Submitted 30 November 2015Accepted 3 April 2016Published 21 April 2016

Corresponding authorXiao Jun Yang yangxjmailkizaccn

Academic editorDonald Kramer

Additional Information andDeclarations can be found onpage 15

DOI 107717peerj1968

Copyright2016 Dong et al

Distributed underCreative Commons CC-BY 40

OPEN ACCESS

Winter diet and food selection of theBlack-necked Crane Grus nigricollis inDashanbao Yunnan ChinaHao Yan Dong12 Guang Yi Lu12 Xing Yao Zhong3 and Xiao Jun Yang1

1 State Key Laboratory of Genetic Resources and Evolution Kunming Institute of Zoology Chinese Academyof Sciences Kunming China

2Kunming College of Life Science University of Chinese Academy of Science Kunming China3Administrative Bureau Dashanbao National Nature Reserve Zhao Tong Yunnan China

ABSTRACTThe Black-necked Crane Grus nigricollis is a globally vulnerable species whose foodis the factor determining its long-term survival Understanding dietary habits foodpreferences and related factors will facilitate the development of effective conservationplans for the protection of this vulnerable species For this purpose we used videorecordings and sampling of food availability to examine the dietary compositionand temporal variation in food selection of Black-necked Cranes wintering in theDashanbao National Nature Reserve China The Black-necked Cranersquos diet consistsprimarily of domestic food crops such as grains (74) and potatoes (8) in additionto invertebrates (14) A much smaller proportion of the diet was comprised ofturnips and wild herbaceous plants and tubers There was monthly variation infood selection partially related to food availability Grains were most available inNovember and decreased through thewinter whereas invertebrates weremore availablein November and February than in December and January Grain consumptionwas lowest in November but higher from December through February Invertebrateconsumptionwas highest in November and February The cranes preferred to eat grainsthroughout winter months while they mainly selected invertebrates in November andFebruary We suggested invertebrate populations sharply declined in December andJanuary due to the low temperature In addition grain consumption was negativelyassociated with invertebrate availability In November when invertebrates were mostabundant and despite a concomitant peak in grain abundance we suggested cranesexhibited a preference for invertebrates over grainsWe recommend that the protectionadministration provide appropriate supplemental foods for cranes during freezingweather

Subjects Animal Behavior Ecology ZoologyKeywords Black-necked Crane Grus nigricollis Diet composition Food availability Foodselection Dashanbao

INTRODUCTIONThe Black-necked Crane Grus nigricollis is a globally vulnerable species with the mainbreeding distribution in the high altitude Tibetan-Qinghai Plateau The cranes migrateshort distances to winter in the lower altitude (primarily 2000ndash3200 m) YunnanndashGuizhouPlateau (Harris amp Mirande 2013) Through telemetry and banding data it has become

How to cite this article Dong et al (2016) Winter diet and food selection of the Black-necked Crane Grus nigricollis in Dashanbao Yun-nan China PeerJ 4e1968 DOI 107717peerj1968

clear that the birds using the Eastern migratory route (in the following referred to as theEastern Black-necked Cranes) breed in northern Sichuan and southern Gansu provincesand mainly winter in northeast Yunnan and southwest Guizhou (Li amp Li 2005 Qian et al2009) More than 50 of the wild populations of this species are currently suffering due tosignificant habitat destruction resulting from grassland degeneration (Li amp Li 2012) andconventional agricultural practices that have decreased the diversity of available food typesfor this species in northeast Yunnan Food is the factor determining the long-term survivalof Black-necked Crane (Liu et al 2014) Thus understanding the Black-necked Cranersquosdietary habits food preferences and the associated factors will facilitate the developmentof effective conservation plans for the protection of this vulnerable species

Determining the dietary composition of wild birds is essential for understanding how theanimals interact with their habitats and consequently for identifying their preferred foodtypes (Baubet Bonenfant amp Brandt 2004) Their late discovery and remote range led to alate start in research pertaining to Black-necked Cranersquos feeding habits (Harris amp Mirande2013) To this point research surrounding the Black-necked Cranersquos diet has includedquantitative studies on various types of domestic and wild plant foods (Li amp Nie 1997Bishop amp Li 2001 Liu et al 2014) and qualitative studies on animal-based foods (Han1995 Hu et al 2002 Li amp Li 2005 Liu Yang amp Zhu 2014) Nonetheless there remains alack of synthetic analyses or comparative data regarding the proportions of domestic foodcrops animal-matter and wild plants consumed by the Blacked-necked Crane duringthe winter

Until now fecal microhistological analysis has been the only method used to identifyplant material consumed by wintering Black-necked Cranes (Li amp Nie 1997 Liu et al2014a) These studies did not mention the consumption of animal-based foods due to theneed for alternative methods to collect this data (Liu Yang amp Zhu 2014b) Generally fecalanalysis can create a bias due to the high variability in digestibility of different food items(Redpath et al 2001) Thus we chose video recording as an alternative method to betterunderstand the food selection of Black-necked Cranes This method provided a simpleminimally invasive manner to directly observe the feeding behavior of the threatenedbird species in order to estimate their dietary composition (Newton 1967 Price 1987Yoshikawa amp Osada 2015)

Previous studies suggest that variations in temperature may impact food availability(Kushlan 1978 Stapanian Smith amp Finck 1999) As mentioned by Alonso Alonso ampBautista (1994) low temperatures may decrease grain availability for CommonCranesGrusgrus by increasing foraging costs due to changes in soil properties Likewise temperatureis an important correlate of insect activity further affecting the invertebrate-feeding birdsHigher temperatures are associated with more frequent droughts and dry soils (Martin1985) while lower temperature cause the soil to freeze Thus both affect the degree ofinsect activity (McCollogh Hayes amp Bryson 1927 Dowdy 1937 Zhou et al 2015) and theiravailability for birds Considering this information we considered that the temperaturechanges would influence the attributes of available foraging sites affecting food availabilityand food selection

Dong et al (2016) PeerJ DOI 107717peerj1968 219

The goal of this research was to better understand factors influencing Black-neckedCranes selection of different feeding habitats during the winter This information mayfacilitate the development of strategies to protect the Eastern Black-necked Crane whoselargest population winters in their most important wintering sites in the DashanbaoNational Nature Reserve on the YunnanndashGuizhou Plateau (Li amp Yang 2002 Qian et al2009) In this report we provided a quantitative and comprehensive assessment of thecranesrsquo wintering diet which included domestic food crops animal-based foods and wildplants We analyzed the cranesrsquo diet composition food selection and any correlationbetween environmental factors food availability and food selection

METHODSEthics statementOur research on Black-necked Cranes in DashanbaoNational Nature Reserve was approvedby the Chinese Wildlife Management Authority and conducted under Law of the PeoplersquosRepublic of China on the Protection of Wildlife (August 28 2004)

Field permitThe Administration of ZhaoTong Forestry Bureau approved our study on behaviorobservation and food availability sampling in the research plot in Dashanbao NationalNature Reserve (IDZTL2008163)

Study siteDashanbao National Nature Reserve (hereafter referred to as Dashanbao Reserve2718prime38

primeprime

N 10314prime55primeprime

E altitudes of 3000ndash3200 m) is located in southwest China(Fig 1) and is listed as awetland of international significance under the Ramsar Conventionon Wetlands The Dashanbao Reserve is considered an important habitat for Black-neckedCranes as well as other wintering water birds It is also known for its upland wetlandecosystem (Zhong amp Dao 2005) The study area covers 19200 ha and is a warm humidplateau with a monsoon climate characterized by cool wet summers and cold dry wintersDuring winter months frequent days of sustained freezing temperatures can be expectedfrom December to January The mean temperature for January is minus1 C and for July127 C The mean annual temperature is 62 C with 123 frost-free days and 346 snowcover days per year The mean annual precipitation is 1165 mm (Li amp Zhong 2010)

A total of c 1200 Black-necked Cranes winter in the Dashanbao Reserve every yearfeeding on agricultural farmlands as well as wild grasslands (Kong 2008) For the purposesof this study supplemental feeding by humans was ignored because only c 3 kg of corn arefed to fewer than 50 cranes every day (Kong et al 2011a) which would have little impacton the overall dietary composition and food selection for the cranes Farmland includedfields of cereal (Avena sativa and Fagopyrum tataricum) potatoes (Solanum tuberosum)and turnip (Brassica rapa var rapa) Local farming uses a 3-year rotation system in whichcereal is grown one year followed by two years of potato or turnip and then back to cerealThus a mosaic of patches of cereal potato and turnip characterizes the farmland witheach occupying about the same surface area each year Wild grasslands were comprised of

Dong et al (2016) PeerJ DOI 107717peerj1968 319

Figure 1 Map of the Dashanbao National Nature Reserve (thin black line) showing the location of ourstudy areas The red dot at the upper right designates the location of the Reserve within China Circles in-dicate sites where we recorded cranes foraging Stars indicate the sites of food availability sampling Thicklines indicate the transects Dahaizi Reservoir (blue) and Tiaodunhe Reservoir (yellow) are indicated Thinblue lines indicate smaller water bodies

meadows withminimal water (Kong et al 2011a) and dominated by orchard grass (Dactylisglomerata) bluegrass (Poa annua) Leontopodium Trifolium Pterospermum heterophyllumPedicularis densispica Luzula multiflora Hemiphragma heterophyllum (Kuang et al 2008)The cranes have been reported to forage on Pedicularis Stellaria Polygonatum andVeronica

Dong et al (2016) PeerJ DOI 107717peerj1968 419

(Kong et al 2011a Liu et al 2014a) The study area covered most of the foraging sites ofBlack-necked Cranes

Bird observationsField data were collected fromNovember 2013 to February 2015 in the Dashanbao ReserveSince Black-necked Cranes are highly vigilant and the landscape of the Dashanbao Reserveconsists of rolling hills and valleys we were unable to adequately observe the flocks fromour vehicles along the main road in the Dashanbao Reserve and we had to walk alongsmaller roads Therefore we selected three transect routes crossing the mountain ridge ofthe reserve at two sites which housed the largest flocks of cranes according to the reservestaffrsquos experience and the suggestions from previous research in October 2013 (Kong et al2011a) (Fig 1) The majority of cranes arrived in early November and remained feeding inDashanbao Reserve until early March We observed cranes for three days every week for15 weeks between the second week of November and the end of February In both yearswe missed one week due to severe weather

We videotaped the birds for 5-min intervals each along all transect routes We walkedtransects once per day and switched direction of travel on subsequent days During thistime the cranes were undisturbed and at a maximum distance of 80 m from our point ofobservation Based on personal observation the cranes would startle and flee their feedingsite when observed from a distance of less than 60 m Thus most sightings were between60 and 80 m from the birds For videotaping we chose cranes at random from a withinthe total number of birds in a flock This resulted in observations of 50ndash70 of all cranesin each flock A total of 505 good quality 5-min videos were recorded ensuring sufficientclarity to accurately differentiate among all the consumed food types For this study poorquality recordings and those lasting less than 5 min were discarded A Canon PowerShotSX30 IS digital camera with a 35times optical zoom was used for all the video recordings

Foraging behaviorWe combined feeding behavior and information about habitat type to determine food typeFood types were classified into 3 categories (1) domestic crops (including a grains bpotatoes and c turnips) (2) animalmatter (d invertebrates including primarily earthwormsand coleopteran larvae) (3) wild plants (including e herbaceous plants f roots or tubers)Videorecordings of foraging cranes were examined in slow motion to quantify numberof pecks per 5-min interval Every video was watched at least three times to confirmaccurate identification of the food types consumed by the feeding crane Depending on thetypes of food being eaten and the peck frequency four different types of feeding patternswere identified (1) high pecking frequency and ingestion of all the target food quicklyin farmland This pattern was used primarily for grain on the surface of the ground (seeVideos S1 and S2) (2) Digging up the soil to find and consume underground food such asroots or tubers (including potato and turnip) (see Video S3) Since tubers are too bulky forcranes to swallow they peck at them repeatedly swallowing smaller pieces until the item iscompletely consumed This behavior facilitates visual identification of tuber consumption(3) Consumption of invertebrates is also easily identifiable by a pattern in which the cranes

Dong et al (2016) PeerJ DOI 107717peerj1968 519

peck at a plot of turf capture their prey and then quickly swallow it (see Videos S4 andS5 Fig S1) This pattern leaves an obvious disturbance of the turf that can be used foridentification (see Figs S2 and S3) (4) Lastly the cranes used tugging (Ellis et al 1991)without digging up the soil primarily for aboveground foods consisting of herbaceousplants We distinguished this from foraging on grains via a lower pecking frequency andslower swallowing movements (see Video S6 Fig S4) We recorded the numbers of pecksfor each food type When there was more than one food type in a 5-min recording werecorded the number of pecks for each type separately

Food availabilityGiven the mosaic landscape of the Dashanbao Reserve the sampling sites for foodavailability were selected based on two criteria (1) The site needed to include a largesection of farmland and grassland bordered by farmland with three types of crops incultivation in the transects (2) The site must have been selected by at least one flockof cranes for foraging across three transects Based on these two criteria twelve plots offarmland (2ndash6 ha) and twenty plots of grassland (13ndash43 ha) were selected using GoogleEarth followed by a field survey (Fig 1) The proportion of land that each crop andgrassland occupied was obtained via monthly sampling The areas of the sampling siteswere calculated using Arcgis 92 (ESRI Inc Redlands CA USA)

To investigate the availability of consumable crops animal matter and wild plantswe proceeded to sample foods using quadrats (50times50times10 cm deep) placed at intervalsof 100 m along a straight line guided by GPS localization We used a direct collectionsampling method for cereal grains on unploughed plots and turned the soil for samplingcereal grains on ploughed lands The latter method was used for sampling potatoes turnipsinvertebrates (eg earthworms and Coleoptera larvae) herbaceous plants as well as tuberswithin a depth of 10 cm The length of a cranersquos bill is 124 cm (n= 10 105ndash140 cm)We only counted invertebrates larger than approximately 4 mm because that appearedto be the minimal size consumed by the cranes The count biomass and the depth offood types available in each quadrat were recorded We recorded the depths of frozen soilduring the sampling We placed 176 quadrats in grain fields and another 222 quadrats inpotato and turnip fields in 2013ndash2015 (sampled monthly for eight months over two years)Earthworms Coleoptera larvae herbaceous plants and roots or tubers were collectedfrom 295 quadrats in grassland in 2013ndash2015 (eight months) The extracted food itemswere stored in plastic bags and frozen until processing After defrosting cereals potatoesturnips invertebrates herbaceous plants and tubers were separated dried (60 C 48 h)and then weighed to determine dry biomass (0001 g precision)

Weather variablesDaily temperature values were taken from Zhonghaizi in the Dashanbao Reserve For ouranalyses we used the mean daily temperature and the mean minimum daily temperatureWe also counted the number of periods with three or more consecutive days of sustainedlow temperature (minimum temperature equal to or less than minus10 C)These would bedays when the ground would remain frozen thus preventing the cranes from being able todig for food

Dong et al (2016) PeerJ DOI 107717peerj1968 619

Statistical analysisMonthly trophic diversity was estimated using Shannonrsquos diversity index H prime =minus

sumPi ln(Pi) (Pielou 1966) where P i represents the proportion of each food type

Pi =NiN where Ni is the total number of ingested items of food type i and N is thetotal number of ingested food items of all types combined We calculated H prime using theproportion estimate derived for each food type present in the sample We used one-wayANOVA to test differences between months in diversity index Subsequently Bonferronitechniques were applied to correct the level of significance of the index We used theKruskalndashWallis nonparametric test to explore monthly differences in available biomassof four foods If the monthly differences were statistically significant Nemenyi tests inSPSS 20 to test for differences between months Statistical significance was obtained afterapplying the Bonferroni correction We estimated the monthly availability by multiplyingthe monthly surface of grassland and each type of farmland by the calculated biomassmeans Monthly percentage of availability for each food type was calculated by dividingfood availability of one food type by the total food availability of all types combined Foodselection by cranes was analyzed using the Savage selectivity index (Savage 1931 cited byManly et al 1993) Wi=Oiπi where Oi is the proportion of the sample of used resourceunits that are in category i and πi is the proportion of available resource units that arein category i The proportion Oi is calculated using the formula Oi = uiu+ where uirepresents the number of ingested items of a specific food type and u+ represents the totalnumber of ingested items of all food types Likewise the proportion of mean availablebiomass for a food type is calculated using the formula πi=AiA+ with Ai representingthe biomass of available resource in category i and A+ the biomass of all available resourceunits (Manly et al 1993 Avileacutes Saacutenchez amp Parejo 2002) This Savage selectivity index canrange from 0 to infinity with 0 indicating maximum negative selection 1 indicating noselection bias and infinity indicating maximum positive selection (Manly et al 1993) Thestatistical significance of the selection for each food type from a distribution proportionalto its availability was tested using the statistic (Wiminus1)2se (Wi)2 (Manly et al 1993)which follows the critical value of a χ2 distribution with one degree of freedom where se(Wi) is the standard error of Wi calculated using the formula

radic(1minusπi)(u+timesπi) (Manly

et al 1993) Statistical significance was obtained after applying the Bonferroni correctionfor the number of statistical tests (Rice 1989)

We determined the relationships between food availability variables and environmentalvariability (the mean daily temperatures minimum daily temperatures and number ofdays with frozen soil) using Pearson correlation coefficients in SPSS 20

To examine the association between food selection and environmental variables weused Canonical Correlation Analysis (CCA)Multivariate analyses were performed with thesoftware CANOCO (Ter Braak amp Smilauer 1998) Preliminary detrended correspondenceanalysis (DCA) was applied to three food selection datasets (grains potatoes invertebrates)to determine the length of the gradient This DCA revealed that the gradient was greaterthan 3 standard deviation units (42) justifying the use of unimodal ordination techniques(Ter Braak amp Verdonschot 1995) Following a preliminary canonical correlation analysis(CCA) we eliminated collinear environmental variables with high variance inflation factors

Dong et al (2016) PeerJ DOI 107717peerj1968 719

Table 1 Monthly yearly and two-year combined percentage of food types in the dietary composition of the Black-necked CraneG nigricolliswintering in the Dashanbao National Nature Reserve China For each month in two years the number of video recordings the total number ofpecks observed and the percent of pecks directed toward each major food type are shown

No of videorecordings

No of pecks Grain()

Potato()

Turnip()

Invertebrate()

Herbaceousplant ()

Tuber()

2013ndash2014 Nov 46 1180 3973 2029 001 3905 093 000Dec 47 1608 9524 188 000 163 100 025Jan 70 1808 8142 581 011 183 597 487Feb 50 1212 8251 495 000 1056 149 050Total year 213 5808 7472 823 003 1327 235 140

2014ndash2015 Nov 105 1342 4955 1118 007 3383 186 350Dec 53 1861 8855 430 000 161 489 064Jan 66 1495 8027 903 007 234 716 114Feb 68 1502 7324 533 000 2084 047 013Total year 292 6200 7290 746 004 1466 359 135

Both years combined 505 12008 7381 784 003 1396 297 138

(VIFgt 20) from further analyses The variable with the highest significant contributionwas included in the analysis (Monte Carlo permutation test P le 005 randomizationtest with 499 unrestricted permutations) Then we recalculated the contribution andsignificance for each variable Again the variable with the highest significant contributionwas included This procedure was repeated until none of the variables had a significantcontribution The variables we included were the distributed depths of grain the depths ofpotato the depths of invertebrate grain availability potato availability and invertebrateavailability Canonical correspondence analysis (CCA) with biplot scaling on inter-speciesdistances was used to display the relationship between food selection structure and the 6environmental variables (main food variables) All multivariate analyses were performedusing CANOCO version 45 software (Ter Braak amp Smilauer 2002)

RESULTSDiet compositionDomestic crops (grains and potatoes) and animal matter (invertebrates) collectivelycomprised the majority of the Black-necked Cranersquos diet followed by wild plants(herbaceous plants tubers) (Table 1) Turnips comprised less than 1 of the diet onaverage When we pooled yearly data domestic crops and animal matter accounted for9561 in total food items of which grains accounted for 7381 potatoes 784 andanimal matter 1396 respectively In November (both years combined) the proportionof grains consumed was the lowest compared to other months From December throughFebruary grain consumption was more than twice as high in 2013ndash2014 and more than14 times as high in 2014ndash2015 In contrast the highest consumption of potato andinvertebrates occurred in November followed by January (for potato) and February (forinvertebrates) while the lowest consumption for both food types occurred in DecemberHerbaceous plants and tubers comprised less than 5 of the diet on average One-way

Dong et al (2016) PeerJ DOI 107717peerj1968 819

Figure 2 Monthly mean Shannon index of diversity (H prime) for the diet of the Black-necked CraneG ni-gricolliswintering in the Dashanbao National Nature Reserve China

ANOVA indicated significant variation on monthly diversity of the diet (F = 2500 df = 3N = 8 P = 0005) Then Bonferroni post-hoc test showed the diversity of the diet washigher in November than in December (P = 0006) January (P = 006) and February(P = 0028) (Fig 2)

Food availabilityWild plant food accounted for the largest proportion of food available in the Black-neckedCranersquos environment (Table 2) When we pooled yearly data herbaceous plants andtubers accounted for 8976 with 4651 for herbaceous plants and 4325 for tubersDomestic crops (grain 117 potato 164 and turnip 294) and invertebrates (448)accounted for a much lower proportion of total food available KruskalndashWallis testindicated significant variation in the monthly availability of grain and invertebrates (grainH = 16402 P = 0001 invertebrate H = 13081 P = 0004) whereas we did not findsignificant effects of month on other types of food (P gt 005) The available biomass ofgrains in November and December was higher than that in the other two months (Table 2Nemenyi test Nov vs Dec H = 114 P = 029 Nov vs Jan H = 753 P = 0006 Novvs Feb H = 1360 P = 0000 Dec vs Jan H = 392 P = 0048 Dec vs Feb H = 646P = 0010) Invertebrate biomass was higher in November and February than that in theother two months (Table 2 Nemenyi test Nov vs Dec H = 755 P = 0006 Nov vsJan H = 456 P = 0033 Nov vs Feb H = 002 P = 0888 Feb vs Dec H = 838P = 0004 Feb vs Jan H = 523 P = 0022)

Dong et al (2016) PeerJ DOI 107717peerj1968 919

Table 2 Monthly availability of biomass of all food in the Dashanbao National Nature Reserve China

2013ndash2014 2014ndash2015 Two yearcombined

Nov Dec Jan Feb Totalyear

Nov Dec Jan Feb Totalyear

Grains Mean food biomass (g025 m2) 079 073 035 018 052 245 099 049 043 107 081Sample N 21 23 20 20 84 23 19 20 30 92 176Percentage of food availability () 045 054 038 026 041 508 117 056 093 194 117

Potatoes Mean food biomass (g025 m2) 216 122 062 077 111 138 118 082 082 102 105Sample N 12 14 18 15 59 26 19 34 30 109 168Percentage of food availability () 107 116 108 156 122 251 181 150 247 207 164

Turnip Mean food biomass (g025 m2) 1250 1601 715 599 1194 421 574 578 338 444 763Sample N 8 8 4 3 23 8 8 3 12 31 54Percentage of food availability () 237 531 332 202 325 291 305 281 170 262 294

Invertebrate Mean food biomass (g025 m2) 103 039 020 081 059 062 029 049 060 049 051Sample N 10 12 12 13 47 35 41 41 31 148 195Percentage of food availability () 377 254 174 565 343 823 305 456 630 554 448

Herbaceousplant

Mean food biomass (g025 m2) 760 379 136 318 412 238 238 255 247 245 299

Sample N 25 22 20 26 93 49 40 51 62 202 295Percentage of food availability () 4353 4505 5073 414 4518 4711 4541 5042 4845 4785 4651

Tubers Mean food biomass (g025 m2) 788 695 817 701 746 181 434 433 385 347 473Sample N 11 19 19 18 67 41 29 28 47 145 212Percentage of food availability () 4882 454 4271 491 4651 3417 4550 4014 4015 3999 4325

Food selectionIn comparing the six types (3 categories) of foods available to the foods selected the Savageindex showed that the cranes preferred grain through the wintering period Invertebrateswere the second preferred food type in November and February Potatoes were preferredin November in 2013ndash2014 (Table 3) whereas they were either avoided or showed nosignificant preference in the other months Turnips herbaceous plants and tubers wereavoided through the wintering period

Relationships between environmental factors and food availability andfood selectionWe used the Pearson correlation coefficient to determine the correlation betweenenvironmental temperatures and the availability of key food items The results showedthe depth distribution of invertebrate was significantly negatively correlated with meantemperature and mean of minimum temperature and was positively correlated with thenumber of days during which the ground was frozen (Table 4) In addition the number ofinvertebrates at depths of 0ndash1 cm and 11ndash2 cm were positively correlated with temperatureand negatively correlated with the number of days with frozen ground (Table 4) The meandepth of the frozen ground was 493 cm in December (n= 10 26ndash69 cm) and 312 cm(n= 5 29ndash35 cm) in January (see Fig S5)

Dong et al (2016) PeerJ DOI 107717peerj1968 1019

Table 3 Food selection of Black-necked Cranes (G nigricollis) for the six most available food types in the Dashanbao National Nature ReserveChina in relation to month and year Shown is number of video recordings the number of pecks food availability (πi) food selection (Oi) theSavage selectivity index (Wi) for each food type standard error of the index (se) and the statistical significance (P) for our results Significance isreached at P lt 0006 after applying the Bonferroni correction

No of videorecordings

No ofpecks

Oi πi Wi se P selection

2013ndash2014 Nov Grain 13 470 040 004 1034 015 +

Potato 3 240 020 006 337 011 +

Turnip 0 0 0 023 0 005 minus

Invertebrate 26 462 039 002 1632 019 +

Herbaceous plants 4 8 001 044 002 003 minus

Tuber 0 0 0 020 0 006 minus

Dec Grain 26 1520 095 004 2149 012 +

Potato 4 30 002 004 042 012 minus

Turnip 0 0 0 034 0 004 minus

Invertebrate 7 26 002 001 132 022 NS +

Herbaceous plants 7 21 001 022 006 005 minus

Tuber 3 11 001 035 002 003 minus

Jan Grain 22 1472 081 003 2671 013 +

Potato 13 105 006 005 121 010 NS +

Turnip 2 2 0 012 001 006 minus

Invertebrate 12 33 002 001 177 023 +

Herbaceous plants 6 108 006 011 053 007 minus

Tuber 16 88 005 067 007 002 minus

Feb Grain 22 1000 083 001 5709 024 +

Potato 5 60 005 005 108 013 NS +

Turnip 0 0 0 007 0 010 minus

Invertebrate 14 128 011 004 254 014 +

Herbaceous plants 5 18 001 033 005 004 minus

Tuber 4 6 0 050 001 003 minus

2014ndash2015 Nov Grain 18 665 050 017 297 006 +

Potato 18 150 011 011 105 008 NS +

Turnip 6 1 0 010 001 008 minus

Invertebrate 41 454 034 006 532 010 +

Herbaceous plants 14 25 002 034 005 004 minus

Tuber 8 47 004 022 016 005 minus

Dec Grain 20 1648 089 006 1507 009 +

Potato 10 80 004 007 062 008 minus

Turnip 0 0 0 014 0 006 minus

Invertebrate 8 30 002 004 043 012 minus

Herbaceous plants 11 91 005 030 016 004 minus

Tuber 4 12 001 039 002 003 minus

Jan Grain 22 1200 08 003 2679 015 +

Potato 19 135 009 009 106 008 NS +

(continued on next page)

Dong et al (2016) PeerJ DOI 107717peerj1968 1119

Table 3 (continued)

No of videorecordings

No ofpecks

Oi πi Wi se P selection

Turnip 3 1 0 005 001 011 minus

Invertebrate 10 35 002 006 038 010 minus

Herbaceous plants 10 107 007 040 018 003 minus

Tuber 2 17 001 037 003 003 minus

Feb Grain 26 1100 073 003 2441 015 +

Potato 7 80 005 006 093 010 NS minus

Turnip 0 0 0 009 0 008 minus

Invertebrate 28 313 021 004 479 012 +

Herbaceous plants 5 7 0 036 001 003 minus

Tuber 2 2 0 042 0 003 minus

NotesNS P gt 0006+ Positive selectionminus Negative selection

Table 4 Pearson correlations between the environmental variables and invertebrate food variables for Black-necked cranes (G nigricollis) inthe Dashanbao National Nature Reserve China

Invertebratedepth

Invertebratenumbers (0ndash1 cm)

Invertebratenumbers (11ndash2 cm)

Mean temperature Correlation coefficients minus0721 0740 0690P-value 0043 0036 0058

Minimum temperature Correlation coefficients minus0730 0843 0775

P-value 0040 0009 0024Number of days with frozen ground Correlation coefficients 0779 minus0842 minus0797

P-value 0023 0009 0018

NotesCorrelation is significant at the 005 level (2-tailed)Correlation is significant at the 001 level (2-tailed)

Canonical correspondence analysis (CCA) exhibited the relationship betweenenvironmental factors and grain selection potato selection and invertebrate selectionin different patterns (Fig 3) The eigenvalues for the first two axes in Fig 3 were 0223 and0007 respectively The food selection-environmental correlations for the first two axeswere 0986 and 0714 respectively The first two axes of the CCA explained 966 of thetotal variance in food selection data and food variables of which 938 was contributedby the first axis and 288 by the second axis Invertebrate availability (077) potatoavailability (065) and grain availability (053) were positively associated with the firstaxis while the distributed depths of invertebrate (minus078) the depths of potato (minus068)and the depths of grain (minus049) were negatively associated with the first axis CCA axes1 and 2 separated the food selection into groups for grain selection potato selection andinvertebrate selection Invertebrate selection was positively associated with invertebrateavailability and was negatively associated with the invertebrate depths Potato selection wasnegatively associated with invertebrate depths Grain selection was positively associatedwith invertebrate depth followed by potato and grain depths which were negatively

Dong et al (2016) PeerJ DOI 107717peerj1968 1219

Figure 3 Canonical Correlation Analysis (CCA) showing the relationship between environmentalvariables and selection for grain potato and invertebrates Environmental variables are represented byarrows and their abbreviation Ia Invertebrate availability Pa Potato availability Ga Gain availability IdInvertebrate depth Gd Grain depth Pd Potato depth The x-axis refers to Axis 1 and the y-axis refers toAxis 2

associated with invertebrate availability As invertebrate availability seasonally decreasedcranes increased their grain consumption

DISCUSSIONDiet compositionThe variation in diet of the Black-necked Cranes was systematically studied for the firsttime using video recording The results revealed that the wintering diet of the Black-neckedCrane in the Dashanbao Reserve mainly consisted of domestic crops (eg grains andpotatoes) and invertebrate animals Turnips and wild plants foods (such as herbaceousplants and tubers) accounted for a much lower proportion of their diet These results aresimilar to those of a previous report in which fecal analysis was used to study the cropand wild plant consumption of a subpopulation of Black-necked Cranes wintering at theYarlung Zangpo Valley National Natural Reserve However the report on the cranes inthe Yarlung Zangpo Valley National Natural Reserve did not calculate the proportionof animal-based food (Bishop amp Li 2001) It is important to note that initial estimatesapproximated that 1396 of the Dashanbao Black-necked Cranersquos diet would consistof invertebrates In comparison animal matter comprises less than 10 of the diet forCommon Cranes in the Holm Oak Dehesas (Avileacutes Saacutenchez amp Parejo 2002) and 2ndash3 ofthe diet for various crane species in different regions of the world (Irene 1980 Reineckeamp Krapu 1986) Certain crane species feed primarily on animal matter while wintering insome sites These include the Lesser Sandhill Crane (G canadensis canadensis) (Davis ampVohs 1993) Whooping Crane (G americana) (Pugesek Baldwin amp Stehn 2013) and Red-crowned Crane (G japonensis) (Li et al 2014) Demoiselle Cranes (Anthropoides virgo)(Sarwar et al 2013) Florida Sandhill Cranes (G c pratensis) (Rucker 1992) and CommonCranes show similar preferences for invertebrates (Avileacutes Saacutenchez amp Parejo 2002)

Current research on the proportion of animal-based foods in the diet of Black-neckedCranes has solely focused on describing species (Han 1995 Hu et al 2002 Li amp Li 2005Liu Yang amp Zhu 2014b) Thus there is a need for additional quantitative investigations

Dong et al (2016) PeerJ DOI 107717peerj1968 1319

into the Black-neckedCranes feeding habits including invertebrate consumption Likewisemore data are needed to study the feeding habits of Black-necked Cranes over a greaterdistribution of locations This would greatly enhance our understanding of the dietaryhabits of this species

Previous studies using fecal analysis to assess the proportion of the mentioned foodcategories in the Black-necked Cranersquos diet have produced results inconsistent with ourstudy These studies largely reported a wild plant diet (leaves roots and tubers) while failingto mention the inclusion of domestic crops or invertebrates in the diet of cranes in theDashanbao Reserve (Liu et al 2014a) This inconsistency has two possible explanationsthe method to analyze the data and the sampling procedures First different methods wereused to analyze the diet With fecal analysis wild plant fiber may therefore have been easierto detect in feces than the potato and grain fibers or invertebrate larvae residues despite thelatter two making up a larger proportion of the diet Liu et al (2014a) mentioned potatocuticles were not detectable in the fecal sample of a crane that due to the digestibility ofthe food type With video observation we were able to directly estimate the frequency onwhich a particular food type was fed on without concern for variations in digestibilityWhile video observation enables the detection of even highly digestible food it is oftenmore difficult to identify the specific food types that are seen consumed Thus it requiresmore careful observation and detection of feeding patterns to identify food items This mayalso be seen as an advantage as it can provide us with more complete foraging informationincluding actual foraging behavior We are thus able to successfully estimate the digestiblecompositions of a birdsrsquo diet (Robinson amp Holmes 1982 Rundle 1982) Second our resultsinfer that the sampling time may have greater impact on identifying food types whichchange with monthly variations For example as a climate-restricted food invertebratesare difficult for Black-necked Cranes to find in December and January (Table 2) (seebelow discussion) Fecal analysis of Black-necked Cranersquos diet in the previous study didnot mention sampling time in Dashanbao Reserve (Liu et al 2014a) It is possible thatdifferent sampling times caused the difference from our results

Monthly variation and diet selectionIn November a high proportion of the Black-necked Cranersquos diet consisted of domesticcrops (principally grain) and invertebrate organisms (Table 1) This may be becausethe availability of those food types was the highest immediately after the birds arrived(November) The birds require a balanced diet including a variety of nutrients fromdifferent food types In November when both grains and invertebrates were most availableinvertebrates were consumed more than at any other time In contrast grains wereconsumed less than in othermonths This suggests that the cranes likely prefer invertebratesover grains potentially because invertebrate organisms provide a greater source of proteinand calcium than available in grains These nutrients are essential for their migration fitnessand overall survival Cranes consumed only a minimal quantity of wild plants despite theirlarger proportion of available biomass as compared to that of domestic crops and animalmatter (Table 2) It is possible that cranes prefer domestic crops or animal matter overwild plants because (1) herbaceous plants may have lower caloric content than grains or

Dong et al (2016) PeerJ DOI 107717peerj1968 1419

animal matter (2) there is insufficient density of vegetation suitable for the cranes to foragepreferred species such as Pedicularis Stellaria Polygonatum and Veronica (Kong et al 2011Liu et al 2014a)

Environmental factors compared to food selectionBased on the results of our CCA the grain selection and invertebrate selection present twodifferent patterns Grain selection was positively correlated with invertebrate depth andnegatively correlated with invertebrate availability However invertebrate selection showsthe opposite pattern Falling temperatures and freezing soils reduced the availability ofinvertebrates and increased the depth of invertebrates especially for December and January(Table 4) Therefore cranes primarily fed on grains during December and January andfed on invertebrate animals in November and February Potato selection was negativelyassociated with the depths of invertebrates When invertebrates are at increased depth dueto low temperatures the cost of digging for potatoes also increases as a result of frozen soil

Management implicationsOur results support previous reports that Black-necked Cranes generally prefer farmlandsand avoid grasslands (Kong et al 2011a) likely due to the availability of domestic crops andinvertebrates to feed on as well as other habitat featuresWe agree with Kongrsquos views (2011)that higher quantities and densities of food as well as looser soil structure in farmlandsfacilitate food collection by the cranes During colder weather (December or January) theinvertebrate shortage is exacerbated We recommend that the protection administrationshould supplement additional foods for cranes during the cold-weather periods and restoregrassland foraging habitat This would support the cranesrsquo need for dietary diversity andwould benefit the farmers by reducing economic losses resulting from the cranes feedingon newly planted crop seeds during their late spring migration (in March) To furtherease the conflict between cranes and local farmers it is advisable to cultivate crops in acertain area that may be left unharvested for the cranes to eat Furthermore it is necessaryto maintain adequate traditional croplands to sustain this vulnerable species as many ofthese conventional cultivations (grains potatoes and turnips) have been replaced by moreeconomic crops (Lepidium meyenii Walp) in the Dashanbao Reserve

ACKNOWLEDGEMENTSWe thank Shimei Li and Yuanjian Zhen for their help in our field work and staff ofDashanbao National Nature Reserve for their valuable support in the field

ADDITIONAL INFORMATION AND DECLARATIONS

FundingThe study was supported by the ICF (International Crane Foundation) The funders hadno role in study design data collection and analysis decision to publish or preparation ofthe manuscript

Dong et al (2016) PeerJ DOI 107717peerj1968 1519

Grant DisclosuresThe following grant information was disclosed by the authorsICF (International Crane Foundation)

Competing InterestsThe authors declare there are no competing interests

Author Contributionsbull Hao Yan Dong conceived and designed the experiments performed the experimentsanalyzed the data contributed reagentsmaterialsanalysis tools wrote the paperprepared figures andor tables reviewed drafts of the paperbull Guang Yi Lu performed the experiments analyzed the data reviewed drafts of the paperbull Xing Yao Zhong performed the experiments contributed reagentsmaterialsanalysistoolsbull Xiao Jun Yang conceived and designed the experiments performed the experimentsanalyzed the data contributed reagentsmaterialsanalysis tools reviewed drafts of thepaper

EthicsThe following information was supplied relating to ethical approvals (ie approving bodyand any reference numbers)

Our research on Black-necked Cranes in Dashanbao National Nature Reserve wasapproved by the Chinese Wildlife Management Authority and conducted under Law of thePeoplersquos Republic of China on the Protection of Wildlife (August 28 2004)

Field Study PermissionsThe following information was supplied relating to field study approvals (ie approvingbody and any reference numbers)

The Administration of ZhaoTong Forestry Bureau approved our study on behaviorobservation and sampling collection in the research plot in Dashanbao National NatureReserve (IDZTL2008163)

Data AvailabilityThe following information was supplied regarding data availability

The raw data was supplied as Supplemental Information

Supplemental InformationSupplemental information for this article can be found online at httpdxdoiorg107717peerj1968supplemental-information

REFERENCESAlonso JC Alonso JA Bautista LM 1994 Carrying capacity of staging areas and

facultative migration extension in common cranes Journal of Applied Ecology31212ndash222 DOI 1023072404537

Dong et al (2016) PeerJ DOI 107717peerj1968 1619

Avileacutes JM Saacutenchez JM Parejo D 2002 Food selection of wintering common cranes(Grus grus) in holm oak (Quercus ilex) dehesas in south-west Spain in a rainy seasonJournal of Zoology 25671ndash79

Baubet E Bonenfant C Brandt S 2004 Diet of the wild boar in the French AlpsGalemys 16101ndash113

BishopMA Li FS 2001 Effects of farming practices in Tibet on wintering Black-neckedCrane (Grus nigricollis) diet and food availability Chinese Biodiversity 10393ndash398

Davis CA Vohs PA 1993 Role of macroinvertebrates in spring diet and habitat use ofsandhill cranes Transaction of the Nebraska Academy of Sciences XX81ndash86

DowdyW 1937 The hibernation of certain arthropod fauna of the soil Proceedings ofthe Missouri Academy of Science 3116ndash117

Ellis DH Archibald GW Swengel SR Kepler CB 1991 Compendium of crane behaviorPart 1 individual (nonsocial) behavior In Harris JT ed Proceedings of the 1987International crane workshop Baraboo Wise International Crane Foundation225ndash234

Han LX 1995 Habitat status and conservation of cranes in Yunnan In Chen YY edWetland research of China Changchun Jilin Science amp Technology Press 256ndash261

Harris J Mirande C 2013 A global overview of cranes status threats and conservationpriorities Chinese Birds 4189ndash209 DOI 105122cbirds20130025

Hu JSWu JL Dang CL Zhong XY DaoMB 2002 A Study on the population ecologyof wintering Black-necked Cranes (Grus nigricollis) at Dasangbao Reserve ZhaotongYunnan Province Journal of Yunnan University 24140ndash143 146

Irene AN 1980 The cranes must live In Lewis JC ed Proceedings of the internationalcrane symposium Sapporo Japan 13ndash14

Kong DJ 2008 Studies on wintering behavior and conservation of Black-necked CranesGrus nigricollis at Dashanbao Yunnan China Masterrsquos Degree Thesis GraduateSchool of Chinese Academy of Sciences

Kong DJ 2011 Studies on vocal behavior and conservation of Black-necked Cranes Grusnigricollis China Doctor dissertation Chinese Academy of Sciences Beijing

Kong DJ Yang XJ Liu Q Zhong XY Yang JX 2011Winter habitat selection by theVulnerable Black-necked Crane Grus nigricollis in Yunnan China implicationsfor determining effective conservation actions Oryx 45258ndash264DOI 101017S0030605310000888

Kuang FL Zhong XY DaoMB Ji YS Liu N 2008 Initial observation on selection fbrforage location for forage location of Grus nigricollis in Dashanbao Forest Inventoryand Planning 33(4)75ndash77

Kushlan JA 1978 Feeding ecology of wading birds In Sprunt IV A Ogden JCWinckler S edsWading birds Research Report Number 7 New York NationalAudubon Society 249ndash297

Dong et al (2016) PeerJ DOI 107717peerj1968 1719

Li DL Ding YQ Yuan Y Lloyd H Zhang ZW 2014 Female tidal mudflat crabsrepresent a critical food resource for migratory Red-crowned Cranes in the YellowRiver Delta China Bird Conservation International 24416ndash428DOI 101017S0959270913000555

Li Z Li F 2005 Research on the Black-necked Crane Shanghai Shanghai ScienceTechnology and Education Press (in Chinese with English preface and summary)

Li R Li H 2012 Eco-life form of plants from Dashanbao Black-necked Crane NationalNature Reserve Journal of West China Forestry Science 169ndash76

Li FS Nie H 1997Microscopic analysis on herbivorous diets of wintering Black-neckedCranes at Cao Hai China Zoological Research 1851ndash57

Li FS Yang F 2002 Population numbers and distribution of Black-necked Cranes (Grusnigricollis) in the Yungui Gaoyuan Plateau Chinese Journal of Zoology 3843ndash46

Li H Zhong XY 2010 Plants of Dashanbao Black-necked Cranes National Nature ReserveBeijing Science Press

Liu Q Yang XJ Zhu JG 2014 Animal food items of wintering Black-necked CranesZoological Research 35197ndash200

Liu Q Zhu XM Li NY Liu ZP Zhong X 2014Microhistological analysis of winteringBlack-necked Cranes herbivorous diets at Dashanbao Wetland China ZoologicalResearch 35201ndash204

Manly BF McDonald L Thomas D McDonald TL EricksonWP 1993 Resourceselection by animals Statistical design and analysis for field studies London Chapmanand Hall

Martin TE 1985 Resource selection by tropical frugivorous birds integrating multipleinteractions Oecologia 66563ndash573 DOI 101007BF00379351

McCollogh J HayesWP Bryson HR 1927 Preliminary notes on the depth of hiber-nation of wireworms (Elateridae Coleoptera) Journal of Economic Entomology20561ndash564 DOI 101093jee204561

Newton I 1967 The adaptive radiation and feeding ecology of some British finches Ibis10933ndash96 DOI 101111j1474-919X1967tb00005x

Pielou EC 1966 The measurement of diversity in different types of biological collectionsJournal of Theoretical Biology 13131ndash144 DOI 1010160022-5193(66)90013-0

Price T 1987 Diet variation in a population of Darwinrsquos finches Ecology 681015ndash1028Pugesek BH BaldwinMJ Stehn T 2013 The relationship of blue crab abundance

to winter mortality of Whooping Cranes The Wilson Journal of Ornithology125658ndash661 DOI 10167612-1591

Qian FWWuH Gao L Zhang H Li F Zhong X Yang X Zheng G 2009Migrationroutes and stopover sites of Black-necked Cranes determined by satellite trackingJournal of Field Ornithology 8019ndash26 DOI 101111j1557-9263200900201x

Redpath SM Clarke R Madders M Thirgood SJ 2001 Assessing raptor diet comparingpellets prey remains and observational data at hen harrier nests The Condor103184ndash188 DOI 1016500010-5422(2001)103[0184ARDCPP]20CO2

Reinecke KJ Krapu GL 1986 Feeding ecology of sandhill cranes during springmigration in Nebraska The Journal of Wildlife Management 5071ndash79

Dong et al (2016) PeerJ DOI 107717peerj1968 1819

RiceWR 1989 Analyzing tables of statistical tests Evolution 43223ndash225DOI 1023072409177

Robinson SK Holmes RT 1982 Foraging behavior of forest birds the relationshipsamong search tactics diet and habitat structure Ecology 631918ndash1931

Rucker CR 1992 Food and feeding habits of released Florida sandhill cranes ProceedingsNorth American Crane Workshop 685ndash89

RundleWD 1982 A case for esophageal analysis in shorebird food studies Journal ofField Ornithology 53249ndash257

SarwarM Hussain I Khan A AnwarM 2013 Diet composition of the Demoisellecrane (Anthropoides virgo) migrating through Lakki Marwat Pakistan Avian BiologyResearch 6269ndash274 DOI 103184175815513X13802893287049

Savage RE 1931 The relation between the feeding of the herring off the east coast ofEngland and the plankton of the surrounding waters Fishery Investigation Ministryof Agriculture Food and Fisheries 121ndash88 Series 2

StapanianMA Smith CC Finck EJ 1999 The response of a Kansas winter bird commu-nity to weather photoperiod and year The Wilson Bulletin 111550ndash558

Ter Braak CJF Smilauer P 1998 CANOCO reference manual and userrsquos guide to canocofor windows software for canonical community ordination (version 4) WageningenCentre for Biometry 351 pp

Ter Braak CJF Smilauer P 2002 CANOCO reference manual and canodraw for windowsuserrsquos guide software for canonical community ordination (version 45) IthacaMicrocomputer Power

Ter Braak CJF Verdonschot PFM 1995 Canonical correspondence analysis andrelated multivariate methods in aquatic ecology Aquatic Science 37130ndash137DOI 101007BF00877430

Yoshikawa T Osada Y 2015 Dietary compositions and their seasonal shifts in Japaneseresident birds estimated from the analysis of volunteer monitoring data PLoS ONE10e0119324 DOI 101371journalpone0119324

Zhong XY DaoMB 2005 Cranes of the world In Li FS Yang XJ Yang F eds Statusand conservation of Black-necked Cranes on the Yunnan and Guizhou Plateau PeoplersquosRepublic of China Kunming Yunnan Nationalities Publishing House 101ndash106

Zhou YB ChenWW Kaneko Y Newman C Liao ZH Zhu XQ Buesching CD Xie ZQMacdonald DW 2015 Seasonal dietary shifts and food resource exploitation by thehog badger (Arctonyx collaris) in a Chinese subtropical forest European Journal ofWildlife Research 61125ndash133 DOI 101007s10344-014-0881-5

Dong et al (2016) PeerJ DOI 107717peerj1968 1919