measures of correlations of species composition. We

found that cross-taxonomic congruence is highest in landscapes that were heavily

disturbed (e.g. cropping, urban) and lowest in fairly intact landscapes. However, we

also found that various interactions between disturbance, spatial scale, or study

duration, best explained the variation in effect sizes. Spatial scale alone was a

relatively weak predictor of cross-taxonomic congruence, becoming important only

at large (>continental) scales.

Main conclusions We showed that the effects of anthropogenic disturbance play an

important, yet overlooked role in shaping patterns of cross-taxonomic congruence.

Shifts in species assemblages resulting from altered environmental gradients, biotic

homogenisation or invasion by generalists may promote stronger cross-taxonomic

177

associations in heavily disturbed landscapes. Identification of cross-taxonomic

congruence patterns in disturbed landscapes may not be transferrable to intact

landscapes, potentially limiting the use of taxonomic surrogates in biodiversity

monitoring and assessment across regions with diverse land uses, especially in

agricultural landscapes.

Introduction

Rapid modification of landscapes resulting from agricultural expansion and

urbanisation has led to significant changes to biodiversity and associated ecosystem

services worldwide (Dobson et al. 2006; Koh & Wilcove 2008; Butchart et al. 2010).

In the face of the global biodiversity crisis (Maxwell et al. 2016), the limited

resources available to conservation practitioners and incomplete knowledge on

many taxonomic groups (Schuldt et al. 2009; McMullan-Fisher et al. 2010) poses

formidable impediments to acquiring biodiversity data efficiently to guide

conservation planning and management (Possingham et al. 2007; Bottrill et al. 2008;

Landeiro et al. 2012). Over time, these challenges have fostered the development of

numerous surrogate approaches for biodiversity assessments and monitoring (Heino

& Soininen 2007; Caro 2010; Lindenmayer et al. 2015),

A widely examined approach to surrogacy in biodiversity conservation is the concept

of cross-taxonomic surrogacy (Schuldt et al. 2009; Caro 2010; Heino 2010; Qian &

Kissling 2010; Lindenmayer et al. 2015). Cross-taxonomic surrogates are underpinned

by the hypothesis that information on one taxonomic group (i.e. the surrogate) co-

varies with, and thus predicts, the occurrence or diversity of one or more other taxa

(i.e. the target) (Sætersdal & Gjerde 2011). The congruence in diversity patterns

between the surrogate and other target groups may arise due to the similar

responses to environmental gradients (Saetersdal et al. 2004; Barlow et al. 2007;

178

Heino 2010), or the structuring effects on species assemblages resulting from biotic

interactions between different groups (Larsen et al. 2012). Vertebrates, in particular,

have been widely proposed as cross-taxonomic surrogates for various component of

biodiversity because of the wide availability of data on well-studied groups such as

birds and mammals (e.g. Larsen et al. 2012; Higa et al. 2016).

Despite its appeal in biodiversity conservation and monitoring, the utility of cross-

taxonomic surrogates is not without problems. A major stumbling block is the often

conflicting outcomes reported by studies of cross-taxonomic surrogates (e.g.

Cushman et al. 2010; Fattorini et al. 2012). Taxa found to be good surrogates in some

cases may perform poorly in others (e.g. Oliver et al. 1998; Eglington et al. 2012).

Furthermore, the patterns of congruence between different taxonomic groups may

be inconsistent across spatial scales and landscape contexts (Hess et al. 2006; Ekroos

et al. 2013) or lack temporal validation (Thomson et al. 2005). From a conceptual

perspective, there remains an absence of a robust, theory-based framework on which

to base the study of surrogates (Lindenmayer & Likens 2011; Sætersdal & Gjerde 2011).

Put together, these shortcomings underscore the need to better clarify the scales and

ecological contexts that influence cross-taxonomic surrogacy, so as to better guide

their application (Gaspar et al. 2010; Barton et al. 2015; Lindenmayer et al. 2015).

One important aspect of cross-taxonomic surrogacy seldom studied is how

anthropogenic disturbance can influence congruence patterns between different

taxa (Wolters et al. 2006; de Andrade et al. 2014; Rooney & Azeria 2015). Disturbance

arising from different human land uses can disrupt ecological processes, altering

species communities and ecosystems over time (e.g. Turner 2010; Banks et al. 2013;

Tonkin et al. 2016). The limited evidence on this topic to date is mixed. For example,

some studies suggest that disturbance may diminish cross-taxonomic congruence

(Anand et al. 2005; Rooney & Bayley 2012), whereas others suggest disturbance may

in fact promote cross-taxonomic congruence (de Andrade et al. 2014; Rooney &

179

Azeria 2015). Given that the world’s landscapes will continue to be modified with

increasing agricultural land use (Koh & Wilcove 2008; FAO 2017), a better

understanding of how disturbance can impact cross-taxonomic surrogacy is

necessary to inform initiatives to conserve biodiversity in farming and other human-

transformed landscapes (e.g. Fischer et al. 2006; Koh & Gardner 2010).

In this study, we systematically reviewed the global literature on cross-taxonomic

surrogates in terrestrial ecosystems. We addressed three questions on the

effectiveness of cross-taxonomic surrogates using a meta-analytical framework

(Koricheva et al. 2013; Nakagawa et al. 2017). First, we asked: (1) How do different

levels of anthropogenic disturbance affect patterns of cross-taxonomic congruence

at two commonly used measures of diversity (species richness, species

composition)? This question has never previously been addressed using a

comprehensive meta-analysis approach. We then asked: (2) How does the spatial

extent of a study influence cross-taxonomic congruency patterns? As disturbance

regimes often lead to altered biotic assemblages (Hobbs & Huenneke 1992; Collins

2000; Miller et al. 2011), we predicted that highly disturbed landscapes such as

grazing or cropping land at small spatial scales will be characterised by lower and

more variable cross-taxonomic congruence patterns than less disturbed landscapes

(e.g. old-growth forests). At the same time, we predicted cross-taxonomic patterns

to vary with spatial scale, increasing with higher spatial extent and resolution (Qian

& Kissling 2010; Westgate et al. 2014). We then asked: (3) How does the length of a

study affect the performance of cross-taxonomic surrogates? Since many studies of

species surrogates tend to be based on short-term data (Thomson et al. 2005;

Favreau et al. 2006), we predicted that cross-taxonomic patterns based on longer-

term datasets may be less consistent due to temporal variation in species

assemblages (e.g. Hewitt et al. 2016; Yong et al. 2016). Such changes may be more

pronounced in disturbed landscapes or regions where successional effects play a

large role in driving shifts in species composition and richness over time (e.g.

Swanson et al. 2010; Sousa-Souto et al. 2016). Drawing on our findings, we discuss

180

the potential ecological processes driving variation in cross-taxonomic congruence

patterns and their implications on the broader use of cross-taxonomic surrogates in

biodiversity conservation.

Methods

Literature search

We systematically searched the ISI Web of Science and Scopus databases for studies

relevant to cross-taxonomic surrogates of biodiversity published up to 31 December

2016. First, we conducted a ‘title’ and ‘topic’ search using five different key word

pairings, including: surrogate* AND species*, surrogate* AND ecology*,

biodiversity* AND surrogate*, cross-taxon* AND surrogate*, and biodiversity* AND

indicator*. We then evaluated the title and abstract of each article to check for its

relevance, and included only studies that compared different species groups within a

defined (e.g. order, family, tribe) but not necessarily equivalent taxonomic hierarchy

(e.g. family versus order) in assessing cross-taxonomic relationships. Studies

examining groups with a taxonomic hierarchy above the species level (e.g. number

of genera or family; i.e. higher-taxonomic surrogates) were excluded, as were species

groups defined by shared functional attributes (e.g. “herbivorous insects”,

“waterbirds”). Studies conducted solely in open aquatic (e.g. rivers) and marine

systems were omitted from our analyses given their very different ecological

dynamics from terrestrial systems (Elser et al. 2000).

Even though a variety of statistical approaches have been used to quantify cross-

taxonomic surrogacy (e.g. species-area index (SAI), spatial representation) (e.g.

Rodrigues & Brooks 2007; Westgate et al. 2014; Albuquerque & Beier 2016), we

focused on studies that applied correlations of species richness and species

181

composition because these formed the largest group of studies of cross-taxonomic

surrogacy. Thus, from an initial set of 1,116 articles on biodiversity surrogates, we

were able to narrow our review to 146 studies (see Appendix II of thesis for full list of

papers).

Data extraction

Studies that met our search criteria were partitioned into two subsets based on two

different approaches for quantifying cross-taxonomic congruency (i.e. species

richness correlations versus species composition correlations) (Sauberer et al. 2004;

Westgate et al. 2014; Rooney & Azeria 2015). We extracted information on: (1) the

measure of effect size (i.e. the type of correlation), (2) the value of the effect size,

and (3) the sample size N (i.e. number of sites). In cases where the effect sizes were

not reported, we checked for data in the paper’s supplementary material sections

and calculated the Pearson’s r. Because studies reported a diversity of effect size

measures (e.g. Spearman’s ρ, Kendall’s τ, linear model R2) besides the Pearson’s r, we

first transformed these different measures into the Pearson’s r following Lajeunesse

(2013). We then performed the Fisher’s z transformation to rescale the effect sizes,

defined as:

𝑧 =1

2𝐼𝑛 (

1 + 𝑟

1 − 𝑟)

Where In is the natural logarithmic function and r the sample effect size. Fisher’s z

follows a normal distribution and has the standard error σ:

σ = 1/√𝑁 − 3

where N is the sample size used in calculating the effect size.

We used the ‘escalc’ function available in the metafor package to perform the

Fisher’s z transformation (Viechtbauer 2010). To assess for publication bias in

reporting of effect sizes, we plotted funnel plots (See Supplementary Figure S2) of

182

effect sizes against its standard deviation to visually assess for asymmetrical

clustering (Sterne et al. 2011). We then proceeded to gather information on six

attributes of each study by checking the ‘methods’ section of each paper. These

attributes were: (1) spatial extent of the sampling sites (in km2), (2) mean latitude of

the study landscape, (3) the study duration (in years), (4) the dominant vegetation

type(s) in the study landscape, (5) level(s) of disturbance (see below for definition),

and (6) the pairs of taxonomic groups compared.

Classification of study spatial scale

We classified studies into five categories based on the geographical extent (spread)

(sensu Whittaker et al. 2001; Westgate et al. 2014) of all their study sites as reported

in each paper. Studies with their sampling sites distributed within an area of <1km2

to 10km2 were considered as ‘local-scale’. Studies with a spatial extent spanning 10-

1,000km2 were considered as ‘landscape-scale’ studies. Studies exceeding 1,000km2,

but less than 1,000,000 km2 were considered as ‘regional-scale’ studies. Any study

that exceeded 1,000,000 km2 in its spatial extent was considered a ‘continental-scale’

study. Studies at the global extent are defined as studies based on data from

locations spanning more than one continent.

Classification of anthropogenic disturbance levels

We classified terrestrial ecosystems into three broad classes based on the level of

anthropogenic disturbances reported in their studies. Only studies that were

conducted at the ‘local’, ‘landscape’ and ‘regional’ scale were classified because

studies at the largest spatial extents invariably incorporate composite landscapes

subjected to mixed and different degrees of disturbance. Our first classification of

‘undisturbed landscape’ was defined if anthropogenic disturbance was not reported,

183

and a study overlapped with a protected area (e.g. national parks) with no indication

of anthropogenic disturbance (e.g. grazing, vegetation clearance). We acknowledged

that such an approach is pragmatic and may overlook protected areas that were

formerly disturbed. Our second class was broadly defined a ‘disturbed landscape’,

and was one that remained predominantly covered with natural vegetation, yet had

been subjected to some disturbance by low intensity agricultural activities (e.g.

grazing) or other kinds of degradation (e.g. fires). This also included study sites with

secondary growth, logged forests and grazed woodland. Our third classification was

‘heavily disturbed landscapes’, and included study regions that incorporate land uses

with high levels of human modification and low natural vegetation cover. This

included cropping and plantation monoculture landscapes, as well as urban

environments such as parkland and remnant woodland patches in cities.

Quantifying the duration of a study

Where possible, we compiled information on the duration of data collection for each

study. As many studies do not provide specific information on the sampling

duration, we used the number of years of data collection as a proxy of sampling

duration. Thus, studies with one or more sampling periods distributed within a year

were considered as one-year studies. For studies with field data collection stretching

multiple years, we calculated the difference between the first and last year of data

collection.

Data preparation and statistical analyses

We analysed our dataset using a two-step approach to assess the effects of

disturbance, spatial scale and study duration on cross-taxonomic congruence

patterns. To estimate the average effect sizes for each category of disturbance and

184

spatial scale level after partitioning our dataset into subgroups for each level, we

used random-effect models (Nakagawa et al. 2017). We then fitted a series of

candidate mixed-effects models to assess the significance of each of the three study

attributes in predicting cross-taxonomic congruence, before carrying out further

post-hoc analyses to assess the significance of pairwise comparisons of levels

(Viechtbauer 2010).

First, we sub-grouped our dataset into each level. We then fitted a set of random-

effects models to estimate the average effect size and 95% confidence intervals at

each level of disturbance and spatial scale. We fitted separate models for the two

measures of effect sizes (correlations of species richness and species composition).

In the random-effects model, the effect sizes from each study are weighted by the

inverse of the within-study variance (σ2) and between-study variance derived from

the effect sample size n (Borenstein et al. 2010). This allocates more weight to effect

sizes based on larger pools of study sites. For each model, we performed the H-test

of heterogeneity to assess within-study heterogeneity in effect sizes. This step in the

analyses was completed using the ‘rma’ function implemented on the metafor

package (Viechtbauer 2010).

Second, we fitted a series of meta-regression (mixed-effects) models to assess how

well each of three predictors (disturbance level, spatial scale, study duration)

explained cross-taxonomic congruence for the two measures of effect sizes

(correlations of species richness, species composition). Categorical predictors

(disturbance level, spatial scale) were coded into factors with multiple levels while

‘duration of study’ (in years) was retained as a continuous variable. In each model,

we related the measure of effect size to each of the three study attributes. Given that

the impact of disturbance regimes on species assemblages manifests at different

spatial scales and can be temporally variable (e.g. Hobson & Schiek 1999; Walker &

Wardle 2014), we additionally factored an interaction term between these attributes

185

and disturbance level in our models. While we acknowledge that multiple effect

sizes compiled from a single study may not necessarily be independent (Koricheva et

al. 2013), we did not fit ‘study’ as a random effect due to the large number of studies

involved. Doing so could potentially conceal significant variation explained by the

study attributes.

The statistical significance of the predictor variables in the model were assessed by

an Omnibus (Wald-type) test implemented in the ‘metafor’ package, which

generates the test statistic Qm and a P-value. If the different levels of the predictors

were found to be significantly different, we performed post-hoc analysis to compare

pairwise differences by adjusting the reference coefficient (intercept) by using the

‘relevel’ function (Viechtbauer 2010). To evaluate model fit, we recorded the amount

of heterogeneity (R2) explained by each candidate model and ranked them with their

corresponding Akaike’s Information Criterion (AIC) values. The statistical

significance of the coefficients in each model was assessed by an Omnibus test,

followed by a post-hoc pairwise comparison of coefficients. All statistical analyses

were conducted using on the R platform (R Development Core Team 2013).

Results

We compiled a total of 1,633 measures of correlations of species richness and 1,030

correlations of species composition. Studies were best represented in Europe (63 of

146 studies; 43.1%) followed by North America (23 studies; 15.7%) (See

Supplementary Information Figure S1). Among studies that assessed pairwise

correlations of species richness between taxa, we found that 40 of 84 studies (47.6%)

were conducted in moderately disturbed landscapes (Table 1). For studies that

assessed cross-taxonomic correlations of species composition, we found that 19 of 39

studies (48.7%) were carried out in moderately disturbed landscapes, while the

remainder were in undisturbed or heavily-disturbed landscapes. When classified by

186

spatial extent, we found that the majority of studies examining correlations of

species richness were conducted at the regional (61 of 127 studies; 48.0%) and

landscape scales (25 studies; 22.0%) (Table 2). Similarly, the majority of studies

examining correlations of species composition were carried out at the landscape (13

of 45 studies; 28.8%) or regional scale (22 studies; 48.8%). The mean duration of a

study was 2.76 ± 3.67 (s.d.) years (N = 89) for correlations of species richness, and

2.95 ± 4.15 years (N = 41) for correlations of species composition. The majority of

studies were based on datasets collected within one year (Supplementary Figure S2).

Question 1. How do different levels of anthropogenic disturbance affect cross-

taxonomic congruence?

We found that average effect sizes for correlations of species richness were

consistently positive and varied across different levels of disturbance. Both levels of

disturbed landscapes exhibited higher average effect sizes than undisturbed

landscapes (x ̅ = 0.167, 0.134–0.194, P < 0.001) (Figure 1, Table 1). Average effect sizes

for correlations of species composition were higher than correlations of species

richness at low and moderate, but not high disturbance levels (Figure 1). Our mixed-

effects models revealed that both correlations of species richness (Model R2 = 0.161,

Qm = 149.462, P < 0.001) and species composition (Model R2 = 0.0790, Qm = 48.835, P

< 0.001) were significantly different across disturbance levels (Table 3). For example,

correlations of species richness for highly disturbed landscapes were significantly

higher than moderately disturbed (mean difference = 0.360, Z = 12.201, P < 0.001) and

undisturbed landscapes (mean difference = 0.236, Z = 7.699, P < 0.001). However,

correlations of species composition were different between highly disturbed and

moderately disturbed landscapes (mean difference = 0.136, Z = 6.154, P < 0.001), but

not with undisturbed landscapes.

Question 2. How does spatial scale affect patterns of cross-taxonomic congruence?

187

Average effect sizes for correlations of species richness were consistently positive,

but increased with spatial scale, being highest at the global scale (x ̅ = 0.928, 0.786–

1.071, P < 0.001) (Figure 2, Table 2). We found that a model incorporating the

interactive effects of spatial scale and disturbance best explained the variation in

effect sizes for correlations of species richness (R2 = 0.208, AIC = 1456.88) (See

Supplementary Table S1). Similarly, the best-fitted model to explain the variation in

effect sizes for correlations of species composition contained disturbance and spatial

scale as interaction terms (R2 = 0.307, AIC = 558.512) (See Supplementary Table S2).

Both models with this interaction term revealed that at the smallest spatial (local)

scale, effect sizes for correlations of species richness and composition declined with

increasing disturbance level (Figure 4). However, at larger spatial scales (landscape,

regional), effect sizes for both kinds of correlations increased with disturbance.

We found that spatial scale on its own was a weaker predictor of correlations of

species richness (model R2 = 0.151, Qm = 204.217, P < 0.001) than disturbance (model

R2 = 0.161, Qm = 149.217, P < 0.001). In this model, effect sizes for studies at the global

extent were significantly different from that at the continental extent (mean

difference = 0.238, Z = 3.032, P < 0.001) (Table 3). However, effect sizes were not

significantly different among local, landscape and regional scales. By comparison,

spatial scale was a stronger predictor of effect size for correlations of species

composition (model R2 = 0.189, Qm = 111.195, P < 0.001). All pairwise comparisons of

effect sizes between spatial scale levels were significantly different except for that

between continental and regional scales (mean difference = 0.103, Z = 1.500, P =

0.134), and between local and continental scales (mean difference = -0.0896, Z = -

1.193, P = 0.233).

Question 3. How does the duration of a study affect cross-taxonomic congruence?

188

Of a set of models fitted to explore the combinatorial effects of study duration and

disturbance level on species richness correlations, we found that the best fitting

model included disturbance and duration of study as an interaction term (R2 =

0.244, AIC = 1339.345) (See Supplementary Table S3). Similarly, we found that the

best-fitting model to explain effect sizes of correlations of species composition

contained study duration and disturbance as interaction terms (R2 = 0.212, AIC =

995.840) (Supplementary Table S4). All coefficients in this model were significant.

Both models with this interaction term revealed that at low disturbance levels, effect

sizes for correlations of species richness and composition declined with study

duration. However, at high disturbance levels, effect sizes for correlations of species

composition declined while that for species richness increased with study duration

(Figure 6). When modelled its own, we found that the duration of a study had a

negative effect on effect sizes for correlations of species richness (coefficient = -

0.0353, Z = -9.874, P < 0.001) (Figure 4, Table S3). However, the amount of variation

explained was low (R2 = 0.0806).

Discussion

Much remains to be learnt about the ecological contexts in which cross-taxonomic

surrogates work best, particularly with respect to anthropogenic disturbances,

spatial scale, and time (Wolters et al. 2006; Tonkin et al. 2016; Rooney & Azeria 2015;

Yong et al. 2016). This is surprising given the fact that disturbance regimes have been

well-recognised in shaping species diversity patterns (Hobbs & Huenneke 1992;

Fraterrigo & Rusak 2008). In this study, we found that anthropogenic disturbance

was an important predictor of patterns of cross-taxonomic congruence, and had

significant interactions with spatial scale and study duration. We discuss the

ecological implications of these findings, and what they mean for the use of cross-

taxonomic surrogates in the context of biodiversity assessments and monitoring.

189

Question 1. How do different levels of anthropogenic disturbance affect cross-

taxonomic congruence?

We found that the most heavily-disturbed landscapes such as farmland exhibited

the highest cross-taxonomic congruence, in contrast with less disturbed landscapes.

Few studies have examined the influence of disturbance regimes on cross-taxonomic

congruence. Comparing tropical rainforests subjected to fire-induced disturbance,

de Andrade et al. (2014) reported an increase in species richness correlations when

data from burnt sites were included in their analysis. However, unburnt forests at

small spatial scales yielded few significant correlations between different taxa (de

Andrade et al. 2014). Similarly, Rooney & Azeria (2015) found that highly disturbed

upland grasslands and boreal forests exhibited stronger cross-taxonomic congruency

compared with sites subjected to low disturbance. However, Anand et al. (2005)

found low cross-taxonomic congruence between taxa compared in heavily disturbed

and subsequently restored landscapes.

Our findings conflict with earlier propositions that heavily disturbed systems should

in principle exhibit weaker cross-taxonomic associations. For instance, Rooney &

Bayley (2012) predicted that cross-taxonomic relationships should decline after a

system has been disturbed. Hypothetically, anthropogenic disturbance can upset

relationships between species assemblages, increase temporal variability and disrupt

dynamics at the community and ecosystem level (Fraterrigo & Rusak 2008; Rooney &

Bayley 2012). Additionally, cross-taxonomic congruence may be diminished given

dissimilar responses to the disturbance gradient by different taxa (Gardner et al.

2009). We suggest that the stronger patterns of cross-taxonomic congruence

observed in highly disturbed landscapes, and at larger spatial scales (landscape,

regional) may have arisen due to similar responses across different taxa to

disturbance regimes (e.g. Pharo et al. 1999; Anand et al. 2005). Anthropogenic

disturbance and the accompanying loss of habitats can drive species extirpations

190

alongside invasion by generalist species (Hobbs & Huenneke 1992; McKinney 2006),

resulting in biotic homogenisation (Smart et al. 2006; Devictor et al. 2008; Dornelas

et al. 2014; Ribeiro-Neto et al. 2016) and other succession-driven shifts in

communities at small spatial scales (Lebrija-Trejos et al. 2010; Sitters et al. 2016).

Over time, highly disturbed landscapes may favour sets of generalists and species

with high dispersal capability, thus offsetting the declines suffered by rare and

specialist species (Ekroos et al. 2010).

Alternatively, it may be that environmental gradients resulting from disturbance can

promote cross-taxonomic congruence by enhancing the pool of species that

characterises different disturbance levels (Rooney & Azeria 2015). Larger study sites

thus encapsulate broader gradients of disturbance, and larger sets of species across

different taxa (Rooney & Azeria 2015). This may explain why cross-taxonomic

congruence at larger spatial scales increased with increasing disturbance level

(Figure 3), but decline at the smallest spatial scales (local) within which disturbance

gradients may be less pronounced. Further research should clarify how the intensity

and regularity of disturbance can shape cross-taxonomic congruence through its

direct effects on species assemblages responding to the disturbance gradient(s),

especially in tropical systems where high species diversity necessitate the use of

surrogate approaches in biodiversity assessments.

Question 2. How does spatial scale affect patterns of cross-taxonomic congruence?

The effects of space are well-documented determinants of patterns of cross-

taxonomic congruence patterns, and can be partitioned into that of spatial extent

(e.g. Pearson & Caroll 1999; Gaspar et al. 2010) and spatial grain (e.g. Hess et al.

2006; Qian & Kissling 2010). Consistent with other studies (Grenyer et al. 2006;

Westgate et al. 2014), our analyses showed that cross-taxonomic congruence can be

explained by the spatial scale of the analysis. Effect sizes for cross-taxonomic

191

congruence were highest in studies at the global extent, declining as spatial extent

decreased. Stronger patterns of cross-taxonomic congruence at large spatial scales

and grains arise because the species diversity of different taxa responds broadly to

climatic and environmental gradients operating at large spatial resolutions (e.g.

McKnight et al. 2007; Qian & Ricklefs 2008). However, at smaller spatial scales of

investigation, our analysis suggested that disturbance regimes play a more

important role than spatial effects (Figure 6). Moreover, other responses driven by

ecological traits of species such as dispersal and sensitivity to fine-scale

environmental variation may manifest more strongly at smaller spatial extents

(Pearson & Caroll 1999; Gardner et al. 2009). Because how spatial effects (at large

scales) influence cross-taxonomic congruence is already well documented, a

research priority is to better understand the ecological processes at the species (e.g.

dispersal) and community level (e.g. co-occurrence, trophic interactions) that drive

these patterns.

Question 3. How does the duration of a study affect cross-taxonomic congruency?

Our analyses revealed that the duration of a study can negatively influence cross-

taxonomic congruence. Such a relationship may arise because studies based on

short-term sampling effort invariably detect a higher proportion of common,

generalist species compared to rare species which are detected with greater

sampling effort (Gotelli & Colwell 2011). Cross-taxonomic patterns among rare and

threatened species are known to be weak (Grenyer et al. 2006). Many rare species

may not necessarily occur in areas of high species richness (Prendergast et al. 1993)

or even occur inconsistently (Hewitt et al. 2016). Increased sampling effort spanning

multiple years may cumulatively add more rare species in the species pool, therefore

weakening patterns of cross-taxonomic congruence.

192

A secondary explanation is that ecological communities are temporally variable in

response to various biotic (i.e. disturbance) and abiotic factors, such as seasonal

effects, fluctuations in resources, species-specific population dynamics and

environmental stochasticity (Collins 2000; Hewitt et al. 2016), all which may interact

synergistically. This may alter co-occurrence patterns, food webs or other biotic

interactions (e.g. Larsen & Ormerod 2014; Tonkin et al. 2016), introducing temporal

variation to species pools and overall species composition (Collins 2000; Magurran

et al. 2010; Hewitt et al. 2016), altering patterns of cross-taxonomic congruence as a

result. Aggregating the temporal variation of different taxa into a long-term dataset

may add ecological ‘noise’, while failing to track the shifts in ecological

communities.

The effects of temporal variation on surrogate efficacy remains poorly understood

(Favreau et al. 2006; Yong et al. 2016). Such a knowledge gap persists because most

surrogate research is focused on addressing short-term questions through ‘snap-

shot’ studies (see Supplementary Figure S2). In spite of better recognition of the

importance of longitudinal studies (Magurran et al. 2010; Lindenmayer et al. 2012),

the long-term data collection needed to address these gaps is impeded by logistical

and funding limitations (Thomson et al., 2005; Turner 2010). Given that human-

altered landscapes are temporally dynamic (e.g. Tscharntke et al. 2005), the

variation in species diversity post-disturbance (e.g. Sitters et al. 2016) may

potentially undermine the use of cross-taxonomic surrogates in long-term

monitoring of biodiversity. It is therefore important to assess if these cross-

taxonomic congruency patterns remain consistent over time (Tulloch et al. 2016),

and if not, can its variation over time be predicted.

Conclusion and implications

193

Our study provides the first global synthesis on the little-examined effects of

anthropogenic disturbance on cross-taxonomic congruence patterns in terrestrial

ecosystems. While our findings validate earlier work highlighting the importance of

spatial effects in driving cross-taxonomic patterns (e.g. Grenyer et al. 2006; Westgate

et al. 2014), we provide new evidence that patterns of cross-taxonomic congruence at

small to moderate spatial scales can arise as a legacy of disturbance. Our findings

pose implications for conservation in four key ways. First, cross-taxonomic

surrogates of biodiversity may best be suited for biodiversity assessments in

disturbed landscapes, and at landscape or regional spatial scales, but not at very

small scales. Surrogate taxa identified at large spatial scales should thus be

cautiously used in assessing biodiversity at the smaller spatial scales at which

conservation management takes place. Second, conservation practitioners will need

to carefully evaluate the kinds of taxa selected as surrogates and how each may

respond to disturbance differently. Taxa found to be good surrogates in landscapes

subjected to one kind of disturbance may not necessary be effective in relatively

undisturbed landscapes or landscapes subjected to different disturbance regimes.

Third, taxa that are highly sensitive to disturbance may not be good cross-taxonomic

surrogates because such groups may be quickly impacted in disturbed landscapes.

Finally, a cross-taxonomic surrogate approach may not necessarily be appropriate

for biodiversity assessments or monitoring in relatively pristine landscapes since one

taxonomic group may poorly reflect diversity or occurrences of other suites of biota.

In such situations, conservation practitioners should consider a direct measure

approach targeting the biodiversity component(s) of interest (see Lindenmayer &

Likens 2011) over a surrogate-based approach.

Acknowledgments

DLY thanks the Lesslie Foundation for supporting this study. DBL is supported by an

Australian Research Council (ARC) Laureate Fellowship. We are grateful to Dr

194

Martin Westgate and Dr M. John Evans for useful discussions during the preparation

of this study.

References

Albuquerque, F., & Beier, P. (2016) Predicted rarity‐weighted richness, a new tool to

prioritize sites for species representation. Ecology and Evolution, 6, 8107-8114.

Anand, M., Laurence, S., & Rayfield, B. (2005) Diversity relationships among

taxonomic groups in recovering and restored forests. Conservation Biology, 19,

955-962.

Banks, S. C., Cary, G. J., Smith, A. L., Davies, I. D., Driscoll, D. A., Gill, A. M.,

Lindenmayer, D. B., & Peakall, R. (2013) How does ecological disturbance

influence genetic diversity? Trends in Ecology & Evolution, 28(11), 670-679.

Barlow, J., Gardner, T. A., Araujo, I. S., Ávila-Pires, T. C., Bonaldo, A. B., Costa, J. E.,

Esposito, M. C., Ferreira, L. V., Hawes, J., Hernandez, M. I. M., Hoogmoed, M. S.,

Leite, R. N., Lo, M. H. N. F., Malcolm, J. R., Martins, M. B., Mestre, L. A. M.,

Miranda-Santos, R., Nunes-Gutjahr, A. L., Overal, W. L., Parry, L., Peters, S. L.,

Ribeiro, M. A., da Silva, M. N. F., da Silva Motta, C., & Peres, C. A. (2007)

Quantifying the biodiversity value of tropical primary, secondary, and plantation

forests. Proceedings of the National Academy of Sciences U.S.A., 104(47), 18555-

18560.

Barton, P. S., Pierson, J. C., Westgate, M. J., Lane, P. W., & Lindenmayer, D. B. (2015)

Learning from clinical medicine to improve the use of surrogates in ecology.

Oikos, 124(4), 391-398.

Borenstein, M., Hedges, L. V., Higgins, J. P. T., & Rothstein, H. R. (2010) A basic

introduction to fixed-effect and random-effects model for meta-analysis. Research

Synthesis Methods 1, 97-111.

195

Bottrill, M. C., Joseph, L. N., Carwardine, J., Bode, M., Cook, C., Game, E. T.,

Grantham, H., Kark, S., Linke, S., McDonald-Madden, E., Pressey, R. L., Walker,

S., Wilson, K., & Possingham, H. P. (2008) Is conservation triage just smart

decision making? Trends in Ecology & Evolution, 23(12), 649-654.

Brooks, T. M., Mittermeier, R. A., da Fonseca, G. A., Gerlach, J., Hoffmann, M.,

Lamoreux, J. F., Mittermeier, C. G., Pilgrim, J. D., & Rodrigues, A. S. (2006) Global

biodiversity conservation priorities. Science, 313(5783), 58-61.

Brown, K. S. (1997) Diversity, disturbance, and sustainable use of Neotropical forests:

insects as indicators for conservation monitoring. Journal of Insect Conservation,

1, 25-42.

Butchart, S. H., Walpole, M., Collen, B., Van Strien, A., Scharlemann, J. P., Almond,

R. E., Baillie, J. E. M., Bomhard, B., Brown, C., Bruno, J., Carpenter, K. E., Carr, G.

M., Chanson, J., Chenery, A. M., Csirke, J., Davidson, N. C., Dentener, F., Foster,

M. Galli, A., Galloway, J. N., Genovesi, P., Gregory, R. D., Hockings, M., Kapos, V.,

Lamarque, J., Leverington, F., Loh, J., McGeoch, M., McRae, L., Minasyan, A.,

Morcillo, M. H., Oldfield, T. E. E., Pauly, D., Quader, S., Revenga, C., Sauer, J. R.,

Skolnik, B., Spear, D., Stanwell-Smith, D., Stuart, S. N., Symes, A., Tierney, M.,

Tyrrell, T. D., Vie, J., & Watson R., (2010) Global biodiversity: indicators of recent

declines. Science, 328, 1164-1168.

Collins, S. L. (2000) Disturbance frequency and community stability in native

tallgrass prairie. The American Naturalist, 155(3), 311-325.

Cushman, S. A., McKelvey, K. S., Noon, B. R., & McGarigal, K. (2010) Use of

abundance of one species as a surrogate for abundance of others. Conservation

Biology, 24(3), 830-840.

De Andrade, R. B., Barlow, J., Louzada, J., Mestre, L., Silveira, J., Vaz-de-Mello, F. Z.,

& Cochrane, M. A. (2014) Biotic congruence in humid tropical forests: a multi-taxa

examination of spatial distribution and responses to forest disturbance. Ecological

Indicators, 36, 572-581.

196

Devictor, V., Julliard, R., Clavel, J., Jiguet, F., Lee, A., & Couvet, D. (2008) Functional

biotic homogenization of bird communities in disturbed landscapes. Global

Ecology and Biogeography, 17(2), 252-261.

Dobson, A., Lodge, D., Alder, J., Cumming, G. S., Keymer, J., McGlade, J., Mooney,

H., Rusak, J. A., Sala, O., Wolters, V., Wall, D., Winfree, R., & Xenopoulos, M. A.

(2006) Habitat loss, trophic collapse, and the decline of ecosystem services.

Ecology, 87(8), 1915-1924.

Dornelas, M., Gotelli, N. J., McGill, B., Shimadzu, H., Moyes, F., Sievers, C., &

Magurran, A. E. (2014) Assemblage time series reveal biodiversity change but not

systematic loss. Science, 344(6181), 296-299.

Eglington, S. M., Noble, D. G., & Fuller, R. J. (2012) A meta-analysis of spatial

relationships in species richness across taxa: birds as indicators of wider

biodiversity in temperate regions. Journal for Nature Conservation, 20(5), 301-309.

Ekroos, J., Heliölä, J., & Kuussaari, M. (2010) Homogenization of lepidopteran

communities in intensively cultivated agricultural landscapes. Journal of Applied

Ecology, 47(2), 459-467.

Elser, J. J., Fagan, W. F., Denno, R. F., Dobberfuhl, D. R., Folarin, A., Huberty, A.,

Interlandi, S., Kilham, S. S., McCauley, E., Schulz, K. L., Siemann, E. H., & Sterner,

R. W. (2000) Nutritional constraints in terrestrial and freshwater food webs.

Nature, 408(6812), 578-580.

FAO. (2017) The future of food and agriculture – Trends and challenges. Food and

Agriculture Organisation of the United Nations, Rome.

Fattorini, S., Dennis, R. L., & Cook, L. M. (2012) Use of cross-taxon congruence for

hotspot identification at a regional scale. PLoS One, 7, e40018.

Favreau, J. M., Drew, C. A., Hess, G. R., Rubino, M. J., Koch, F. H., & Eschelbach, K.

A. (2006) Recommendations for assessing the effectiveness of surrogate species

approaches. Biodiversity and Conservation, 15(12), 3949-3969.

197

Fischer, J., Lindenmayer, D. B., & Manning, A. D. (2006) Biodiversity, ecosystem

function, and resilience: ten guiding principles for commodity production

landscapes. Frontiers in Ecology and the Environment, 4(2), 80-86.

Fraterrigo, J. M., & Rusak, J. A. (2008) Disturbance‐driven changes in the variability

of ecological patterns and processes. Ecology Letters, 11(7), 756-770.

Gardner, T. A., Barlow, J., Chazdon, R., Ewers, R. M., Harvey, C. A., Peres, C. A., &

Sodhi, N. S. (2009) Prospects for tropical forest biodiversity in a human‐modified

world. Ecology Letters, 12(6), 561-582.

Gaspar, C., Gaston, K. J., & Borges, P. A. (2010) Arthropods as surrogates of diversity

at different spatial scales. Biological Conservation, 143(5), 1287-1294.

Gotelli, N. J., & Colwell, R. K. (2011) Estimating species richness. In: Magurran, A. E.,

& McGill, B. J. eds. Biological diversity: frontiers in measurement and assessment.

Oxford University Press, U.K.

Grenyer, R., Orme, C. D. L., Jackson, S. F., Thomas, G. H., Davies, R. G., Davies, T. J.,

Jones, K. E., Olson, V. A., Ridgely, R. S., Rasmussen, P. C., Ding, T-S., Bennett, P.

M., Blackburn, T. M., Gaston, K. J., Gittleman, J. L., & Owens, I. P. (2006) Global

distribution and conservation of rare and threatened vertebrates. Nature 444

(7115): 93-96.

Heino, J. (2010) Are indicator groups and cross-taxon congruence useful for

predicting biodiversity in aquatic ecosystems? Ecological Indicators, 10(2), 112-117.

Heino, J., & Soininen, J. (2007) Are higher taxa adequate surrogates for species-level

assemblage patterns and species richness in stream organisms? Biological

Conservation, 137(1), 78-89.

Hess, G. R., Bartel, R. A., Leidner, A. K., Rosenfeld, K. M., Rubino, M. J., Snider, S. B.,

& Ricketts, T. H. (2006) Effectiveness of biodiversity indicators varies with extent,

grain, and region. Biological Conservation, 132(4), 448-457.

198

Hewitt, J. E., Thrush, S. F., & Ellingsen, K. E. (2016) The role of time and species

identities in spatial patterns of species richness and conservation. Conservation

Biology, 30(5), 1080-1088.

Higa, M., Yamaura, Y., Senzaki, M., Koizumi, I., Takenaka, T., Masatomi, Y., &

Momose, K. (2016) Scale dependency of two endangered charismatic species as

biodiversity surrogates. Biodiversity and Conservation, 25(10), 1829-1841.

Hobbs, R. J., & Huenneke, L. F. (1992) Disturbance, diversity, and invasion:

implications for conservation. Conservation Biology, 6(3), 324-337.

Hobson, K. A., & Schieck, J. (1999) Changes in bird communities in boreal mixed

wood forest: harvest and wildfire effects over 30 years. Ecological Applications,

9(3), 849-863.

Koh, L. P., & Wilcove, D. S. (2008) Is oil palm agriculture really destroying tropical

biodiversity? Conservation Letters, 1(2), 60-64.

Koh, L. P., & Gardner, T. A. (2010) Conservation in human-modified landscapes. In

Sodhi, N. S., & Ehrlich, P. R. eds. Conservation biology for all. Oxford University

Press, U.K.

Koricheva, J., Gurevitch, J., & Mengersen, K. eds. (2013) Handbook of Meta-analysis in

Ecology and Evolution. Princeton University Press, U.S.A.

Lajeunesse, M. J. (2013) Recovering missing or partial data from studies: a survey of

conversions and imputations for meta-analysis. In: Koricheva, J., Gurevitch, J., &

Mengersen, K. eds. Handbook of Meta-analysis in Ecology and Evolution.

Princeton University Press, U.S.A.

Landeiro, V. L., Bini, L. M., Costa, F. R., Franklin, E., Nogueira, A., de Souza, J. L.,

Moraes, J., & Magnusson, W. E. (2012) How far can we go in simplifying

biomonitoring assessments? An integrated analysis of taxonomic surrogacy,

taxonomic sufficiency and numerical resolution in a megadiverse region.

Ecological Indicators, 23, 366-373.

199

Larsen, S., & Ormerod, S. J. (2014) Anthropogenic modification disrupts species co‐

occurrence in stream invertebrates. Global Change Biology, 20(1), 51-60.

Larsen, F. W., Bladt, J., Balmford, A., & Rahbek, C. (2012) Birds as biodiversity

surrogates: will supplementing birds with other taxa improve effectiveness?

Journal of Applied Ecology, 49(2), 349-356.

Lindenmayer, D. B., & Likens, G. E. (2011) Direct measurement versus surrogate

indicator species for evaluating environmental change and biodiversity loss.

Ecosystems, 14(1), 47-59.

Lindenmayer, D. B., Likens, G. E., Andersen, A., Bowman, D., Bull, C. M., Burns, E.,

Dickman, C. R., Hoffman, A. A., Keith, D. A., Liddell, M. J., Lowe, A. J., Metcalfe,

D. J., Phinn, S. R., Russell-Smith, J., Thurgate, N., & Wardle, G. M. (2012) Value of

long‐term ecological studies. Austral Ecology, 37(7), 745-757

Lindenmayer, D.B., Barton, P., & Pierson, J. eds. (2015) Indicators and Surrogates of

Biodiversity and Environmental Change. CSIRO Publishing, Australia & CRC Press,

USA.

Magurran, A. E., Baillie, S. R., Buckland, S. T., Dick, J. M., Elston, D. A., Scott, E. M.,

Smith, R. I., Somerfield, P. J., & Watt, A. D. (2010) Long-term datasets in

biodiversity research and monitoring: assessing change in ecological communities

through time. Trends in Ecology & Evolution, 25(10), 574-582.

Maxwell, S. L., Fuller, R. A., Brooks, T. M., & Watson, J. E. (2016) Biodiversity: The

ravages of guns, nets and bulldozers. Nature, 536(7615), 143-145.

McKinney, M. L. (2006) Urbanization as a major cause of biotic homogenization.

Biological Conservation, 127(3), 247-260.

McKnight, M. W., White, P. S., McDonald, R. I., Lamoreux, J. F., Sechrest, W.,

Ridgely, R. S., & Stuart, S. N. (2007) Putting beta-diversity on the map: broad-

scale congruence and coincidence in the extremes. PLoS Biology, 5, e272.

200

McMullan‐Fisher, S. J., Kirkpatrick, J. B., May, T. W., & Pharo, E. J. (2010) Surrogates

for macrofungi and mosses in reservation planning. Conservation Biology, 24(3),

730-736.

Miller, A. D., Roxburgh, S. H., & Shea, K. (2011) How frequency and intensity shape

diversity–disturbance relationships. Proceedings of the National Academy of

Sciences U.S.A., 108(14), 5643-5648.

Nakagawa, S., Noble, D. W. A., Senior, A. M., & Lagisz, M. (2017) Meta-evaluation of

meta-analysis: ten appraisal questions for biologists. BMC Biology 15, 18.

Oliver, I., Beattie, A. J., & York, A. (1998) Spatial fidelity of plant, vertebrate, and

invertebrate assemblages in multiple‐use forest in eastern Australia. Conservation

Biology, 12(4), 822-835.

Pharo, E. J., Beattie, A. J., & Binns, D. (1999) Vascular plant diversity as a surrogate

for bryophyte and lichen diversity. Conservation Biology, 13(2), 282-292.

Possingham, H. P., Grantham, H., & Rondinini, C. (2007) How can you conserve

species that haven't been found? Journal of Biogeography, 34(5), 758-759.

Prendergast, J. R., Quinn, R. M., Lawton, J. H., Eversham, B. C., & Gibbons, D. W.

(1993) Rare species, the coincidence of diversity hotspots and conservation

strategies. Nature, 365, 335-337.

Qian, H., & Kissling, W. D. (2010) Spatial scale and cross‐taxon congruence of

terrestrial vertebrate and vascular plant species richness in China. Ecology, 91(4),

1172-1183.

R Development Core Team. (2013) R: A language and environment for statistical

computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-

900051-07-0, URL http://www.R-project.org/.

Ribeiro-Neto, J. D., Arnan, X., Tabarelli, M., & Leal, I. R. (2016) Chronic

anthropogenic disturbance causes homogenization of plant and ant communities

in the Brazilian Caatinga. Biodiversity and Conservation, 25(5), 943-956.

201

Rodrigues, A. S., & Brooks, T. M. (2007) Shortcuts for biodiversity conservation

planning: the effectiveness of surrogates. Annual Review of Ecology, Evolution, and

Systematics, 38, 713-737.

Rooney, R. C., & Azeria, E. T. (2015) The strength of cross‐taxon congruence in

species composition varies with the size of regional species pools and the

intensity of human disturbance. Journal of Biogeography, 42(3), 439-451.

Sætersdal, M., & Gjerde, I. (2011) Prioritising conservation areas using species

surrogate measures: consistent with ecological theory? Journal of Applied Ecology,

48(5), 1236-1240.

Sauberer, N., Zulka, K. P., Abensperg-Traun, M., Berg, H. M., Bieringer, G.,

Milasowszky, N., Moser, D., Plutzar, C., Pollheimer, M., Storck, C., Tröstl, R.,

Zechmeister, H., & Grabherr, G. (2004) Surrogate taxa for biodiversity in

agricultural landscapes of eastern Austria. Biological Conservation, 117(2), 181-190.

Schuldt, A., Wang, Z., Zhou, H., & Assmann, T. (2009) Integrating highly diverse

invertebrates into broad‐scale analyses of cross‐taxon congruence across the

Palaearctic. Ecography, 32(6), 1019-1030.

Sitters, H., Di Stefano, J., Christie, F., Swan, M., & York, A. (2016) Bird functional

diversity decreases with time since disturbance: Does patchy prescribed fire

enhance ecosystem function? Ecological Applications, 26(1), 115-127.

Smart, S. M., Thompson, K., Marrs, R. H., Le Duc, M. G., Maskell, L. C., & Firbank, L.

G. (2006) Biotic homogenization and changes in species diversity across human-

modified ecosystems. Proceedings of the Royal Society of London B: Biological

Sciences, 273(1601), 2659-2665.

Sousa-Souto, L., Figueiredo, P., Ambrogi, B., Oliveira, A., Ribeiro, G. T., & Neves, F.

(2016) Composition and richness of arboreal ants in fragments of Brazilian

Caatinga: effects of secondary succession. Sociobiology, 63(2), 762-769.

Sterne, J. A., Sutton, A. J., Ioannidis, J. P., Terrin, N., Jones, D. R., Lau, J., Carpenter,

J., Rucker, G., Harbord, R. M., Schmid, C. H., Tetzlaff, J., Deeks, J. J., Peters, J.,

202

Macaskill, P., Schwarzer, G., Duval, S., Altman, D. G., Moher, D., & Higgins, J. P.

T. (2011) Recommendations for examining and interpreting funnel plot

asymmetry in meta-analyses of randomised controlled trials. The BMJ, 343, d4002.

Swanson, M. E., Franklin, J. F., Beschta, R. L., Crisafulli, C. M., DellaSala, D. A.,

Hutto, R. L., Lindenmayer, D. B., & Swanson, F. J. (2011) The forgotten stage of

forest succession: early‐successional ecosystems on forest sites. Frontiers in

Ecology and the Environment, 9(2), 117-125.

Thomson, J. R., Fleishman, E., Mac Nally, R., & Dobkin, D. S. (2005) Influence of the

temporal resolution of data on the success of indicator species models of species

richness across multiple taxonomic groups. Biological Conservation, 124(4), 503-

518.

Tonkin, J. D., Stoll, S., Jähnig, S. C., & Haase, P. (2016) Anthropogenic land-use stress

alters community concordance at the river-riparian interface. Ecological

Indicators, 65, 133-141.

Tscharntke, T., Klein, A. M., Kruess, A., Steffan‐Dewenter, I., & Thies, C. (2005)

Landscape perspectives on agricultural intensification and biodiversity–ecosystem

service management. Ecology Letters, 8(8), 857-874.

Tulloch, A. I., Chadès, I., Dujardin, Y., Westgate, M. J., Lane, P. W., & Lindenmayer,

D. (2016) Dynamic species co–occurrence networks require dynamic biodiversity

surrogates. Ecography, 39, 1185-1196.

Turner, M. G. (2010) Disturbance and landscape dynamics in a changing world.

Ecology, 91(10), 2833-2849.

Viechtbauer, W. (2010). Conducting meta-analyses in R with the metafor package.

Journal of Statistical Software, 36(3), 1-48.

Walker, L. R., & Wardle, D. A. (2014) Plant succession as an integrator of contrasting

ecological time scales. Trends in Ecology & Evolution, 29(9), 504-510.

203

Westgate, M. J., Barton, P. S., Lane, P. W., & Lindenmayer, D. B. (2014) Global meta-

analysis reveals low consistency of biodiversity congruence relationships. Nature

communications, 5.

Whittaker, R. J., Willis, K. J., & Field, R. (2001) Scale and species richness: towards a

general, hierarchical theory of species diversity. Journal of Biogeography, 28(4),

453-470.

Wiens, J. A., Hayward, G. D., Holthausen, R. S., & Wisdom, M. J. (2008) Using

surrogate species and groups for conservation planning and management.

Bioscience, 58(3), 241-252.

Wolters, V., Bengtsson, J., & Zaitsev, A. S. (2006) Relationship among the species

richness of different taxa. Ecology, 87(8), 1886-1895.

Yong, D. L., Barton, P. S., Okada, S., Crane, M., & Lindenmayer, D. B. (2016) Birds as

surrogates for mammals and reptiles: Are patterns of cross-taxonomic

associations stable over time in a human-modified landscape? Ecological

Indicators, 69, 152-164.

204

Figures and tables

Figure 1. Average effect sizes for correlations of species composition and richness

estimated using a random-effects model across three classes of habitat disturbance.

205

Figure 2. Average effect sizes for correlations of species composition and species

richness estimated using a random-effects model across five classes of spatial extent.

206

Figure 3. Fitted regression lines (with 95% confidence regions) showing the

(multiplicative) effects of interactions between spatial scale and disturbance level,

on: (a) effect sizes of correlations of species richness and, (b) correlations of species

composition. At low (local) spatial extents, effect sizes declined when disturbed.

However, at the landscape and regional extents, effect sizes increased with

disturbance levels.

207

Figure 4. Regression lines (with 95% confidence regions) relating study duration

(number of years) to effect sizes for correlations of species richness.

208

Figure 5. Fitted regression lines (with 95% confidence regions) showing the

(multiplicative) effects of interactions between duration of study and disturbance

level on: (a) effect sizes of correlations of species richness and, (b) correlations of

species composition. Effect sizes for both measures declined over time at low

disturbance levels. At moderate levels of disturbance, effect sizes increased over

time.

209

Figure 6. A conceptual framework showing the increased role of disturbance in

shaping cross-taxonomic congruency patterns at low spatial scales. At large spatial

scales, disturbance becomes relatively less important.

210

Table 1. Summary statistics (to 3 significant figures) for random-effects models

fitted for correlations of two measures of diversity (species richness, species

composition) across three broad classes of habitat disturbances.

Random

effects model

Mean effect

size ( ± se)

Sample

size (N)

Number

of studies

τ2 I2 (%) H2 Q

Species richness

High

disturbance

0.531 ± 0.0242 359 20 0.144 74.00 3.85 1351.027

Moderate

disturbance

0.292 ± 0.0220 452 40 0.151 80.40 5.10 1882.684

Low

disturbance

0.167 ± 0.0167 459 31 0.0934 79.13 4.79 1973.650

Species composition

High

disturbance

0.206 ± 0.0354 52 5 0.0293 50.13 2.01 98.999

Moderate

disturbance

0.435 ± 0.0173 440 19 0.0819 65.82 2.93 1476.883

Low

disturbance

0.304 ± 0.0167 446 17 0.0904 79.07 4.78 2020.077

All values of the Q statistic (test for heterogeneity) was highly significant (P < 0.001).

211

Table 2. Summary statistics (to 3 significant figures) for random-effects models

fitted for effect sizes of two measures of species diversity (species richness, species

composition) across spatial scale levels.

Random

effects model

Mean effect

size ( ± se)

Sample

size (N)

Number

of

studies

τ2 I2 (%) H2 Q

Species richness

Global 0.928 ± 0.0727 37 6 0.167 95.07 20.26 587.124

Continental 0.688 ± 0.0371 226 22 0.294 98.26 57.31 13485.899

Regional 0.365 ± 0.0189 526 61 0.140 93.41 15.18 4786.884

Landscape 0.277 ± 0.0166 637 28 0.130 80.53 5.14 2649.491

Local 0.310 ± 0.0338 204 13 0.153 80.65 5.17 893.209

Species composition

Continental 0.503 ± 0.0865 18 3 0.116 99.04 104.38 2054.077

Regional 0.431 ± 0.0214 329 22 0.0984 75.58 4.09 1293.715

Landscape 0.278 ± 0.0123 588 15 0.048 61.99 2.63 1615.763

Local 0.571 ± 0.0482 92 6 0.155 80.37 5.09 420.941

a. All values of the Q statistic (test for heterogeneity) was highly significant (P < 0.0001).

b. There was no data on correlation of species composition at the global scale

212

Table 3. Summary of results for post-hoc (Wald-test type) analysis for mixed-effects

models examining the effects of (1) disturbance and (2) spatial scale on effect sizes of

species richness and species composition

Factor

Pairwise comparison Post-hoc test parameters

Estimate

( ± se)

Z-value P-value 95% conf. interval

Lower Upper

Species richness

d.f. = 1247 Model Qm = 149.462** Model R2 = 16.10%

Disturbance Low Moderate 0.124 ± 0.0277 4.487 <0.001 0.0699 0.178

High 0.360 ± 0.0295 12.201 <0.001 0.303 0.418

Moderate High 0.236 ± 0.0307 7.699 <0.001 0.176 0.297

d.f. = 1605 Model Qm = 204.217** Model R2 = 15.14%

Spatial Local Landscape -0.0258 ± 0.0396 -0.651 0.515 -0.104 0.0519

Regional 0.0565 ± 0.0406 1.393 0.164 -0.0230 0.136

Continental 0.382 ± 0.0453 8.416 <0.001 0.293 0.471

Global 0.619 ± 0.0811 7.641 <0.001 0.461 0.779

Landscape Regional 0.0823 ± 0.0275 2.997 0.0027 0.0285 0.136

Continental 0.407 ± 0.0341 11.943 <0.001 0.341 0.474

Global 0.646 ± 0.0754 8.561 <0.001 0.498 0.794

Regional Continental 0.325 ± 0.0352 9.237 <0.001 0.256 0.394

Global 0.563 ± 0.0759 7.421 <0.001 0.415 0.712

Continental Global 0.238 ± 0.0786 3.032 0.0024 0.084 0.392

Species composition

d.f. = 984 Model Qm = 48.835** Model R2 = 7.86%

Disturbance Low Moderate 0.136 ± 0.0221 6.154 <0.001 0.0927 0.179

High -0.092 ± 0.0478 -1.929 0.0536 -0.186 0.00140

Moderate High -0.228 ± 0.0482 -4.737 <0.001 -0.323 -0.134

d.f. = 1017 Model Qm = 111.195** Model R2 = 18.9%

Spatial Local Landscape -0.341 ± 0.0378 -9.028 <0.001 -0.415 -0.267

Regional -0.193 ± 0.0399 -4.838 <0.001 -0.271 -0.115

Continental -0.0896 ± 0.0751 -1.193 0.233 -0.237 0.0576

Landscape Regional 0.148 ± 0.0227 6.536 <0.001 0.104 0.193

Continental 0.251 ± 0.0676 3.719 <0.001 0.119 0.384

Regional Continental 0.103 ± 0.0688 1.500 0.1336 -0.0316 0.238

Qm is the test statistic for the Omnibus (Wald-type) test of moderator significance

P-value: <0.001***, <0.01** , <0.05*

213

Supplementary figures and tables

Figure S1. Map of geographical location of (local to regional scales) studies included in the meta-analysis. Study sites (up to

regional scale) are represented as red circles. Countries that contain at least one study are shaded brown. Shaded countries that

lack dots contain studies that are conducted at the national-level.

214

Figure S2. Studies of cross-taxonomic surrogates classified by duration of study (in

years). Grey bars represent studies measuring correlations of species richness.

Orange bars represent studies measuring correlations of species composition.

0

5

10

15

20

25

30

35

40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Nu

mb

er

of

stu

die

s

Duration of study (years)

215

Figure S3. Funnel plots showing scatter of effect sizes with respect to the standard

deviation of each effect size for, (a) correlations of species richness and, (b)

correlation of species composition.

216

Table S1. Summary statistics of best-fitted models ranked by AIC values relating

spatial scale and disturbance level to effect size (correlation of species richness).

Model Estimate std

error

Z-value P-value Goodness-of-

fit statistics Model parameter(s)

Effect size ~ spatial* disturb

R2 0.208

AIC 1456.880

∆AIC 0.00

Qe 4911.006**

Qm 211.052**

Intercept 0.366 0.0480 7.628 <0.001

Mod. disturbed -0.0786 0.0629 -1.250 0.211

High. disturbed -0.221 0.184 -1.202 0.229

Landscape -0.281 0.0529 -5.308 <0.001

Regional -0.0442 0.0641 -0.690 0.491

Mod. disturbed : landscape 0.194 0.0827 2.342 0.0192

Highly disturbed : landscape 0.686 0.187 3.660 <0.001

Moderately disturbed : regional 0.0808 0.0803 1.006 0.314

Highly disturbed : regional 0.393 0.195 2.018 0.0436

Effect size ~ spatial + disturb

R2 0.191

AIC 1480.375

∆AIC 23.495

Qe 4970.034**

Qm 177.174**

Intercept 0.269 0.0355 7.605 <0.001

Mod. disturbed 0.0598 0.0306 1.957 0.0504

High. disturbed 0.374 0.0295 12.682 <0.001

Landscape -0.144 0.0375 -3.839 <0.001

Regional -0.0190 0.0372 -0.509 0.6109

Effect size ~ disturb

R2 0.161

AIC 1499.684

∆AIC 42.804

Qe 5207.362**

Qm 149.462**

Intercept 0.170 0.0186 9.116 <0.001

Mod. disturbed 0.124 0.0277 4.4874 <0.001

High. disturbed 0.360 0.0295 12.201 <0.001

Effect size ~ spatial

R2 0.151

AIC 2158.806

∆AIC 701.926

Qe 22402.607**

Qm 204.217*

Intercept 0.309 0.0351 8.793 <0.001

Landscape -0.0258 0.0396 -0.651 0.515

Regional 0.0565 0.0406 1.393 0.164

Continental 0.382 0.0453 8.416 <0.001

Global 0.620 0.0811 7.641 <0.001

Qm is the test statistic for the Omnibus (Wald-type) test of moderator significance

Qe is the test statistic for the test of residual (unexplained) heterogeneity

P-value: <0.001 ***, <0.01 ** , <0.05 *

217

Table S2. Summary statistics of best-fitted models ranked by AIC values relating

spatial extent and disturbance level to effect size (correlation of species

composition).

Model Estimate Std error Z-value P-value Goodness-of-

fit statistics Model parameter(s)

Effect size ~ spatial * disturb

R2 0.307

AIC 558.512

∆AIC 0.00

Qe 2520.638**

Qm 217.582**

Intercept 0.839 0.0494 16.996 <0.001

Landscape -0.603 0.0516 -11.675 <0.001

Regional -0.483 0.0643 -7.509 <0.001

Mod. disturbed -0.410 0.0678 -6.047 <0.001

High. disturbed -0.116 0.217 -0.533 0.594

Landscape : mod. disturbed 0.567 0.0734 7.728 <0.001

Regional : mod. disturbed 0.518 0.0821 6.309 <0.001

Regional : high. disturbed 0.0459 0.228 0.202 0.840

Landscape : high. disturbed -0.0785 0.227 -0.346 0.729

Effect size ~ spatial + disturb

R2 0.245

AIC 611.361

∆AIC 52.849

Qe 2633.452**

Qm 143.540**

Intercept 0.583 0.0364 16.005 <0.001

Landscape -0.326 0.0370 -8.826 <0.001

Regional -0.198 0.0393 -5.031 <0.001

Mod. disturbed 0.0772 0.0231 3.339 <0.001

High. disturbed -0.137 0.0460 -2.970 0.003

Effect size ~ spatial

R2 0.189

AIC 667.455

∆AIC 108.943

Qe 4885.419**

Qm 111.195**

Intercept 0.614 0.0354 17.367 <0.001

Landscape -0.341 0.0378 -9.028 <.0001

Regional -0.193 0.0399 -4.838 0.0004

Continental -0.0896 0.0751 -1.193 0.2328

Effect size ~ disturb

R2 0.0786

AIC 686.120

∆AIC 127.608

Qe 3101.374**

Qm 48.835*

Intercept 0.295 0.0150 19.692 <0.001

Mod. disturbed 0.136 0.0221 6.154 <0.001

High. disturbed -0.092 0.0478 -1.929 0.0536

Qm is the test statistic for the Omnibus (Wald-type) test of moderator significance

Qe is the test statistic for the test of residual (unexplained) heterogeneity

P-value: <0.001***, <0.01** , <0.05*

218

Table S3. Summary statistics of best-fitted models ranked by AIC values relating

duration of study and disturbance level to effect size (correlation of species richness)

Model Estimate Std error Z-value P-value Goodness-of-fit

statistics Model parameter(s)

Effect size ~ year * disturb

R2 0.244

AIC 1339.345

∆AIC 0.00

Qe 4486.684**

Qm 251.761**

Intercept 0.492 0.0486 10.122 <0.001

Mod. disturbed -0.341 0.0615 -5.543 <0.001

High. disturbed 0.117 0.0710 1.649 0.0993

Year -0.048 0.00650 -7.386 <0.001

Mod. disturbed : year 0.0754 0.00960 7.868 <0.001

High. disturbed : year 0.0180 0.0190 0.949 0.3426

Effect size ~ year + disturb

R2 0.204

AIC 1397.548

∆AIC 58.203

Qe 4581.288**

Qm 181.862**

Intercept 0.257 0.0373 6.904 <0.001

Mod. disturbed 0.0841 0.0291 2.889 0.0039

High. disturbed 0.311 0.0349 8.892 <0.001

Year -0.0143 0.00470 -3.038 0.0024

Effect size ~ disturb

R2 0.161

AIC1499.684

∆AIC 160.339

Qe 5207.362**

Qm 149.462**

Intercept 0.170 0.0186 9.116 <0.001

Mod. disturbed 0.124 0.0277 4.487 <0.001

High. disturbed 0.360 0.0295 12.201 <0.001

Effect size ~ year

R2 0.0806

AIC 2172.269

∆AIC 832.924

Qe 26009.651**

Qm 97.490*

Intercept 0.549 0.0203 27.049 <0.001

Year -0.0353 0.00360 -9.874 <0.001

Qm is the test statistic for the Omnibus (Wald-type) test of moderator significance

Qe is the test statistic for the test of residual (unexplained) heterogeneity

P-value: <0.001***, <0.01** , <0.05*

219

Table S4. Summary statistics of best-fitted models ranked by AIC values relating

duration of study and disturbance level to effect size (correlation of species

composition)

Model Estimate Std error Z-value P-value Goodness-of-fit

statistics Model parameter(s)

Effect size ~ year * disturb

R2 0.212

AIC 995.840

∆AIC 0.00

Qe 3054.446**

Qm 128.943**

Intercept 0.614 0.0425 14.425 <0.001

Mod. disturbed -0.259 0.0504 -5.126 <0.001

High. disturbed -0.912 0.143 -6.389 <0.001

Year -0.0784 0.0101 -7.755 <0.001

Mod. disturbed : year 0.114 0.0143 7.980 <0.001

High. disturbed : year 0.434 0.0956 4.538 <0.001

Effect size ~ year + disturb

R2 0.0942

AIC 1067.518

∆AIC 71.678

Qe 3335.156**

Qm 44.260**

Intercept 0.3778 0.0333 11.351 <0.001

Mod. disturbed 0.0907 0.0271 3.343 <0.001

High. disturbed -0.186 0.0626 -2.962 0.0031

Year -0.0182 0.00740 -2.448 0.0144

Effect size ~ disturb

R2 0.0773

AIC 1085.475

∆AIC 89.635

Qe 3538.691**

Qm 39.412**

Intercept 0.307 0.0166 18.485 <0.001

Mod. disturbed 0.128 0.0241 5.315 <0.001

High. disturbed -0.139 0.0607 -2.280 0.0226

Effect size ~ year

R2 0.0104

AIC 1112.347

∆AIC 116.507

Qe 3622.267**

Qm 3.307

Intercept 0.386 0.0205 18.846 <0.001

Year -0.0102 0.00560 -1.819 0.0690

Qm is the test statistic for the Omnibus (ANOVA) test of moderator significance

Qe is the test statistic for the test of residual (unexplained) heterogeneity

P-value: <0.001 ***, <0.01 ** , <0.05 *

220

Appendix I. Accepted abstract for ICE 2016 XXV

Identifying surrogates for conserving insect diversity in human-

modified landscapes in south-eastern Australia

Ding Li Yong1 & Philip S. Barton1

1Fenner School of Environment and Society, the Australian National University

Abstract

Landscape modification has led to significant loss and degradation of natural

habitats worldwide, with associated declines in biodiversity. Despite this, there

remains limited knowledge on the diversity patterns of many taxonomic groups

across different landscapes and at different spatiotemporal scales, especially for

speciose taxa like insects. Discrepancies in the effort and resources needed to collect

data on different taxonomic groups further accentuate this problem, thereby

hampering the effective conservation of insect communities. Additionally, the

difficulty in identifying invertebrates necessitates the use of robust surrogate

measures that can account for insect diversity in conservation assessments. In our

study, we aimed to identify such surrogates for insect diversity across a human-

modified landscape in southeastern Australia. We asked if birds, mammals, or

habitat structural surrogates could act as reliable indicators of bee diversity in a

complex and highly fragmented landscape. Using vane traps, our surveys yielded 35

species of native bees of more than 1,700 individuals, notably halictid (e.g.

Lasioglossum, Sphecodoides spp.), megachilid (e.g. Megachile spp.) and apid bees

(e.g. Amegilla, Exoneura spp.). Bee communities were most diverse in remnant

Eucalypt woodland surrounded by agriculture, and least so in pine monoculture.

Our preliminary analyses, which adopted a multivariate approach, found that bird

and mammal diversity were poorly concordant with native bee diversity. However,

we found that a number of habitat structural variables significantly explained bee

richness and composition. We suggest these habitat structure can be better

developed as surrogates for native bee communities to facilitate the conservation of

this ecologically important insect group.

221

Appendix II. List of cross-taxonomic surrogacy papers compiled for meta-analysis,

classified by continent.

Global studies

McInnes, L., Jones, F. A., Orme, C. D. L., Sobkowiak, B., Barraclough, T. G., Chase,

M. W., Govaerts, R., Soltis, D. E., & Savolainen, V. (2013). Do global diversity

patterns of vertebrates reflect those of monocots? PloS One, 8(5), e56979.

McKnight, M. W., White, P. S., McDonald, R. I., Lamoreux, J. F., Sechrest, W.,

Ridgely, R. S., & Stuart, S. N. (2007). Putting beta-diversity on the map: broad-

scale congruence and coincidence in the extremes. PLoS Biology, 5(10), e272.

Mittermeier, R. A., Mittermeier, C. G., Brooks, T. M., Pilgrim, J. D., Konstant, W. R.,

Da Fonseca, G. A., & Kormos, C. (2003). Wilderness and biodiversity conservation.

Proceedings of the National Academy of Sciences U.S.A., 100(18), 10309-10313.

Pearson, D. L., & Cassola, F. (1992). World‐Wide Species Richness Patterns of Tiger

Beetles (Coleoptera: Cicindelidae): Indicator Taxon for Biodiversity and

Conservation Studies. Conservation Biology, 6(3), 376-391.

Pearson, D. L., & Carroll, S. S. (1998). Global patterns of species richness: spatial

models for conservation planning using bioindicator and precipitation data.

Conservation biology, 12(4), 809-821.

Schall, J. J., & Pianka, E. R. (1978). Geographical trends in numbers of species.

Science, 201(4357), 679-686.

Schuldt, A., Wang, Z., Zhou, H., & Assmann, T. (2009). Integrating highly diverse

invertebrates into broad‐scale analyses of cross‐taxon congruence across the

Palaearctic. Ecography, 32(6), 1019-1030.

Schmit, J. P., Mueller, G. M., Leacock, P. R., Mata, J. L., & Huang, Y. (2005).

Assessment of tree species richness as a surrogate for macrofungal species

richness. Biological Conservation, 121(1), 99-110.

Sisk, T. D., Launer, A. E., Switky, K. R., & Ehrlich, P. R. (1994). Identifying extinction

threats. BioScience, 44(9), 592-604.

Europe

Anderson, A., McCormack, S., Helden, A., Sheridan, H., Kinsella, A., & Purvis, G.

(2011). The potential of parasitoid Hymenoptera as bioindicators of arthropod

diversity in agricultural grasslands. Journal of Applied Ecology, 48(2), 382-390.

222

Araújo, M. B., Densham, P. J., & Williams, P. H. (2004). Representing species in

reserves from patterns of assemblage diversity. Journal of Biogeography, 31(7),

1037-1050.

Báldi, A. (2003). Using higher taxa as surrogates of species richness: a study based on

3700 Coleoptera, Diptera, and Acari species in Central-Hungarian reserves. Basic

and Applied Ecology, 4(6), 589-593.

Beck, J., Pfiffner, L., Ballesteros‐Mejia, L., Blick, T., & Luka, H. (2013). Revisiting the

indicator problem: can three epigean arthropod taxa inform about each other's

biodiversity? Diversity and Distributions, 19(7), 688-699.

Benayas, J. M. R., & de la Montaña, E. (2003). Identifying areas of high-value

vertebrate diversity for strengthening conservation. Biological Conservation,

114(3), 357-370.

Billeter, R., Liira, J., Bailey, D., Bugter, R., Arens, P., Augenstein, I., Aviron, S.,

Baudry, J., Bukacek, R., Burel, F., Cerny, M., De Blust, G., De Cock, R., Diekotter,

T., Dietz, H., Dirksen, J., Dormann, C., Durka, W., Frenzel, M., Hamersky, R.,

Hendrickx, F., Herzog, F., Klotz, S., Koolstra, B., Lausch, A., Le Couer, D.,

Maelfait, J. P., Opdam, P., Roubalova, M., Schermann, A., Schermann, N.,

Schmidt, T., Schweiger, O., Smulders, M. J. M., Speelmans, M., Simova, P.,

Verboom, J., Van Wingerden, W. K. R. E., Zobel, M., & Edwards, P. J. (2008).

Indicators for biodiversity in agricultural landscapes: a pan‐European study.

Journal of Applied Ecology, 45(1), 141-150.

Brändle, M., Amarell, U., Auge, H., Klotz, S., & Brandl, R. (2001). Plant and insect

diversity along a pollution gradient: understanding species richness across trophic

levels. Biodiversity & Conservation, 10(9), 1497-1511.

Bräuniger, C., Knapp, S., Kühn, I., & Klotz, S. (2010). Testing taxonomic and

landscape surrogates for biodiversity in an urban setting. Landscape and Urban

Planning, 97(4), 283-295.

Brose, U. (2003). Bottom-up control of carabid beetle communities in early

successional wetlands: mediated by vegetation structure or plant diversity?

Oecologia, 135(3), 407-413.

Buse, J., & Griebeler, E. M. (2012). Determinants and congruence of species richness

patterns across multiple taxonomic groups on a regional scale. International

Journal of Zoology, 2012, 297657.

Chiarucci, A., D'auria, F., De Dominicis, V., Laganà, A., Perini, C., & Salerni, E.

(2005). Using vascular plants as a surrogate taxon to maximize fungal species

richness in reserve design. Conservation Biology, 19(5), 1644-1652.

223

Chiarucci, A., D’auria, F., & Bonini, I. (2007). Is vascular plant species diversity a

predictor of bryophyte species diversity in Mediterranean forests? Biodiversity and

Conservation, 16(2), 525-545.

Dauber, J., Hirsch, M., Simmering, D., Waldhardt, R., Otte, A., & Wolters, V. (2003).

Landscape structure as an indicator of biodiversity: matrix effects on species

richness. Agriculture, Ecosystems & Environment, 98(1), 321-329.

Dirilgen, T., Arroyo, J., Dimmers, W. J., Faber, J., Stone, D., da Silva, P. M., Carvalho,

F., Schmelz, R., Griffiths, B. S., Francisco, R., Creamer, R. E., Sousa, J., & Bolger, T.

(2016). Mite community composition across a European transect and its

relationships to variation in other components of soil biodiversity. Applied Soil

Ecology, 97, 86-97.

Djupström, L. B., Perhans, K., Weslien, J., Schroeder, L. M., Gustafsson, L., &

Wikberg, S. (2010). Co-variation of lichens, bryophytes, saproxylic beetles and

dead wood in Swedish boreal forests. Systematics and Biodiversity, 8(2), 247-256.

Duelli, P., & Obrist, M. K. (1998). In search of the best correlates for local organismal

biodiversity in cultivated areas. Biodiversity and Conservation, 7(3), 297-309.

Dynesius, M., & Zinko, U. (2006). Species richness correlations among primary

producers in boreal forests. Diversity and Distributions, 12(6), 703-713.

Ekroos, J., Kuussaari, M., Tiainen, J., Heliölä, J., Seimola, T., & Helenius, J. (2013).

Correlations in species richness between taxa depend on habitat, scale and

landscape context. Ecological Indicators, 34, 528-535.

Essl, F., Dullinger, S., Plutzar, C., Willner, W., & Rabitsch, W. (2011). Imprints of

glacial history and current environment on correlations between endemic plant

and invertebrate species richness. Journal of Biogeography, 38(3), 604-614.

Estrada, A., Real, R., & Vargas, J. M. (2011). Assessing coincidence between priority

conservation areas for vertebrate groups in a Mediterranean hotspot. Biological

Conservation, 144(3), 1120-1129.

Fattorini, S. (2010). Biotope prioritisation in the Central Apennines (Italy): species

rarity and cross-taxon congruence. Biodiversity and Conservation, 19(12), 3413-

3429.

Fattorini, S., Dennis, R. L., & Cook, L. M. (2011). Conserving organisms over large

regions requires multi-taxa indicators: one taxon’s diversity-vacant area is another

taxon’s diversity zone. Biological Conservation, 144(5), 1690-1701.

Finch, O. D., & Löffler, J. (2010). Indicators of species richness at the local scale in an

alpine region: a comparative approach between plant and invertebrate taxa.

Biodiversity and Conservation, 19(5), 1341-1352.

224

Gaspar, C., Gaston, K. J., & Borges, P. A. (2010). Arthropods as surrogates of diversity

at different spatial scales. Biological Conservation, 143(5), 1287-1294.

Giorgini, D., Giordani, P., Casazza, G., Amici, V., Mariotti, M. G., & Chiarucci, A.

(2015). Woody species diversity as predictor of vascular plant species diversity in

forest ecosystems. Forest Ecology and Management, 345, 50-55.

Grill, A., Knoflach, B., Cleary, D. F., & Kati, V. (2005). Butterfly, spider, and plant

communities in different land-use types in Sardinia, Italy. Biodiversity and

Conservation, 14(5), 1281-1300.

Guareschi, S., Abellán, P., Laini, A., Green, A. J., Sánchez-Zapata, J. A., Velasco, J., &

Millán, A. (2015). Cross-taxon congruence in wetlands: assessing the value of

waterbirds as surrogates of macroinvertebrate biodiversity in Mediterranean

Ramsar sites. Ecological Indicators, 49, 204-215.

Hansson, L. (1997). Environmental determinants of plant and bird diversity in

ancient oak-hazel woodland in Sweden. Forest Ecology and Management, 91(2),

137-143.

Hawkins, B. A., & Pausas, J. G. (2004). Does plant richness influence animal

richness? the mammals of Catalonia (NE Spain). Diversity and Distributions,

10(4), 247-252.

Hayes, M., Boyle, P., Moran, J., & Gormally, M. (2015). Assessing the biodiversity

value of wet grasslands: can selected plant and insect taxa be used as rapid

indicators of species richness at a local scale? Biodiversity and Conservation,

24(10), 2535-2549.

Irwin, S., Pedley, S. M., Coote, L., Dietzsch, A. C., Wilson, M. W., Oxbrough, A.,

Sweeney, O., Moore, K. M., Martin R., Kelly, D. L., Mitchell, F. J., Kelly, T. C., &

O’Halloran, J. (2014). The value of plantation forests for plant, invertebrate and

bird diversity and the potential for cross-taxon surrogacy. Biodiversity and

Conservation, 23(3), 697-714.

Jeanneret, P., Schüpbach, B., Pfiffner, L., Herzog, F., & Walter, T. (2003). The Swiss

agri-environmental programme and its effects on selected biodiversity indicators.

Journal for Nature Conservation, 11(3), 213-220.

Jonsson, B. G., & Jonsell, M. (1999). Exploring potential biodiversity indicators in

boreal forests. Biodiversity and conservation, 8(10), 1417-1433.

Kati, V., Devillers, P., Dufrêne, M., Legakis, A., Vokou, D., & Lebrun, P. (2004).

Testing the value of six taxonomic groups as biodiversity indicators at a local

scale. Conservation biology, 18(3), 667-675.

Keith, A. M., Boots, B., Hazard, C., Niechoj, R., Arroyo, J., Bending, G. D., Bolger, T.,

Breen, J., Clipson, N., Doohan, F. M., Griffin, C. T., & Schmidt, O. (2012). Cross-

225

taxa congruence, indicators and environmental gradients in soils under

agricultural and extensive land management. European Journal of Soil Biology, 49,

55-62.

Kivinen, S., Luoto, M., Kuussaari, M., & Helenius, J. (2006). Multi‐species richness of

boreal agricultural landscapes: effects of climate, biotope, soil and geographical

location. Journal of Biogeography, 33(5), 862-875.

Koch, B., Edwards, P. J., Blanckenhorn, W. U., Buholzer, S., Walter, T., Wüest, R. O.,

& Hofer, G. (2013). Vascular plants as surrogates of butterfly and grasshopper

diversity on two Swiss subalpine summer pastures. Biodiversity and Conservation,

22(6-7), 1451-1465.

Lüscher, G., Schneider, M. K., Turnbull, L. A., Arndorfer, M., Bailey, D., Herzog, F.,

Pointereau, P., Richner, N., & Jeanneret, P. (2014). Appropriate metrics to inform

farmers about species diversity. Environmental Science & Policy, 41, 52-62.

Maccherini, S., Bacaro, G., Favilli, L., Piazzini, S., Santi, E., & Marignani, M. (2009).

Congruence among vascular plants and butterflies in the evaluation of grassland

restoration success. Acta Oecologica, 35(2), 311-317.

Maes, D., Bauwens, D., De Bruyn, L., Anselin, A., Vermeersch, G., Van Landuyt, W.,

De Knijf, G., & Gilbert, M. (2005). Species richness coincidence: conservation

strategies based on predictive modelling. Biodiversity & Conservation, 14(6), 1345-

1364.

Mandelik, Y., Dayan, T., Chikatunov, V., & Kravchenko, V. (2012). The relative

performance of taxonomic vs. environmental indicators for local biodiversity

assessment: a comparative study. Ecological Indicators, 15(1), 171-180.

Marini, L., Fontana, P., Battisti, A., & Gaston, K. J. (2009). Agricultural management,

vegetation traits and landscape drive orthopteran and butterfly diversity in a

grassland–forest mosaic: a multi‐scale approach. Insect Conservation and

Diversity, 2(3), 213-220.

Mazaris, A. D., Kallimanis, A. S., Sgardelis, S. P., & Pantis, J. D. (2008). Does higher

taxon diversity reflect richness of conservation interest species? the case for birds,

mammals, amphibians, and reptiles in Greek protected areas. Ecological

Indicators, 8(5), 664-671.

Myšák, J., & Horsák, M. (2014). Biodiversity surrogate effectiveness in two habitat

types of contrasting gradient complexity. Biodiversity and Conservation, 23(5),

1133-1156.

Nordén, B., Paltto, H., Götmark, F., & Wallin, K. (2007). Indicators of biodiversity,

what do they indicate?–Lessons for conservation of cryptogams in oak-rich forest.

Biological Conservation, 135(3), 369-379.

226

Oertli, S., Müller, A., Steiner, D., Breitenstein, A., & Dorn, S. (2005). Cross-taxon

congruence of species diversity and community similarity among three insect taxa

in a mosaic landscape. Biological Conservation, 126(2), 195-205.

Pautasso, M., & Powell, G. (2009). Aphid biodiversity is positively correlated with

human population in European countries. Oecologia, 160(4), 839-846.

Pearman, P. B., & Weber, D. (2007). Common species determine richness patterns in

biodiversity indicator taxa. Biological Conservation, 138(1), 109-119.

Pérez-Fuertes, O., García-Tejero, S., Hidalgo, N. P., Mateo-Tomás, P., Cuesta-Segura,

A. D., & Olea, P. P. (2016). Testing the effectiveness of surrogates for assessing

biological diversity of arthropods in cereal agricultural landscapes. Ecological

Indicators, 67, 297-305.

Sabatini, F. M., Burrascano, S., Azzella, M. M., Barbati, A., De Paulis, S., Di Santo, D.,

Facioni, L., Giuliarelli, D., Lombardi, F., Maggi, O., Mattioli, W., Parisi, F.,

Persiani, A., & Ravera, S. (2016). One taxon does not fit all: Herb-layer diversity

and stand structural complexity are weak predictors of biodiversity in Fagus

sylvatica forests. Ecological Indicators, 69, 126-137.

Sætersdal, M., Gjerde, I., Blom, H. H., Ihlen, P. G., Myrseth, E. W., Pommeresche, R.,

Skartveit, J., Solhoy, T., & Aas, O. (2004). Vascular plants as a surrogate species

group in complementary site selection for bryophytes, macrolichens, spiders,

carabids, staphylinids, snails, and wood living polypore fungi in a northern forest.

Biological Conservation, 115(1), 21-31.

Salmela, J. (2011). The semiaquatic nematoceran fly assemblages of three wetland

habitats and concordance with plant species composition, a case study from

subalpine Fennoscandia. Journal of Insect Science, 11(35), 1-28.

Santi, E., Maccherini, S., Rocchini, D., Bonini, I., Brunialti, G., Favilli, L., Perini, C.,

Pezzo, F., Piazzini S., Rota, E., Salerni, E., & Chiarucci, A. (2010). Simple to

sample: vascular plants as surrogate group in a nature reserve. Journal for Nature

Conservation, 18(1), 2-11.

Sauberer, N., Zulka, K. P., Abensperg-Traun, M., Berg, H. M., Bieringer, G.,

Milasowszky, N., Moser, D., Plutzar, C., Pollheimer, M., Storck, C., Tröstl, R.,

Zechmeister, H., & Grabherr, G. (2004). Surrogate taxa for biodiversity in

agricultural landscapes of eastern Austria. Biological Conservation, 117(2), 181-190.

Schouten, M. A., Barendregt, A., Verweij, P. A., Kalkman, V. J., Kleukers, R. M. J. C.,

Lenders, H. J. R., & Siebel, H. N. (2010). Defining hotspots of characteristic species

for multiple taxonomic groups in the Netherlands. Biodiversity and Conservation,

19(9), 2517-2536.

227

Schwab, A., Dubois, D., Fried, P. M., & Edwards, P. J. (2002). Estimating the

biodiversity of hay meadows in north-eastern Switzerland on the basis of

vegetation structure. Agriculture, Ecosystems & Environment, 93(1), 197-209.

Steck, C. E., & Pautasso, M. (2008). Human population, grasshopper and plant

species richness in European countries. Acta Oecologica, 34(3), 303-310.

Taboada, Á., Tárrega, R., Calvo, L., Marcos, E., Marcos, J. A., & Salgado, J. M. (2010).

Plant and carabid beetle species diversity in relation to forest type and structural

heterogeneity. European Journal of Forest Research, 129(1), 31.

Tonkin, J. D., Stoll, S., Jähnig, S. C., & Haase, P. (2016). Anthropogenic land-use

stress alters community concordance at the river-riparian interface. Ecological

Indicators, 65, 133-141.

Vera, P., Sasa, M., Encabo, S. I., Barba, E., Belda, E. J., & Monrós, J. S. (2011). Land use

and biodiversity congruences at local scale: applications to conservation

strategies. Biodiversity and Conservation, 20(6), 1287-1317.

Vessby, K., Söderström, B. O., Glimskär, A., & Svensson, B. (2002). Species‐Richness

Correlations of Six Different Taxa in Swedish Seminatural Grasslands.

Conservation Biology, 16(2), 430-439.

Weibull, A. C., Östman, Ö., & Granqvist, Å. (2003). Species richness in

agroecosystems: the effect of landscape, habitat and farm management.

Biodiversity and conservation, 12(7), 1335-1355.

Williams, P. H., & Gaston, K. J. (1998). Biodiversity indicators: graphical techniques,

smoothing and searching for what makes relationships work. Ecography, 21(5),

551-560.

Asia

Axmacher, J. C., Liu, Y., Wang, C., Li, L., & Yu, Z. (2011). Spatial α-diversity patterns

of diverse insect taxa in Northern China: Lessons for biodiversity conservation.

Biological Conservation, 144(9), 2362-2368.

Beck, J. A. N., Schulze, C. H., Linsenmair, K. E., & Fiedler, K. (2002). From forest to

farmland: diversity of geometrid moths along two habitat gradients on Borneo.

Journal of Tropical Ecology, 18(01), 33-51.

Chettri, N. (2010). Cross-taxon congruence in a trekking corridor of Sikkim

Himalayas: Surrogate analysis for conservation planning. Journal for Nature

Conservation, 18(2), 75-88.

Duan, M., Liu, Y., Yu, Z., Baudry, J., Li, L., Wang, C., & Axmacher, J. C. (2016).

Disentangling effects of abiotic factors and biotic interactions on cross-taxon

228

congruence in species turnover patterns of plants, moths and beetles. Scientific

Reports, 6, 23511.

Gillison, A. N., Bignell, D. E., Brewer, K. R., Fernandes, E. C., Jones, D. T., Sheil, D.,

May, P. H., Watt, A. D., Constantino, R., Couto, E. G., Hairiah, K., Jepson, P.,

Kartono, A. P., Maryanto, I., Neto, G. G., van Noordwijk, M., Silveira, E. A., Susilo,

F., Vosti, S. A., & Nunes, P. C. (2013). Plant functional types and traits as

biodiversity indicators for tropical forests: two biogeographically separated case

studies including birds, mammals and termites. Biodiversity and Conservation,

22(9), 1909-1930.

Kessler, M., Abrahamczyk, S., Bos, M., Buchori, D., Putra, D. D., Gradstein, S. R.,

Hohn, P., Kluge, J., Orend, F., Pitopang, R., Saleh, S. Schulze, C. H., Sporn, S. G.,

Steffan-Dewenter, I., Tjitrosoedirdjo, S. S., & Tscharntke, T. (2011). Cost‐

effectiveness of plant and animal biodiversity indicators in tropical forest and

agroforest habitats. Journal of Applied Ecology, 48(2), 330-339.

Kitahara, M., Yumoto, M., & Kobayashi, T. (2008). Relationship of butterfly diversity

with nectar plant species richness in and around the Aokigahara primary

woodland of Mount Fuji, central Japan. Biodiversity and Conservation, 17(11), 2713-

2734.

Kunte, K., Joglekar, A., Utkarsh, G., & Padmanabhan, P. (1999). Patterns of butterfly,

bird and tree diversity in the Western Ghats. Current Science, 577-586.

Pawar, S. S., Birand, A. C., Ahmed, M. F., Sengupta, S., & Raman, T. R. (2007).

Conservation biogeography in North‐east India: hierarchical analysis of cross‐

taxon distributional congruence. Diversity and Distributions, 13(1), 53-65.

Qian, H. (2007). Relationships between plant and animal species richness at a

regional scale in China. Conservation Biology, 21(4), 937-944.

Qian, H., & Kissling, W. D. (2010). Spatial scale and cross‐taxon congruence of

terrestrial vertebrate and vascular plant species richness in China. Ecology, 91(4),

1172-1183.

Schulze, C. H., Waltert, M., Kessler, P. J., Pitopang, R., Veddeler, D., Mühlenberg,

M., Gradstein, S. R., Leuschner, C., Steffan-Dewenter, I., & Tscharntke, T. (2004).

Biodiversity indicator groups of tropical land-use systems: comparing plants,

birds, and insects. Ecological Applications, 14(5), 1321-1333.

van Weerd, M., & De Haes, H. A. U. (2010). Cross-taxon congruence in tree, bird and

bat species distributions at a moderate spatial scale across four tropical forest

types in the Philippines. Biodiversity and Conservation, 19(12), 3393-3411.

229

Wang, Y., Bao, Y., Yu, M., Xu, G., & Ding, P. (2010). Nestedness for different reasons:

the distributions of birds, lizards and small mammals on islands of an inundated

lake. Diversity and distributions, 16(5), 862-873.

Xu, H., Wu, J., Liu, Y., Ding, H., Zhang, M., Wu, Y., Xi, Q., & Wang, L. (2008).

Biodiversity congruence and conservation strategies: a national test. BioScience,

58(7), 632-639.

Zhang, K., Lin, S., Ji, Y., Yang, C., Wang, X., Yang, C., Wang, H., Jiang, H., Harrison,

R. D., & Yu, D. W. (2016). Plant diversity accurately predicts insect diversity in

two tropical landscapes. Molecular Ecology, 25(17), 4407-4419.

Zhao, S., Fang, J., Peng, C., & Tang, Z. (2006). Relationships between species richness

of vascular plants and terrestrial vertebrates in China: analyses based on data of

nature reserves. Diversity and Distributions, 12(2), 189-194.

Africa and Madagascar

Chambers, B. Q., & Samways, M. J. (1998). Grasshopper response to a 40-year

experimental burning and mowing regime, with recommendations for

invertebrate conservation management. Biodiversity & Conservation, 7(8), 985-

1012.

Decher, J., & Bahian, L. K. (1999). Diversity and structure of terrestrial small

mammal communities in different vegetation types on the Accra Plains of Ghana.

Journal of Zoology, 247(3), 395-408.

Kissling, W. D., Field, R., & Böhning‐Gaese, K. (2008). Spatial patterns of woody

plant and bird diversity: functional relationships or environmental effects? Global

Ecology and Biogeography, 17(3), 327-339.

Lewin, A., Feldman, A., Bauer, A. M., Belmaker, J., Broadley, D. G., Chirio, L., Itescu,

Y., LeBreton, M., Maza, E., Meirte, D., Nagy, Z. T., Novosolov, M., Roll, U.,

Tallowin, O., Trape, J., Vidan, E., & Meiri, S. (2016). Patterns of species richness,

endemism and environmental gradients of African reptiles. Journal of

Biogeography, 43(12), 2380-2390.

Lovell, S., Hamer, M., Slotow, R., & Herbert, D. (2007). Assessment of congruency

across invertebrate taxa and taxonomic levels to identify potential surrogates.

Biological Conservation, 139(1), 113-125.

Moore, J. L., Balmford, A., Brooks, T., Burgess, N. D., Hansen, L. A., Rahbek, C., &

Williams, P. H. (2003). Performance of Sub‐Saharan Vertebrates as Indicator

Groups for Identifying Priority Areas for Conservation. Conservation Biology, 17(1),

207-218.

230

Muldoon, K. M., & Goodman, S. M. (2015). Primates as predictors of mammal

community diversity in the forest ecosystems of Madagascar. PloS One, 10(9),

e0136787.

Tushabe, H., Kalema, J., Byaruhanga, A., Asasira, J., Ssegawa, P., Balmford, A.,

Davenport, T., Fjeldsa, J., Friis, I., Pain, D., Pomeroy, D., Williams, P., & Williams,

C. (2006). A nationwide assessment of the biodiversity value of Uganda's

important bird areas network. Conservation Biology, 20(1), 85-99.

North America

Blair, R. B. (1999). Birds and butterflies along an urban gradient: surrogate taxa for

assessing biodiversity? Ecological Applications, 9(1), 164-170.

Brown, J. W., & Opler, P. A. (1990). Patterns of butterfly species density in peninsular

Florida. Journal of Biogeography, 17(6), 615-622.

Carroll, S. S., & Pearson, D. L. (1998). Spatial modeling of butterfly species richness

using tiger beetles (Cicindelidae) as a bioindicator taxon. Ecological Applications,

8(2), 531-543.

Chase, M. K., Kristan, W. B., Lynam, A. J., Price, M. V., & Rotenberry, J. T. (2000).

Single species as indicators of species richness and composition in California

coastal sage scrub birds and small mammals. Conservation Biology, 14(2), 474-487.

Diffendorfer, J. E., Fleming, G. M., Duggan, J. M., Chapman, R. E., Rahn, M. E.,

Mitrovich, M. J., & Fisher, R. N. (2007). Developing terrestrial, multi-taxon indices

of biological integrity: an example from coastal sage scrub. Biological

Conservation, 140(1), 130-141.

Estrada, C. G., & Rodríguez‐Estrella, R. (2016). In the search of good biodiversity

surrogates: are raptors poor indicators in the Baja California Peninsula desert?

Animal Conservation, 19, 360-368.

Gagné, S. A., & Fahrig, L. (2011). Do birds and beetles show similar responses to

urbanization? Ecological Applications, 21(6), 2297-2312.

Joern, A. (2005). Disturbance by fire frequency and bison grazing modulate

grasshopper assemblages in tallgrass prairie. Ecology, 86(4), 861-873.

Jonas, J. L., Whiles, M. R., & Charlton, R. E. (2002). Aboveground invertebrate

responses to land management differences in a central Kansas grassland.

Environmental Entomology, 31(6), 1142-1152.

Kemp, W. P., Harvey, S. J., & O'neill, K. M. (1990). Patterns of vegetation and

grasshopper community composition. Oecologia, 83(3), 299-308.

231

Kerr, J. T. (1997). Species richness, endemism, and the choice of areas for

conservation. Conservation Biology, 11(5), 1094-1100.

Kerr, J. T., Sugar, A., & Packer, L. (2000). Indicator taxa, rapid biodiversity

assessment, and nestedness in an endangered ecosystem. Conservation Biology,

14(6), 1726-1734.

Kirkman, L. K., Smith, L. L., Quintana-Ascencio, P. F., Kaeser, M. J., Golladay, S. W.,

& Farmer, A. L. (2012). Is species richness congruent among taxa? Surrogacy,

complementarity, and environmental correlates among three disparate taxa in

geographically isolated wetlands. Ecological Indicators, 18, 131-139.

Perfecto, I., Mas, A., Dietsch, T., & Vandermeer, J. (2003). Conservation of

biodiversity in coffee agroecosystems: a tri-taxa comparison in southern Mexico.

Biodiversity & Conservation, 12(6), 1239-1252.

Peters, V. E., Campbell, K. U., Dienno, G., García, M., Leak, E., Loyke, C., Ogle, M.,

Steinly, B., & Crist, T. O. (2016). Ants and plants as indicators of biodiversity,

ecosystem services, and conservation value in constructed grasslands. Biodiversity

and Conservation, 25(8), 1481-1501.

Ricketts, T. H., Dinerstein, E., Olson, D. M., & Loucks, C. (1999). Who's where in

North America? Patterns of species richness and the utility of indicator taxa for

conservation. Bioscience, 49(5), 369-381.

Ricketts, T. H., Daily, G. C., & Ehrlich, P. R. (2002). Does butterfly diversity predict

moth diversity? Testing a popular indicator taxon at local scales. Biological

Conservation, 103(3), 361-370.

Rooney, R. C., & Bayley, S. E. (2012). Community congruence of plants, invertebrates

and birds in natural and constructed shallow open-water wetlands: Do we need to

monitor multiple assemblages? Ecological Indicators, 20, 42-50.

Rooney, R. C., & Azeria, E. T. (2015). The strength of cross‐taxon congruence in

species composition varies with the size of regional species pools and the

intensity of human disturbance. Journal of Biogeography, 42(3), 439-451.

Simonson, S. E., Opler, P. A., Stohlgren, T. J., & Chong, G. W. (2001). Rapid

assessment of butterfly diversity in a montane landscape. Biodiversity &

Conservation, 10(8), 1369-1386.

Shuey, J. A., Metzler, E. H., & Tungesvick, K. (2012). Moth communities correspond

with plant communities in Midwestern (Indiana, USA) sand prairies and oak

barrens and their degradation endpoints. The American Midland Naturalist,

167(2), 273-284.

232

Su, J. C., Debinski, D. M., Jakubauskas, M. E., & Kindscher, K. (2004). Beyond species

richness: Community similarity as a measure of cross‐taxon congruence for

coarse‐filter conservation. Conservation Biology, 18(1), 167-173.

Warman, L. D., Forsyth, D. M., Sinclair, A. R. E., Freemark, K., Moore, H. D., Barrett,

T. W., Pressey, R. L., & White, D. (2004). Species distributions, surrogacy, and

important conservation regions in Canada. Ecology letters, 7(5), 374-379.

Weaver, J. C. (1995). Indicator species and scale of observation. Conservation Biology,

9(4), 939-942.

Central and South America

Barlow, J., Gardner, T. A., Araujo, I. S., Ávila-Pires, T. C., Bonaldo, A. B., Costa, J. E.,

Esposito, M. C., Ferreira, L. V., Hawes, J., Hernandez, M. I. M., Hoogmoed, M. S.,

Leite, R. N., Lo, M. H. N. F., Malcolm, J. R., Martins, M. B., Mestre, L. A. M.,

Miranda-Santos, R., Nunes-Gutjahr, A. L., Overal, W. L., Parry, L., Peters, S. L.,

Ribeiro, M. A., da Silva, M. N. F., da Silva Motta, C., & Peres, C. A. (2007).

Quantifying the biodiversity value of tropical primary, secondary, and plantation

forests. Proceedings of the National Academy of Sciences U.S.A., 104(47), 18555-

18560.

Cabra-García, J., Bermúdez-Rivas, C., Osorio, A. M., & Chacón, P. (2012). Cross-taxon

congruence of α and β diversity among five leaf litter arthropod groups in

Colombia. Biodiversity and Conservation, 21(6), 1493-1508.

De Andrade, R. B., Barlow, J., Louzada, J., Mestre, L., Silveira, J., Vaz-de-Mello, F. Z.,

& Cochrane, M. A. (2014). Biotic congruence in humid tropical forests: a multi-

taxa examination of spatial distribution and responses to forest disturbance.

Ecological Indicators, 36, 572-581.

Gillison, A. N., Bignell, D. E., Brewer, K. R., Fernandes, E. C., Jones, D. T., Sheil, D.,

May, P. H., Watt, A. D., Constantino, R., Couto, E. G., Hairiah, K., Jepson, P.,

Kartono, A. P., Maryanto, I., Neto, G. G., van Noordwijk, M., Silveira, E. A., Susilo,

F., Vosti, S. A., & Nunes, P. C. (2013). Plant functional types and traits as

biodiversity indicators for tropical forests: two biogeographically separated case

studies including birds, mammals and termites. Biodiversity and Conservation,

22(9), 1909-1930.

Goehring, D. M., Daily, G. C., & Şekerçioglu, Ç. H. (2002). Distribution of ground-

dwelling arthropods in tropical countryside habitats. Journal of Insect

Conservation, 6(2), 83-91.

233

Hilt, N., & Fiedler, K. (2006). Arctiid moth ensembles along a successional gradient

in the Ecuadorian montane rain forest zone: how different are subfamilies and

tribes? Journal of Biogeography, 33(1), 108-120.

Landeiro, V. L., Bini, L. M., Costa, F. R., Franklin, E., Nogueira, A., de Souza, J. L.,

Moraes, J., & Magnusson, W. E. (2012). How far can we go in simplifying

biomonitoring assessments? An integrated analysis of taxonomic surrogacy,

taxonomic sufficiency and numerical resolution in a megadiverse region.

Ecological Indicators, 23, 366-373.

Leal, I. R., Bieber, A. G. D., Tabarelli, M., & Andersen, A. N. (2010). Biodiversity

surrogacy: indicator taxa as predictors of total species richness in Brazilian

Atlantic forest and Caatinga. Biodiversity and Conservation, 19(12), 3347-3360.

Loyola, R. D., Kubota, U., & Lewinsohn, T. M. (2007). Endemic vertebrates are the

most effective surrogates for identifying conservation priorities among Brazilian

ecoregions. Diversity and Distributions, 13(4), 389-396.

Mandl, N., Lehnert, M., Kessler, M., & Gradstein, S. R. (2010). A comparison of alpha

and beta diversity patterns of ferns, bryophytes and macrolichens in tropical

montane forests of southern Ecuador. Biodiversity and Conservation, 19(8), 2359-

2369.

Oldekop, J. A., Bebbington, A. J., Truelove, N. K., Tysklind, N., Villamarín, S., &

Preziosi, R. F. (2012). Co-occurrence patterns of common and rare leaf-litter frogs,

epiphytic ferns and dung beetles across a gradient of human disturbance. PloS

one, 7(6), e38922.

Pomara, L. Y., Ruokolainen, K., Tuomisto, H., & Young, K. R. (2012). Avian

Composition Co‐varies with Floristic Composition and Soil Nutrient

Concentration in Amazonian Upland Forests. Biotropica, 44(4), 545-553.

Rompré, G., Douglas Robinson, W., Desrochers, A., & Angehr, G. (2007).

Environmental correlates of avian diversity in lowland Panama rain forests.

Journal of Biogeography, 34(5), 802-815.

Torrusio, S., Cigliano, M. M., & De Wysiecki, M. L. (2002). Grasshopper (Orthoptera:

Acridoidea) and plant community relationships in the Argentine pampas. Journal

of Biogeography, 29(2), 221-229.

Oceania (Australia and Pacific Islands)

Bennett, J. M., Kutt, A. S., Johnson, C. N., & Robson, S. K. A. (2009). Ants as

indicators for vertebrate fauna at a local scale: an assessment of cross-taxa

surrogacy in a disturbed matrix. Biodiversity and Conservation, 18(13), 3407-3419.

234

McMullan‐Fisher, S. J., Kirkpatrick, J. B., May, T. W., & Pharo, E. J. (2010). Surrogates

for macrofungi and mosses in reservation planning. Conservation Biology, 24(3),

730-736.

Moir, M. L., Brennan, K. E. C., Koch, J. M., Majer, J. D., & Fletcher, M. J. (2005).

Restoration of a forest ecosystem: the effects of vegetation and dispersal

capabilities on the reassembly of plant-dwelling arthropods. Forest Ecology and

Management, 217(2), 294-306.

Moritz, C., Richardson, K. S., Ferrier, S., Monteith, G. B., Stanisic, J., Williams, S. E.,

& Whiffin, T. (2001). Biogeographical concordance and efficiency of taxon

indicators for establishing conservation priority in a tropical rainforest biota.

Proceedings of the Royal Society of London B: Biological Sciences, 268(1479), 1875-

1881.

Oliver, I., Beattie, A. J., & York, A. (1998). Spatial fidelity of plant, vertebrate, and

invertebrate assemblages in multiple‐use forest in eastern Australia. Conservation

Biology, 12(4), 822-835.

Pharo, E. J., Beattie, A. J., & Binns, D. (1999). Vascular plant diversity as a surrogate

for bryophyte and lichen diversity. Conservation Biology, 13(2), 282-292.

Pharo, E. J., Beattie, A. J., & Pressey, R. L. (2000). Effectiveness of using vascular

plants to select reserves for bryophytes and lichens. Biological Conservation,

96(3), 371-378.

Powney, G. D., Grenyer, R., Orme, C. D. L., Owens, I. P. F., & Meiri, S. (2010). Hot,

dry and different: Australian lizard richness is unlike that of mammals,

amphibians and birds. Global Ecology and Biogeography, 19(3), 386-396.

Yong, D. L., Barton, P. S., Okada, S., Crane, M., & Lindenmayer, D. B. (2016). Birds as

surrogates for mammals and reptiles: Are patterns of cross-taxonomic

associations stable over time in a human-modified landscape? Ecological

Indicators, 69, 152-164.