Section IV: Special topics - USDA ARS · indicators from the core measurements described in Quick...

107

Section IV: Special topics

Riparian (Ch. 18)

Wildlife habitat (Ch. 20)

Fire (Ch. 22)

Livestock production (Ch. 19)

Vehicles/Recreation (Ch. 21)

Invasive species (Ch. 23)

Remote sensing (Ch. 25)State and transition models (p. Ch. 24)

GrasslandState

ShrublandState

Shrub-invadedGrassland

State

Pho

to b

y B

ob G

ibbe

nsP

hoto

by

Jayn

e B

elna

pP

hoto

by

Mik

e P

ella

nt

Pho

to b

y P

atri

ck S

have

rP

hoto

by

Dou

g B

urke

tt

Soil carbon (Ch. 26)

108

T he monitoring methods included in theprevious chapters of this manual wereselected because they generate indicators

relevant to the three key attributes: soil and sitestability, hydrologic function and biotic integrity.These attributes represent the foundation fornearly every land management objective,including livestock production, wildlife habitat,recreation and watershed protection. The first sixchapters, and the ninth chapter, of this sectionprovide additional guidance on how to adapt theseprotocols to address more specific managementand monitoring objectives.

Each of these chapters is organized into foursections: an introduction, a summary table,methods notes and additional resources. Notes areincluded only for methods that requiremodifications or for which there are additionalindicators that are not described in the methodschapters (Section II, Chs. 7 through 15). Theadditional resources portion describes printed andonline resources. When possible, local experts(NRCS, USFS, Extension, etc.) should be consulted,particularly for projects involving multipleobjectives in complex systems.

Each chapter addresses three strategies. Thefirst and simplest strategy is to calculate additionalindicators from the core measurements describedin Quick Start (Vol. I). Many of the measurementsincluded in this manual were selected in partbecause they can be used to easily generate a largenumber of indicators. For example, Line-pointintercept was selected instead of vegetationfrequency or density because it can be used togenerate cover and composition indicators, as wellas information on soil surface properties such asrock and lichen cover. Unlike Daubenmirequadrats, Line-point intercept data can easilygenerate vegetation structure indicators. Line-point intercept also can quantify ground cover inplant interspaces.

The second strategy involves making relativelysimple modifications or additions to the coremeasurements, such as adding height to the Line-point intercept measurements.

The third and most expensive strategy is toincorporate supplementary measurements.

Each of these strategies increases monitoringcosts. The first six chapters, and the ninth chapterof Section IV, include tables defining the relativepriority of each measurement for typicalapplications. These tables can be used, togetherwith the time estimates in Quick Start, to comparethe relative costs and benefits of eachmeasurement for the particular management ormonitoring objective. Because each situation isunique, these rankings should be used only as arough guide for selecting measurements.

The lists of additional resources are by nomeans complete. There are hundreds ofmonitoring guides available now and many moreare becoming available on the Internet. Most arespecific to particular uses or values, and most canbe adapted to and integrated with the flexiblemonitoring system described here.

Please note that there is potential for overlapamong the special topics. The first (Riparian) is atype of land. The next three (Livestock production,Wildlife habitat and Off-road vehicles) are mostcommonly thought of as land uses or values. Thefifth (Fire) is often applied as a management toolbut it, like the sixth (Invasive species), can also beviewed as a threat. It is increasingly common tofind that all six topics need to be addressedsimultaneously. For example, fire is used to controlinvasive species in riparian zones that aresimultaneously managed for livestock, wildlife,recreation and carbon sequestration. Theadvantage of using an integrated system is that thedata are relevant to all six topics. While the timeallocated to different measurements may varydepending on the relative importance of eachtopic, the basic structure should remain constant.

Chapters 24 and 25 provide a briefintroduction to state and transition models, andremote sensing. Both of these tools can beextremely useful in monitoring program designand data interpretation.

109

Chapter 18

Riparian

Important indicators in most riparian systemsare plant community composition andstructure. These can be monitored using a

combination of one or more of the followingmethods: Line-point intercept, Riparian channelvegetation survey, Belt transect and Tree density.Additional long-term monitoring methods canprovide more complete information onrelationships between changes in vegetation andchannel morphology.

Figure 18.1. Riparian vegetation along Rio Peñasco,New Mexico

.1.81elbaT .smetsysnairapirotsdohtemgnirotinomgniylpparofsenilediuG

tratSkciuQsnoitacifidoM

lanoitiddA lacipyT

)I.loV(stnemerusaem srotacidni ytiroirp

stniopotohP sotohplennahcedulcnI a/n hgiH

eniL tpecretnitniop-prep( *)lennahcotralucidne

thgiehedulcnI seY hgiH

tpecretnipaG enoN oN etaredoM

*tsetytilibatslioS enoN oN etaredoM

*tcesnarttleB enoN oN etaredoM

yratnemelppuS

)II.loV(stnemerusaem

*tsetnoitcapmoC enoN oN woL

*retemortlifnignir-elgniS enoN oN woL

noitcudorptnalP enoN oN woL

*ssenhcirseicepstnalP enoN oN woL

erutcurtsnoitategeV enoN oN etaredoM

*ytisnedeerT enoN oN etaredoM

yevrusnoitategevlennahcnairapiR enoN oN hgiH

eliforpyllugdnalennahcnairapiR enoN oN hgiH

.wolebsetoneesesaelP*

Riparian notesLine-point intercept. The Line-point interceptmethod can be used to effectively monitorchanges in cover and composition across thewidth of the riparian zone. Install transectsperpendicular to the channel with at least threetransects per area of interest. Set transect ends atleast 5 m (15 ft) outside the maximum potential

riparian zone. Transects should be extendedfurther in areas where the riparian area is expectedto expand. For extremely wide riparian areas,reduce the frequency of measurements along thetransect. Adding height measurements at eachpoint provides useful information on vegetationstructure. Use the Line-point Intercept with HeightData Form in Chapter 15 to record heightmeasurements.

110

In addition to the standard indicators,calculate the proportion of the transect covered bywoody riparian species, and average plant height.Standard deviation of height measurements andthe average number of species recorded at eachpoint are useful indicators of structural diversity.Changes in the width of riparian zones can bemonitored by noting where riparian vegetationbegins and ends along the transect. Line-pointintercept data also can be used for this purpose,but will often underestimate riparian zone width.

Soil stability test. The Soil stability test can be auseful indicator of changes in soil structuraldevelopment, but results are often difficult tointerpret in riparian systems. Interpretation of datais limited by the fact that soil texture often varieswidely within a riparian zone. Texture also maychange during the year as flood events depositnew material. Sediment deposition may result in anegative change in average soil stability followinga flood. However, sediment deposition by floods isoften a positive indicator of riparian zonerecovery, despite its initially low stability.

Belt transect and Tree density. Either of thesemethods can be used to monitor woody plantdensity by size class. Belt transects are moreappropriate for monitoring recruitment of newindividuals, while the Tree density method is moreapplicable for areas with a few highly dispersedindividuals. The Belt transect can be applied eitheralong the greenline (edge of the channel) or onthe Line-point intercept transect that crosses thechannel.

Compaction and Infiltration. High rates ofrecreational or grazing use, especially on moist orwet soil, can cause degradation of soil structure,including compaction. Where compaction appears

to have resulted in reduced infiltration, both theinfiltrometer and the penetrometer may be used.Infiltration measurements are usually low prioritybecause they are relatively time consuming (highcost-benefit ratio). Compaction test measurementsare relatively rapid, but the data are difficult tointerpret unless the measurements are made in soilwith the same moisture content each year. This ismore likely to be possible in arid ecosystems whenmeasurements can often be made following aperiod with no precipitation.

Plant species richness. Plant species richness canbe a valuable indicator of riparian recovery anddegradation. It is useful when biodiversity is amanagement objective. The method generally hasa low priority because a minimum estimate ofrichness can be calculated from the Line-pointintercept and Riparian channel vegetation survey.The method is also quite time consuming, addingsignificantly to costs in most cases.

Additional resourcesA large number of riparian monitoring systemshave been developed for perennial streams. Manysystems focus on specific stream characteristicsbelieved to be important for fish habitats,including water temperature and chemistry. Oneof the most widely applied riparian vegetationmethods is described in Winward (2000). Thismethod, like the Riparian channel vegetationsurvey, depends on identifying the greenline.Researchers are continuing to develop appropriatemethods for intermittent streams, washes andarroyos where the greenline is often difficult toidentify. Using aerial photography andvideography to monitor (Prichard et al. 1996) isbecoming increasingly popular, particularly wheredense vegetation and accessibility make groundmeasurements difficult or impossible.

Riparian

111

Chapter 19

Livestock production

Figure 19.1. Herding cattle in the ChihuahuanDesert.

L ong-term sustainability of livestockproduction in upland areas depends on thethree key ecosystem attributes: soil and site

stability, hydrologic function and biotic integrity.The Quick Start measurements should be adequatefor monitoring these attributes, except where thereis a specific problem such as compaction, or aconcern such as biodiversity (species richness).Where the flexibility exists to make short-termchanges in stocking rates or grazing patterns (e.g.,by moving water, salt blocks or supplementalfeed), conduct short-term monitoring (QuickStart).

.1.91elbaT yramirpehtsinoitcudorpkcotsevilerehwsmetsysotsdohtemgnirotinomgniylpparofsenilediuG .esu


lanoitiddA lacipyT


stniopotohP enoN a/n hgiH

tpecretnitniop-eniL enoN seY hgiH

tpecretnipaG enoN oN hgiH

tsetytilibatslioS enoN oN hgiH

tcesnarttleB enoN oN hgiH

yratnemelppuS


tsetnoitcapmoC enoN oN woL

retemortlifnignir-elgniS enoN oN woL

*noitcudorptnalP enoN oN hgiH

ssenhcirseicepstnalP enoN oN woL

erutcurtsnoitategeV enoN oN woL

ytisnedeerT enoN oN woL

yevrusnoitategevlennahcnairapiR enoN oN **woL

eliforpyllugdnalennahcnairapiR enoN oN woL.wolebsetoneesesaelP*

.hgihsiytiroirperehw,senoznairapirnitpecxE**

Livestock production notesPlant production. Annual forage production issometimes considered to be one of the mostimportant indicators for livestock management.This is generally calculated as part of total plantproduction. It can be used to plan annual stocking

rates. It is also a long-term indicator of changes inland status. Plant production is one of the mostdifficult and costly indicators to accuratelymonitor, especially in arid and semi-aridecosystems. It can also be difficult to interpret,particularly in areas with highly variableprecipitation.

Pho

to b

y B

ob G

ibbe

ns

112

An aboveground standing biomass of foragespecies is a useful short-term indicator, which canbe used to determine how many animals a pasturewill support for a particular period of time. It canbe calculated from the plant production datacollected by simply setting utilization to 0 and thegrowth adjustment factor to 1.0 (Rules 9 and 10 inChapter 9). Residual (standing) biomass, likeresidual cover, is also an excellent short-termindicator for determining when to removelivestock from a pasture.

Additional resourcesMost monitoring systems used by federal agenciesin the United States were designed to monitorlivestock grazing impacts on plant communitiesand, to a lesser extent, production. They ofteninclude a mixture of short-term indicators (such asstubble height and estimated utilization) and long-term indicators (such as similarity to a

hypothesized historic plant community). TheNatural Resources Conservation Service (NRCS)and Bureau of Land Management (BLM) bothcurrently rely on the NRCS National Range andPasture Handbook (USDA-NRCS 1997) formonitoring guidance. However, there is significantvariability at both the state and local level. TheUnited States Forest Service (USFS) has relied onthe Parker Three Step method for monitoring inmost regions, although other methods areincreasingly used. The basic approach is describedin Parker (1951). The local office should beconsulted to find out exactly how the method wasand is being applied in each forest. In addition tofederal handbooks, most state extension serviceshave developed and published rangelandmonitoring guides. Again, these generally focus oneffects of livestock grazing and include a mix ofshort- and long-term indicators. Contact yourlocal extension office or land-grant university forcurrent versions.

Livestock production

113

Chapter 20

Wildlife habitat

Important characteristics for wildlifemanagement are vegetation composition andstructure. These can be monitored using

augmented versions of the Line-point interceptand Belt transect methods, as well as adding acover pole or cover board measurement(Vegetation structure).

Every species has unique habitat requirements.These requirements may be poorly understood andthey can change during the year. Therefore, pleaseread the Wildlife habitat notes section below todetermine which combination of methods bestsuits your needs.

.1.02elbaT .evitcejbotnemeganamyramirpehtsitatibahefildliwnehwsdohtemgnirotinomgniylpparofsenilediuG


lanoitiddA lacipyT



*tpecretnitniop-eniL thgiehedulcnI seY hgiH

*tpecretnipaGthgiehedulcniyaM

tnemeriuqerseY etaredoM

*tsetytilibatslioS enoN oN woL

*tcesnarttleB enoN oN hgiH

yratnemelppuS




*noitcudorptnalP enoN oN etaredoM


*erutcurtsnoitategeV enoN oN hgiH

ytisnedeerT enoN oN hgiH




Wildlife habitat notesLine-point intercept. Line-point intercept can beused to assess plant composition. In savannas andother systems with widely scattered trees, it shouldbe supplemented with the Belt transect and/orTree density methods.

Where vertical vegetation structure is ofinterest, height estimates should be included for atleast every fifth point. Use the Line-point Interceptwith Height Data Form in Chapter 15. Line-pointintercept (with height) should be combined withCanopy Gap intercept to best understandvegetation structure.

Figure 20.1. Mule deer habitat.

Pho

to b

y D

oug

Bur

kett

114

In addition to the standard indicators,calculate the proportion of the line covered bywoody species and average height. The standarddeviation of the height measurements and theaverage number of species recorded at each pointare useful indicators of structural diversity. Thedistribution of species along a transect can beused, together with Gap intercept from the sametransect, to characterize individual vegetationpatches.

Gap intercept. Canopy Gap intercept is generallymore useful than Basal Gap intercept foraddressing horizontal vegetation structure withrespect to wildlife. Canopy Gap intercept shouldbe combined with Line-point intercept (withheight) to best understand vegetation structure.Basal Gap intercept may be used as a surrogate insystems where canopy cover is extremely dynamic.The standard Gap intercept indicators can be usedto estimate the proportion of area in which ananimal would be exposed to aerial predators or todirect sunlight. The standard indicator classes (25-50 cm, 51-100 cm, etc.) were selected based onerosion criteria. The proportion of land in evenlarger gaps (e.g., > 500 cm, approximately 15 ft)can be used to examine the extent to whichvegetation is clumped or dispersed.

The Gap intercept method can be modified toexamine gaps between tall clumps of vegetation byestablishing a minimum height or by recordingtwo separate gap types (greater than x cm and lessthan x cm tall) for canopy intercepts. Theindicator calculations and data forms are identicalto the standard technique.

Soil stability test. While not directly related tohabitat for most wildlife species, soil stability isessential to the sustainability of the system. It isalso an important indicator of the integrity of soilprocesses, including the activity of soil-dwellinganimals responsible for root and plant litterdecomposition.

Belt transect. Belt transects can be used to assessplant species composition in communitiescontaining widely scattered trees. Additional sizeclasses can be included for species measured withthe Belt transect to better estimate verticalvegetation structure (but see the Vegetationstructure method discussed below).

Plant production. Please see discussion in“Livestock production notes” in Chapter 19.

Plant species richness. Please see “Riparian”discussion in Chapter 18.

Vegetation structure (cover pole). Cover poles andboards are among the most widely used tools forcharacterizing habitat structure. The proportion ofthe pole that is obscured by vegetation in eachheight increment when viewed from a specifieddistance reflects the proportion of an animal thatwould be obstructed from view at that distance.

Additional resourcesWe found few generic resources for wildlife habitatmonitoring, although there are literally hundredsof protocols available for individual species. Awide range of literature does exist linkingvegetation structure to wildlife habitat, birddiversity, visual obstruction and production (e.g.,Robel 1970, Robel et al. 1970, and Harrell andFuhlendorf 2002).

If a particular species or group of species is ofconcern, try contacting a local wildlife biologist orsearching the Internet. Keywords that may assist inInternet searches include: foliage height diversity(FHD), vegetation structure, vertical structuraldiversity, wildlife habitat structure, cover pole,cover board and Robel Pole. Krebs (1998) lists anumber of techniques for measuring animalpopulations directly. Measuring and MonitoringPlant and Animal Populations (Elzinga et al. 2001)also has information on animal populationmonitoring, although the primary focus is onvegetation monitoring. Research and ManagementTechniques for Wildlife and Habits (1994) is anotherresource for wildlife habitat methods.

Wildlife habitat

115

Chapter 21

Off-road vehicle use and otherrecreational land uses

A reas impacted by off-road vehicles andother recreational land uses are oftencharacterized by linear surface

disturbances. While these disturbances can cover arelatively small portion of the landscape, theireffects on ecosystem function can be significant,especially in steeply sloping terrain and riparianzones. Recent research (Herrick et al. unpublisheddata) has shown that even a single pass of arelatively small vehicle can compact some soils,significantly reducing water infiltration and soilstability for extended periods of time.

.1.12elbaT .sesudnallanoitaercerrehtodnaesuelcihevdaor-ffootsdohtemgnirotinomgniylpparofsenilediuG


lanoitiddA lacipyTytiroirp)I.loV(stnemerusaem srotacidni


*tpecretnitniop-eniL nmuloclanifsa"kcart"ddA oN hgiH

*tpecretnipaG 3sa"kcart"ddA dr epytpag seY hgiH

tsetytilibatslioS kcartffo/noybyfitartS seY hgiH


yratnemelppuS


tsetnoitcapmoC kcartffo/noybyfitartS oN hgiH

retemortlifnignir-elgniS kcartffo/noybyfitartS oN etaredoM

noitcudorptnalP enoN oN woL





eliforpyllugdnalennahcnairapiR enoN oN woL

.wolebsetoneesesaelP*hgihsiytiroirperehw,senoznairapirnitpecxE**

Figure 21.1. Off-road vehicle trails north of Salt LakeCity.

Pho

to b

y Ja

yne

Bel

nap

116

Recreation

In order to make a sufficient number ofmeasurements in tracked areas, it may benecessary to pre-stratify soil measurements intoareas that both do and do not appear to be intracks. Randomly select locations for an equalnumber of soil measurements (Soil stability test,Compaction test and/or infiltration) in trackedand non-tracked areas. If this approach is used, itis essential that track intercepts be recorded on theLine-point Intercept or Gap Intercept Data Form sothat a weighted average can be calculated for eachsoil indicator.

Recreation notesLine-point intercept. Where vehicle tracks arerelatively distinct, the proportion of area theycover can be quantified by recording the numberof Line-point intercept points that fall on them.Use the Line-point Intercept with Height DataForm and change the “Height” column to “Track”(or add another column). The track cover estimateis likely to be less precise than other coverestimates, such as bare ground, because of thedifficulty in defining what constitutes a track.Observer ability varies, and tracks tend to be moreapparent early and late in the day. On some soils itmay be possible to define a minimum depthrequired for tracks to be recorded.

Gap intercept. The Gap intercept method can alsobe used to quantify the proportion of the areacovered by tracks. This is particularly useful inareas where tracks cross the transects relativelyinfrequently (e.g., less than five percent of the

transect). On the Gap Intercept Data Form simplyuse the last few columns of the “Basal Gapintercept” side of the page and record where eachtrack or contiguous set of tracks begins and endsalong the transect.

Belt transect. Belt transect measurements andother strategies to monitor invasive species (see“Invasive species,” Ch. 23) should be given highpriority due to the potential for vehicles totransport invasive species relatively large distances.It is important to train field workers to identify allspecies that could potentially invade a site, basedon soil and climate requirements, whether or notthe species is already present in the area.

Plant species richness. Please see the discussion in“Riparian notes” in Chapter 18.

Additional resourcesDavid Cole of the USFS Rocky Mountain Forestand Range Experiment Station has written anumber of publications on monitoringrecreational impacts. They are available on theUSFS websites (http://fsinfo.fs.fed.us/cgi-bin/gw/chameleon, http://www.srs.fs.fed.us/pubs/index.jsp or http://leopold.wilderness.net/pubs.cfm). Most of these focus on the effects ofhikers, campers and mountain bikers. RichardKnight of Colorado State University and othershave also published extensively on monitoringrecreational impacts. However, there aresurprisingly few protocols available for monitoringoff-road vehicle effects.

117

Chapter 22

Fire

T here are two general types of fire monitoring:fire risk and fire recovery. Fire riskmonitoring is a relatively well-developed

science based on estimates of fuel availability,vertical and horizontal continuity of fuel, moisturecontent and weather conditions. Fire riskmonitoring is not addressed here.

Fire recovery monitoring is generally initiatedfollowing fire. Where possible (e.g., prescribedburns), pre-fire baseline data should be collected atthe same time of year that monitoring will becontinued following fire. It is more important tomonitor at the same time of year before and afterthan to take measurements immediately followingthe fire.

Figure 22.1. Prescribed fire in an old world bluestem,sideoats grama, little bluestem and blueberry junipergrassland community.

.1.22elbaT .yrevocererif-tsopotsdohtemgnirotinomgniylpparofsenilediuG


lanoitiddA lacipyT



*tpecretnitniop-eniL enoN seY hgiH

*tpecretnipaGydoowesraocedulcniyaM

deddebmero/dnasirbedrettil

oN hgiH

*tsetytilibatslioS "cibohpordyh"sayfissalC oN hgiH


yratnemelppuS


*tsetnoitcapmoC enoN oN woL


*noitcudorptnalP enoN oN woL


*erutcurtsnoitategeV enoN oN etaredoM

*ytisnedeerT enoN oN hgiH-woL




Pho

to b

y P

at S

have

r

118

Fire

The most common post-fire and fire recoveryconcerns are runoff, erosion and regeneration ofthe plant community. Runoff and erosion areexpensive to measure directly. The Quick Startmethods and indicators reflect changes in plantcommunities and in the risk of runoff and erosion.Consequently, the basic measurements often canbe applied to post-fire recovery monitoring withrelatively little modification.

The method modifications described here arebased on experience from northern New Mexico ingrasslands invaded by piñon pine, juniper, oakand/or ponderosa pine. Additional modificationsmay be useful in other ecosystems.

Fire notesLine-point intercept. The Line-point intercept canbe applied with little modification. For savannasand woodlands with significant coarse woodydebris, it may be useful to split the woody litterclass (WL on the Line-point Intercept Data Form)into multiple size classes. Where short-termmortality estimates are required, the heightcolumn from the Line-point Intercept with HeightData Form (Ch. 15) can be changed to “Dead?”and used as a checkbox. However, mortality maybe more precisely quantified using the Belttransect method, especially for woody species.Differentiating between dead and live herbaceousplants is normally not recommended because ofthe high level of uncertainty associated with theseassessments. In addition, plant mortality is usuallymore accurately reflected in increased bare groundand reduced plant cover the following year.

Gap intercept. Gap intercept is one of the moreuseful measurements for monitoring post-firerecovery. It distinguishes between recoveryoccurring uniformly across a site, and recoveryconcentrated in dense vegetation patches. Someorganizations have modified the Gap interceptmethod to include embedded litter because of itsrole in helping to slow runoff. Embedded litter isassumed to have a similar effect on runoff as aplant base does. While this may be true forsystems in which litter is firmly anchored to the

soil by fungal mats, it is probably not appropriatein all cases. Coarse woody debris can also act like aplant base where it is in direct contact with thesoil surface.

Soil stability test. Prescribed burns rarely causeshort-term changes in soil stability. Stability canbegin to decline over time, however, if plantrecovery is slow. This is due to reduced root,fungal and litter inputs necessary for soil aggregateformation.

Intense fires where a large amount of fuel isburned at the soil surface can actually increase soilstability by making it hydrophobic. Unfortunately,because these surfaces repel water, they ultimatelyincrease erosion downslope because they increasesurface runoff. Other factors can contribute tohydrophobicity, including high fungalconcentrations. Hydrophobicity can be easilyquantified by recording the number of soilstability samples that float when they are placed inwater.

Belt transect and Tree density. The Belt transectand Tree density methods can be used to quantifymortality and recruitment by simply recording liveand dead individuals, and new seedlings, indifferent columns. The Belt transect method is alsouseful for monitoring the invasive plantpopulations after they have become established(see “Invasive species,” Chapter 23).

Compaction. Fire does not cause compaction.However, fire-fighting activities often do,especially when vehicles are driven off road. Inaddition to the burned area itself, firebreaks andaccess points for fire crews should be consideredfor inclusion in post-fire recovery monitoring.Where time permits, infiltration may also bemeasured.

Plant production. Please see the discussion in“Livestock production notes” (Chapter 19).

Plant species richness. Please see the discussion in“Riparian notes” (Chapter 18).

119

Fire

Vegetation structure. Vegetation structureindicators can be used as a relative indicator of thepresence of “ladder fuels” in savannas. Tallerherbaceous plant material and low branchesfacilitate the movement of ground fires into treecanopies.

Additional resourcesIn the past, fire recovery monitoring wastraditionally limited to photographs andoccasional quadrat or transect measurements.

Funding was rarely available for repeatedmeasurements, or to develop and test protocols.Increased interest in response to large burned areashas sparked the development of a large number ofmonitoring systems, many of which are becomingavailable on the Internet. Many of the systemsconsist of separate methods for each monitoringobjective (runoff, erosion, wildlife, vegetation,etc.). Where possible, the methods should becombined in order to limit costs associated withredundant measurements.

120

Chapter 23

Invasive species

I nvasive species may be the most important,ecologically sensitive and profitable singlefactor to monitor in many areas. The amount of

money that can be saved through early detectionof a new population can often exceed the currentvalue of the land. Unfortunately, establishmentcan be difficult to detect remotely, and it isimpossible to search every acre every year. Thefollowing protocol can be used to reducemonitoring costs while increasing the probabilityof early detection. It is based on rapid assessmentof nonpermanent plots in areas with a high risk ofinvasion.

The methods included in this manual can beused to address two objectives related to invasivespecies:(1) To monitor changes in invasive species after

they have become established (Belt transect forlow cover and Line-point intercept for highcover).

(2) To monitor changes in the susceptibility of asite to invasion (Line-point intercept and Gapintercept) where there is a high risk of seeddispersal, or it is known that invasive speciesalready exist in the seed bank.

The “Invasive Species Detection Protocol” atthe end of this chapter is designed to detectinvasive species in the early stages ofestablishment on a site.

.1.32elbaT .gnirotinomseicepsevisavniotsdohtemgniylpparofsenilediuG


lanoitiddA lacipyT


stniopotohP enoN oN etaredoM

*tpecretnitniop-eniLliosdebrutsidddaylbissoP

ssalcecafrusliossaoN etaredoM

*tpecretnipaG enoN oN etaredoM

tsetytilibatslioS enoN oN woL

*tcesnarttleB aerahcraesesaercnI oN hgiH

yratnemelppuS




*noitcudorptnalP enoN oN etaredoM




yevrusnoitategevlennahcnairapiR enoN oN woL



Figure 23.1. Cheatgrass grassland with sagebrush.

Pho

to b

y M

ike

Pel

lant

121

Invasive species notesLine-point intercept. The Line-point interceptmethod can be used to quantify invasive speciescover changes where the species is a significantcomponent of the plant community (generallygreater than five percent cover). Line-pointintercept cover and composition indicators alsooften reflect the resistance of a site to invasivespecies establishment.

The plant community can affect resistancedirectly by competing with the invasive species. Itcan affect resistance indirectly through its effectson herbivore populations and soil microbialcommunities. It can also indirectly affectresistance to invasion through its effect on thetiming, frequency and intensity of disturbances,which then modify conditions for establishmentof both invasive and non-invasive species.Relevant indicators are site specific and mayinclude percent bare ground or percent cover of aparticular functional group.

Gap intercept. The proportion of the land coveredby large gaps in foliar or basal cover directly affectsinvasive plant establishment through its effect oncompetition and soil stability. It can indirectlyaffect invasive plant establishment through itseffects on small herbivore activity and larger scaledisturbances such as fire. No new indicators arerequired, but the gap sizes of interest may varydepending on species.

Belt transect. The Belt transect is one of the mostrapid methods for monitoring invasive species thatcover too little area to be reliably detected with theLine-point intercept method (generally less thanfive percent cover). It can also be used toquantitatively monitor the appearance of smallseedlings where it is known that the speciesalready exists in the seedbank, or where there is ahigh risk of introduction.

Plant production. Please see the discussion inChapter 19, “Livestock production notes.”

Plant species richness. Please see the discussion inChapter 18, “Riparian notes.”

Invasive species

Invasive Species DetectionProtocol(1) Use existing information to stratify the

landscape into areas that have an inherentlyhigh invasion risk for each species, based onsoil and climate. Ecological Site Descriptions(Chapter 2) can be extremely helpful and oftenlist potentially invasive species. Aerialphotographs and other remote sensing toolscan be extremely useful in developing risk-based landscape stratification.

(2) Within high risk monitoring units, identifyareas most susceptible to invasion. Thisanalysis should be based on risk of dispersal(the risk that seeds will be brought to the site)and risk of successful establishment (theprobability seeds will land in an area favorablefor establishment). For example, trails arehighly susceptible to invasion. Trails have anincreased risk of invasion because of the highprobability of dispersal from distant plantpopulations. Trails are also at risk because trailmargins are often disturbed, reducingcompetition against invasives.

(3) Identify additional high-risk areas each year.For example, the establishment of a newcampground, road or mineral survey operationcan increase the risk of invasive speciesestablishment. Again, aerial photographs andother remote sensing tools can be invaluablein this process.

(4) Randomly select areas for ground-based surveysbased on risk analyses in 1 through 3 above.

(5) Visit each area and complete a rapidassessment that includes the following:• Estimate and record presence, number and

size of invasive species.• Predict the probability that population size

will increase for all invasive speciesencountered, based on site characteristics,climate and disturbance regime.

• Evaluate future invasion risk, including theneed to return to the area within aspecified period of time.

• Record GPS locations of invasive plantsand populations.

(6) Revise risk analysis (steps 1 through 3) basedon field observations.

122

Chapter 24

State and transition models: anintroductionState and transition (S&T) models (Figs. 24.1 and24.2) illustrate possible changes in plantcommunities and soil properties and theirinteractions. They can be used, together withassessments of the current status, to help decidewhere to monitor based on where change is mostlikely to occur. They can also be used to helpdecide what to monitor, because they oftenprovide information on soil and vegetationchanges that are likely to precede a change instate. States are distinguished by transitions, whichmay be relatively irreversible, reflecting asignificant increase in energy required to shiftback to the previous state.

Individual S&T models are usually developedfor each ecological site. Ecological sites are definedas land that has a similar potential to support aparticular range of plant communities based onsoils and climate. Land included in each ecologicalsite is expected to respond similarly to differenttypes of disturbance, climate and management.

State and transition models generally includeat least two states, and one or more plantcommunity within each state. Plant communitieswithin a state are similar in their speciescompositions. Plant communities within a state

are generally functionally similar in their capacityto limit soil loss, cycle water and producevegetative biomass. Changes among plantcommunities within states are considered to bereversible through simple changes in grazingmanagement (in grazed ecosystems) or fluctuatingclimatic conditions. The S&T diagrams (Fig. 24.1)show possible transitions between states. Thediagrams also illustrate the factors that increasethe probability that changes will occur. Transitionsbetween states are reversible only throughgenerally costly, intensive practices such as shrubremoval or soil modification.

The NRCS, BLM, The Nature Conservancy andother organizations are currently developing stateand transition models, and similar types ofmodels. Many are available from NRCS. Pleasecontact your local NRCS field office or refer to theNRCS website for state and transition modelspertaining to your ecological sites. All indicatorsdescribed in this manual can be used to helpquantitatively define states and the probabilitythat transitions will occur. For more informationon the development of these models, seeBestelmeyer et al. (2003) and Stringham et al.(2001).

Figure 24.1. Typical state and transition model structure (based on Bestelmeyer et al. 2003 and Stringham et al.2003). Large boxes are states defined by relatively irreversible transitions. Small boxes within states representplant communities. Transitions (dashed lines) are relatively reversible. Single-state systems are possible whereno thresholds have been identified.

------ ------

----------

------

----------

----------

----------

123

Sideoats gramaShrub live oak

Blue gramaHairy grama

Black gramaSideoats grama

Hairy gramaShrub live oak

Mixed grass savanna

Juniper(Blue grama)

SacahuistaShrub live oak

Live oak(Blue grama)

Woody/succulent dominated

SparseBlue grama

Blue grama/bare

1a

1b

2a 2b 3

1a. Heavy grazing, summer drought,decreased fire

1b, Clearing, seeding, increased firefrequency

2a. Heavy grazing, erosion, reductionof A horizon

2b. Erosion control, gully destruction3. Clearing after erosion is severe

Figure 24.2. State and transition conceptual model for the “Breaks” ecological site in west-central New Mexico(Major Land Resource Area 36, Land Resource Unit WP-3). General structure follows Bestelmeyer et al. (2003)and Stringham et al. (2001, 2003). See description on following page.

S & T models

----------

----------

----------

----------

124

Description for state and transition model for an ecological site (“Breaks”) in west-central New Mexico (MLRA WP-3)(Fig. 24.2).

OverviewThe Breaks sites intergrade with Hills sites and often contain Loamy sites occurring as narrow to broad drainageways. Thehistoric plant communities of the Breaks sites are dominated by black grama (Bouteloua eriopoda) and sideoats grama(Bouteloua curtipendula) and/or blue grama (Bouteloua gracilis) among others, depending on soil types and aspect. Underheavy grazing pressure, especially on steeper slopes and on soils with strong argillic (clay-rich) horizons, erosion may lead to apersistent loss of vegetation. A decline in fire frequencies, or perhaps regional increases in the relative amount of winter rainfallor grazing, may lead to significant increases in the abundance of woody plants and succulents including sacahuista (Nolinamicrocarpa), shrub liveoak (Quercus spp.), and one-seed juniper (Juniperus monosperma). The established woody plants maycompete with grasses and lead to persistent reductions in grass abundance. No systematic studies of communities, states ortransitions have been performed in the Breaks site.

Catalog of states, community pathways, and transitionsMixed-grass savanna: The expression of the community depends upon aspect and soil. On south-facing slopes, black gramatends to dominate and there may be some sideoats grama among other grasses. On north-facing slopes, sideoats grama domi-nates, with blue grama and hairy grama (Bouteloua hirsuta) as subordinates; black grama occurs in smaller amounts. In somecases (especially west of Silver City), sacahuista (Nolina microcarpa) may be dense enough to be considered a secondarydominant. Live oak, sacahuista, and juniper exist in low densities giving the site a savanna aspect. Grazing and drought-inducedmortality may lead to reductions in black and sideoats grama and dominance by hairy grama, blue grama, or annuals.Diagnosis: Sacahuista, oak and juniper are present and scattered; most of the ground surface is grassy, with few large bare areas.

Transition to woody/succulent-dominated state (1a): It is unclear why succulents or trees increase in abundance,although it is likely that the subsequent decline in grasses is due to competition for water and nutrients. The formationof bare ground patches due to grazing, decreases in fire frequency, and increases in winter precipitation, eitherindependently or in concert, may be responsible for the transition.Key indicators of approach to transition: Increases in bare ground, decreases in litter cover and grass cover, increasedfrequency of oak seedlings and small sacahuista (threshold may have been crossed), decreased fire frequency.Transition to blue grama/bare state (2a): Heavy grazing, especially in drought conditions on steeper slopes and onsoils with shallow, strong argillic horizons (e.g., Lonti gravelly loam) may result in grass loss and subsequent erosionof the organic matter-rich A horizon.Key indicators of approach to transition: Increases in bare ground, decreases in litter cover and grass cover, surface soilloss, water flow patterns, rills, pedestalling of plants and stones.

Woody/succulent-dominated: Grass cover is often highly reduced and shrubs, trees, or succulents become dominant. Bareground is extensive, and scattered, small blue grama or hairy grama plants represent the dominant grass cover. West of SilverCity, sacahuista tends to dominate in this state, and liveoak may or may not be a secondary dominant. In other cases, juniper oroak may dominate.Diagnosis: Oak, sacahuista, and/or juniper are the dominant perennial species and the bare ground areas between them areinterconnected. Grass clumps are small and scattered. Evidence of erosion (rills, water flow patterns, pedestalling) is common.

Transition to woody/succulent-dominated state (1b): Thinning of woody or succulent species may release grasses fromcompetitive suppression and grasses may colonize patches where trees or sacahuista were present. If erosion ininterspaces has not been severe, recolonization may take place there over several years.Transition to blue grama/bare state (3): Tree and succulent removal, especially on slopes, may accelerate erosion ifgrasses do not respond to the treatment and the soil is exposed to raindrop impact and erosion.

Blue grama/bare: This state is characterized by extreme erosion and tends to occur on steeper slopes. Bare ground cover isextreme, gullies may be present, and few small perennial plants, usually blue grama, are present. Trees and succulents are notespecially abundant.Diagnosis: Bare ground is interconnected, and trees and succulents are not especially abundant. Evidence of erosion is common,the mollic A horizon is very shallow (a few cm) or missing.

Transition to mixed grass savanna state (2b): The placement of structures (e.g., terraces) to retard erosion and thataccumulate soil, in addition to the destruction of gullies, may be used to initiate the eventual recovery of perennialgrass dominance.

Information sources and theoretical background: Communities, states, and transitions are based upon information in theEcological Site Description and observations by Gene Adkins, NRCS and Brandon Bestelmeyer, USDA-ARS Jornada Experi-mental Range.

S & T models

125

Chapter 25

Remote sensing

R emote sensing includes any data that arecollected remotely, including aerialphotographs, satellite imagery and digital

elevation models generated from aircraft orsatellites.

Remote sensing can increase the quality andcost-effectiveness of monitoring programs in anumber of different ways. It can be used to stratifythe landscape into relatively homogeneous units,to extrapolate ground-based measurements and, insome cases, to quantify properties and processes inthe absence of ground-based measurements usingpreviously established relationships.

Figure 25.1. Example of a color IR aerial photo ofMimbres Watershed.Increasing monitoring cost-

effectiveness with remote sensingIncorporating remote sensing imagery into themonitoring design process at an early stage candramatically increase cost-effectiveness andreliability. It helps to focus monitoring onrepresentative areas with a high potential forchange, while avoiding areas that have alreadycrossed a threshold. Although imagery used forthis step should be as recent as possible, the actualdate is not as critical for the monitoring designstep as when used as monitoring data. Additionally,variability in image quality is much less criticalthan when the imagery is being used directly formonitoring.

Options for incorporating remote sensinginto monitoring programs are summarized inTable 25.1. Option 1 can be done with or withoutGIS knowledge. Options 2 and 3 (Table 25.1)require training or extensive experience in remotesensing and GIS. Option 3 is difficult, but notimpossible, to apply to larger areas. It can be moreeasily applied to relatively small areas (farms,ranches or conservation areas). All three optionsoften can be applied together.

Option 1. It is appropriate to use remote sensingimagery for monitoring unit stratification andextrapolation where the imagery lends itself to

visual classification of geomorphic and vegetationunits. During stratification, use remote sensingimagery (e.g., aerial photographs), together withother available spatial data, to stratify thelandscape into relatively similar landscape units(Figs. 2.1 through 2.4 in Ch. 2). Where possible,further subdivide landscape units based on currentvegetation, management and the status of thethree ecological attributes (soil and site stability,hydrologic function and biotic integrity).

The next step in stratification is to combinethese spatial data with state and transition models(Bestelmeyer et al. 2003; Briske et al. 2003;Stringham et al. 2001, 2003) and information oncurrent and potential drivers (Brown and Havstad2004). All of this information can be used toidentify landscape units with a relatively highpotential for degradation or recovery.

Extrapolation using remotely sensed datarequires an adequate number of plots to representthe landscape. Develop a good relationshipbetween these ground-based measurements andremote-sensing indicators. If this is not feasible, itis possible to extrapolate using remotely sensedimagery if extensive, long-term knowledge of thelandscape, its ecological communities, and theirinteractions and drivers exists.

126

Option 2. The ability to make coarse-scaleextrapolations based on repeated, ground-truthedimagery depends on the scale of the imagery andthe scale of the vegetation heterogeneity anddynamics. Imagery and ground-based data must becollected in the same time frame (either within thesame month or within the same season). Expertisein image classification is required for this option.

Option 3. Fine-scale extrapolation based onrepeated, ground-truthed imagery has the samerequirements as described in Option 2, but to ahigher degree. Fine-scale extrapolation requires thehighest level of GIS expertise, field sampling, andimage quality. Imagery must be at a fine enoughresolution to detect the same community levelchanges as the ground-based measurements.Defining the relationship(s) between the field-based indicators and remote-sensing indicators canbe challenging. It can even, at times, be impossible(see “Monitoring with remote sensing alone”below).

Monitoring with remote sensingaloneA fourth option for incorporating remote sensinginto monitoring programs is to use predefinedrelationships between remote sensing indicators

and ground-based measurements. This option isproblematic because of the high spatial andtemporal variability in soil and vegetationrelationships. Confounding this factor is therelatively low vegetation cover typical forrangelands. In addition, vegetation reflectance andtemperature change rapidly and unpredictably inresponse to highly variable soil moisture. All ofthese caveats make monitoring solely via remotesensing in arid and semi-arid communitieschallenging.

New techniques that take advantage of greatercomputing power, higher resolution images andintegration of information using different types ofimages are currently being developed at theJornada Experimental Range and elsewhere (Rangoet al. 2003). While these techniques are likely to bemore sensitive and reliable, it is unlikely that wewill ever be able to design comprehensivemonitoring programs based exclusively on remotesensing. Periodic ground-truthing is likely to berequired for most applications.

ConclusionsBy using remote sensing imagery primarily toimprove monitoring program design, we exploitthe strengths of remote sensing technologies.Using remote sensing imagery only forstratification allows us to avoid the pitfalls of over-

Remote sensing

.1.52elbaT .smargorpgnirotinomdesab-dnuorgotnignisnesetomergnitargetnirofsnoitpofonosirapmoC.1002hcsannUdnayenohcuMees,yregamifosepyttnereffidfonosirapmocaroF

noitpO noitacilppA elacsdnaepytyregamI egdelwonK tsoC

1noitacifitartstinugnirotinoM

gnilpmasdesaercnirofycneiciffe

tsalniemityna(sotohpriA)sraeynet

yfissalcyllausivotytilibAnoitategevdnacihpromoeg

stinu

woL

2noitalopartxeelacs-esraoC

,detaepernodesabyregamidehturt-dnuorg

rehtodnaSIDOM,tasdnaLsitahtyregamilartcepsitlum

dnadetarenegyralugerdevihcra

dnassecorpotytilibAatadlartcepsitlumyfissalc

.deM

3noitalopartxeelacs-eniF

,detaepernodesabyregamidehturt-dnuorg

ria,SONOKI,driBkciuQelgnisrehtodnasotohplartcepsitlumdnadnab

yregami

dnassecorpotytilibAatadlartcepsitlumyfissalc

hgiH

127

Remote sensing

reliance on relatively abstract indicators, many ofwhich require new ground-based calibration datafor each new set of imagery. By combining remotesensing with qualitative assessments and state andtransition models, we can target both managementand monitoring to those parts of the landscapewith the highest probability of change. Where it ispossible to obtain repeated, concurrent ground-based and remote-sensing data, imagery can be

used to generate a more precise extrapolation thanis possible with the initial stratification alone.However, the ability to make such extrapolationsis tightly linked to the type of vegetationcommunity and the resolution of the imagery.

Parts of this Chapter were adapted from Herrick et al.(2003).

128

Chapter 26

Soil carbon

S oil carbon can be a useful and accessiblelong-term indicator of change in thefunctioning of an ecosystem. Soil carbon is

directly related to soil organic matter content, akey indicator of soil quality. Soil organic matter isimportant for maintaining soil structure. Soilswith good soil structure generally have lowererosion rates, higher water infiltration rates andhigher water-holding capacities. Soil organicmatter also serves as an important nutrientreservoir.

Typically, increasing soil carbon has positiveeffects on soil and ecosystem health. But simplyincreasing soil carbon may not always be the landmanagement goal. For instance, replacement ofgrasslands by woody-dominated plant communitiesmay increase total carbon sequestration at thelandscape level, but reduce soil quality near the soilsurface in plant interspaces. This reduction in soilquality associated with woody plant invasion isparticularly common in arid ecosystems.

Figure 26.1 Soil organic matter and soil carbonare usually higher near the soil surface.

In most cases, meeting land managementobjectives will require tracking changes in soilcarbon over time. There are three options forcarbon monitoring: measurement, modeling, andmonitoring changes in vegetation cover,composition and production.

Soil carbon measurement is currently tooexpensive in most arid and semi-arid ecosystems.This is due to a combination of high sampling andanalysis costs and the large number of samplesrequired to detect a change.

Soil carbon models predict changes in soilcarbon based on soil properties, current vegetationand climate. However, most available carbonmodels focus on agricultural, forest and grasslandecosystems, and do not reliably predict soil carbondynamics in diverse arid and semi-arid ecosystems.Given their drawbacks, carbon measurement andmodeling are not yet recommended as viablemonitoring options. However, the accuracy ofboth measurement and modeling is improving.Cost-effective rangeland carbon monitoringsystems integrating the two approaches should beavailable within the next decade.

A third, more practical, option for the presenttime is to simply monitor changes in vegetationcover and composition (Line-point interceptmethod) and production (Plant productionmethod). These indicators cannot currently predict

Carbon sequestrationIn addition to being a good indicator of soilquality, sequestering (storing for long periods)carbon in the soil keeps it out of theatmosphere, where it occurs as carbon dioxideand contributes to the greenhouse effect andglobal warming. The United States hasadopted a market-based approach to provideincentives for reducing greenhouse gases inthe atmosphere. Efforts are currentlyunderway to establish standard systems for“trading” carbon released into the atmosphere(e.g., from fossil fuel combustion) foradditional carbon stored in the soil as soilorganic matter. From a practical perspective,this means that carbon producers (e.g., powerplant operators) can purchase credits in amarket. Those credits may be supplied by avariety of sources, including increased soilcarbon sequestration. However, formal tradingprocedures are not currently in place and thedetails are still uncertain.

129

soil carbon changes, but they are associated withchanges in carbon inputs. In general (but notalways), soil carbon increases with cover andproduction. In systems in which a significantportion of the production is consumed bylivestock or wildlife, utilization records (“Short-term monitoring” in Quick Start) should also becarefully maintained.

Carbon notesPhoto points. Soil profile photos showing near-surface carbon accumulation where accumulationrates are high can supplement vegetation photos.They can help substantiate changes recorded inthe quantitative soil and vegetation data. They arealso useful communication tools.

Line-point intercept. Line-point intercept data areused in models to estimate carbon inputs. Bothcover and species composition are required forcarbon models.

Gap intercept. Gap intercept may be used as anindex of soil erosion risk. The highestconcentration of soil organic carbon is usually in

.1.62elbaT .gnirotinomnobracliosotsdohtemgnirotinomgniylpparofsenilediuG


lanoitiddA lacipyT


*stniopotohP enoN a/n etaredoM

*tpecretnitniop-eniL enoN seY hgiH

*tpecretnipaG enoN oN etaredoM

*tsetytilibatslioS enoN oN etaredoM

*tcesnarttleB enoN oN etaredoM

yratnemelppuS




noitcudorptnalP enoN oN hgiH

ssenhcirseicepstnalP enoN oN woL


*ytisnedeerT enoN oN etaredoM

yevrusnoitategevlennahcnairapiR enoN oN woL



Soil carbon

the top ten centimeters (4 in), which is also thelayer that is most susceptible to soil erosion.

Soil stability test. Soil stability is closely related tothe creation of new soil organic matter, and maybe a good early warning indicator of changes intotal soil organic carbon. However, therelationship between soil stability and soil organiccarbon is highly variable. It should only be used asa general indicator to compare amongmanagement systems. When testing for soilstability as an indicator of soil carbon, be sure totest soil from different depths. In many arid soils,the stability of the top few millimeters of the soilsurface is stabilized by cyanobacteria. Changes incyanobacterial biomass are not necessarily relatedto changes in root production, which is theprimary source of soil organic matter in mostrangeland ecosystems.

Belt transect and Tree density. Changes in boththe density and cover of shrubs and trees have thepotential to significantly modify soil carbonsequestration. Either of these two methods, Belttransect or Tree density, can be used to detect

130

changes in woody species when cover is belowthat which can be monitored using Line-pointintercept methods (generally five percent).

Additional resourcesGuidelines for soil carbon monitoring arecurrently in the process of being established.Because this field is so dynamic, the best approachto locating the most current and relevant resourcesis an Internet search, focusing on those resources

that include evaluations of the cost, accuracy andprecision of the proposed methods. A recentCouncil on Agricultural Science and Technologyreport (CAST 2004) provides a good overview ofmany of the issues associated with soil carbonsequestration. For information on CENTURY, oneof the models currently being applied in theUnited States, see www.nrel.colostate.edu/projects/century5/reference/html/Century/desc-intro.htm(accessed June 23, 2008).

Soil carbon

Section IV: Special topics - USDA ARS · indicators from the core measurements described in Quick...

Documents

Transcript of Section IV: Special topics - USDA ARS · indicators from the core measurements described in Quick...