12 pages

Habitat and Grazing Influence on Terrestrial Ants in Subtropical Grasslands and Savannas of Argentina

of 12
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Habitat and Grazing Influence on Terrestrial Ants in Subtropical Grasslands and Savannas of Argentina
  C ONSERVATION  B IOLOGY AND  B IODIVERSITY Habitat and Grazing Influence on Terrestrial Ants in SubtropicalGrasslands and Savannas of Argentina  LUIS A. CALCATERRA, 1 SONIA M. CABRERA, 1 FABIANA CUEZZO, 2 IGNACIO JIME´ NEZ PERE´ Z, 3 AND  JUAN A. BRIANO 1 Ann. Entomol. Soc. Am. 103(4): 635Ð646 (2010); DOI: 10.1603/AN09173 ABSTRACT  ThemaintenanceofspeciesdiversityinmodiÞedandnaturalhabitatsisacentralfocusof conservation biology. The Ibera  ´   Nature Reserve (INR) protects highly diverse ecosystems innortheasternArgentina,includingoneofthelargestfreshwaterwetlandsinSouthAmerica.Livestockgrazing is one of the major disturbances to these ecosystems; however, its effect on ant diversity ispoorly known. The objective of this work was to study the effect of savanna versus grassland andgrazing on the structure and composition of subtropical terrestrial ants focusing on the particularresponseofthenativeredimportedÞreant, Solenopsisinvicta Buren.Unbaitedpitfalltrapswereusedto capture worker ants in 25 grazed and 20 nongrazed sites. Fifty ant species were collected. Thesavanna showed more species, individuals, biomass, and functional groups of ants than the grassland. S .  invicta  was the most frequently captured (61.4%) and numerically dominant species; however Camponotus punctulatus punctulatus  Mayr showed the highest biomass. Grazing simpliÞed vegetationstructureinbothhabitats,butitsimpactonvegetationseemedonlytopromoteahighertotalbiomassespecially in the grassland, and/or functional groups, favoring occurrence of hot-climate specialistsinthesavanna.Thisstudyrevealedthathabitattypestronglyaffectedtheorganizationoftheterrestrialant assemblages at the INR. However, as in other studies, we did not Þnd clear evidence that habitatmodiÞcation by grazing signiÞcantly affected terrestrial ant assemblages. The weak grazing inßuencecould be the consequence of the short enclosure time as to recover the srcinal ant communities, thedifferential response of ant species to habitat type, and/or the resilience of ants. KEY WORDS  richness, species composition, disturbance, vegetation structure,  Solenopsis invicta Ants represent a high proportion of the biomass anddiversity on earth. They constitute an important partof the edaphic mesofauna that enriches the soil, dis-perses seeds, tends homopterans, and controls insects(Wilson1971,FittkauandKlinge1973,Ho  ¨  lldoblerandWilson 1990). The structure and composition of antcommunities are inßuenced by competition, naturalenemies, resource availability, habitat, and distur-bance (Wilson 1971; Ho  ¨  lldobler and Wilson 1990;Bestelmeyer and Wiens 1996, 2001; Andersen 1997;Kaspari and Majer 2000).Habitatdegradationandbiologicalinvasionsarethetwo greatest threats to global biodiversity. The main-tenance of species diversity in modiÞed and naturalhabitats is a central focus of conservation biology. Inthis regard, the Ibera  ´   Nature Reserve (INR) repre-sents a unique system of 1,300,000 ha in northeasternArgentina that protects one of the largest freshwaterwetlands in South America (Canziani et al. 2003,Parera 2006, The Ramsar Convention on Wetlands2008). The reserve includes a large marsh with la-goons, small rivers, streams, and extensive ßoatingvegetation, surrounded by a diverse mosaic of terres-trialecosystems,suchastemporaryßoodedgrasslands,savannas, and forests (Neiff and Poi de Neiff 2006,Parera 2006). These terrestrial ecosystems show dif-ferent degrees of modiÞcation through ranching, ag-riculture, and wood production (Parera 2006). Morethan 100 native ant species occur in the INR (Calca-terraetal.2010),severalofwhichareinvasiveinothercountries(Loweetal.2000,PestDirectory2008),suchas the red imported Þre ant,  Solenopsis invicta  Buren,which was accidentally introduced into the UnitedStatesfromnortheasternArgentinaintheearly1900s.This species occurs in several natural habitats of theINR (Calcaterra et al. 2010).Natural and anthropogenic disturbances observedintheINR,suchasßoods,Þres,orgrazing,areessentialfactorsaffectingspeciesstructureandcomposition,asin many natural ecosystems (Cooperrider 1991, Su-arez et al. 2001, Adis and Junk 2002, Parera 2006). Asone of the most extensive forms of land use (FAO1996), cattle grazing is one of the main potentialthreats to biological diversity in the INR. Its effect on 1 Corresponding author: USDAÐARS, South American BiologicalControl Laboratory, Bolõ ´  var 1559 (B1686EFA) Hurlingham, BuenosAires, Argentina (e-mail: 2 CONICET-InstitutoSuperiordeEntomologõ ´  a(INSUE),Facultadde Ciencias Naturales e Instituto Museo Lillo, Miguel Lillo 205(T4000JFE), San Miguel de Tucuma  ´  n, Argentina. 3 TheConservationLandTrustArgentina,Cuba3129,Apartado15,(1429) Ciudad de Buenos Aires, Argentina.0013-8746/10/0635Ð0646$04.00/0  2010 Entomological Society of America   many biological components of wetlands has beenwidely studied in other regions (Reeves and Cham-pion 2004, Marty 2005, Casanova 2008); however, fewstudiesinvestigateditseffectonantfaunaindifferenthabitattypes.Dependingongrazingintensity,habitatcomplexity, and hierarchies in the ant community(Farji-Breneretal.2002),grazingmayhavepositiveornegativeeffectsonants(BestelmeyerandWiens1996,2001;KerleyandWhitford2000;Woinarskietal.2002;Hoffmann and Andersen 2003; Ratchford et al. 2005).Many studies have reported that intensive grazingaffects habitat structure facilitating the dominance of highlycompetitivespecies(suchastheinvasiveant S.invicta ), but with scarce consequences on the orga-nization of assemblages of ants (Orians 1986, Deyrupet al. 2000, Hoffmann and Andersen 2003). To knowwhether the presence of livestock grazing affects theorganization of the ant assemblages of the INR, it isessential to establish future management policies thatcontributetothebiodiversitymaintenanceofthispro-tected area.The main objective of this work was to study theinßuence of the habitat type (savanna versus grass-land)andgrazing(nongrazedversusmoderatetoveryheavy grazing) on the structure and composition of subtropicalterrestrialantassemblagesoftheINR.Wealsoinvestigatedtheresponseoftheinvasive S.invicta to grazing to conÞrm whether or not this Þre ant ismore abundant in disturbed (grazed) habitats in itsnative range, as has been reported for its introducedrange, mainly in North America (King and Tschinkel2006, 2008; Tschinkel 2006; Hill et al. 2008). Materials and MethodsStudy Area.  The study was conducted in June (latefall) and November (spring) 2007 and September(late spring) 2008 in an area of    300 km 2 mostlylocatedinthesoutheastoftheINR,5kmsouthwestof ColoniaPellegrini(Fig.1).Studysiteswerelocatedintwo habitats: 1) savanna, an open forest belonging tothe  “ Espinal ”  ecoregion, with short  Prosopis  L. and  Acacia (Tourn.)trees,bushes,andgrasses(Tressensetal.2002);and2)grassland,atemporarilyßoodedgrass-land, typical of the  “ Campos y Malezales ”  ecoregion,dominated mainly by  Andropogon lateralis  Nees(Tressens et al. 2002, Neiff and Poi de Neiff 2006,Parera2006).Ineachhabitat,wesampledbothgrazedand nongrazed areas with cattle exclusion for  5 yr(strict conservation areas). Besides cattle, the regionharbors wild grazers such as capybara ( Hydrochaerishydrochaeris  L.), plains viscacha ( Lagostomus maxi-mus  Desmarest), marsh deer ( Blastocerus dichotomus Illiger), and brocket deer ( Mazama gouazoupira  Fi-scher).The climate in the region is humid and subtropicalwith drier winters. Mean temperature varies from15  C in July to 26  C in January. Mean annual rainfallis 1,500 mm (De Fina 1992, Fontan and Sierra 2004). Fig. 1.  Studyareashowingthelocationofthe39samplingstationswithintheIbera  ´   NatureReserve(sixadjacentstationsat El Porvenir ranch are not included in the map). Abbreviations refer to nongrazed savanna (S), grazed savanna (Sg),nongrazed grassland (G), and grazed grassland (Gg). 636 A NNALS OF THE  E NTOMOLOGICAL  S OCIETY OF  A MERICA  Vol. 103, no. 4  SamplingDesign. Intotal,45samplingstationswereestablished as follows: 1) nongrazed savanna: 10 sta-tions; 2) grazed savanna: 14 stations; 3) nongrazedgrassland: 10 stations; and 4) grazed grassland: 11 sta-tions (Fig. 1). The nongrazed savanna stations andthose in the grassland were located in two privatereserves owned by The Conservation Land Trustwithin the INR, whereas all the grazed savanna sta-tions were located in Paraje Uguay (eight stations),andElPorvenirRanch(sixstations),30kmsouthfromthe other sampling areas (not included in Fig. 1).All stations were located at least 1Ð2 km apart and  300 m away from habitat limits (ecotones), roadsand trails to avoid border effect. According to theranchers,theaveragelivestockloadsduringthepast5yr were as follows: Paraje Uguay, four animals per ha (very heavy grazing; cows, sheep, and horses); ElPorvenir Ranch, 0.8 animal per ha (heavy grazing;cows); and Ibera  ´   Ranch, 0.4 animal per ha (moderategrazing;cows).Thus,thesavannapresentedeightsiteswith very heavy grazing and six with heavy grazing,and the grassland had all 11 sites with moderate graz-ing. To evaluate the effect of grazing on vegetationrichness and structure (height and cover) and on theant assemblage structure, we used srcinal data fromavegetationsurveycarriedoutbyMaturoetal.(2007)in 25 of the 45 sampling stations used in this study,mainlylocatedinthesavanna.Thisvegetationanalysisdid not include information from El Porvenir Ranch. Ant Survey.  Ants were sampled using unbaited pit-fall traps following Calcaterra et al. (2010). At eachstation, one trap was set up every 10 m along a lineartransect and exposed for 48 h. Each trap consisted of a 50-ml plastic centrifuge tube (  3 cm in diameter)buried in the ground and half-Þlled with soapy water.In total, Þve traps were considered the sampling unit.After 48 h, the content of the traps was removed,rinsed with water, and preserved in 96% ethanol. Allstationsweresampledonce.Thus,intotal45samplingunits (225 pitfalls traps) were considered. Althoughthe habitats were sampled simultaneously, the sam-pling effort differed (Tables 1 and 2).Ants were separated from other arthropods andidentiÞed to species or morphospecies by L.A.C. andF.C. under a dissecting scope and by using keys as inCalcaterra et al. 2010. Voucher specimens were de-posited at the Instituto-Fundacio ´  n Miguel Lillo(IFML),Tucuma  ´  n,Argentina,andattheSouthAmer-ican Biological Control Laboratory (SABCL). Data Analysis.  We used a combination of diversityindices(antspeciesrichness,relativeabundance,bio-mass, Shannon, dominance, similarity, and composi-tion) to study the effects of habitat and livestockgrazing on ant assemblages. Because we were notinterested in temporal effects on the ant assemblages,samplingperiodsdidnotcoverallseasonsandsampleswere grouped for the analysis. The number of antspecies at each station was recorded by pooling Þvepitfall traps (sampling unit). Spatial abundance wasmeasuredastheproportionofunitsinwhichaspecieswas present. Numerical abundance was measured as1)thenumberofindividualstrappedand2)theequiv-alent biomass. Biomass was estimated for the 26 mostcommon species by multiplying the number of indi-viduals by the individual dry weight after exposed to60  C for 48 h. Data on vegetation richness and struc-turepersamplingstationwereextractedfromMaturoet al. (2007). Relationship between the numericalabundances of the six spatially more abundant antspecies and between these and the vegetation vari-ableswereanalyzedusingSpearmanrankcorrelation.Shannon (H  ) and Dominance (D) indices (Magur-ran 1988, Gotelli and Colwell 2001) also were used asestimators of diversity and dominance for compari-sons with other studies. Indices were calculated usingthe number of individuals of each species per samplewith the statistical package Past version 1.82 (Ryan etal. 1995). A two-way Þxed factor analysis of variance(ANOVA) was used to analyze the effect of habitat,grazing, and their interaction on richness (square-root  0.5 transformed), number of workers (square-root  0.5 transformed), Shannon index, Dominanceindex(1/square-root  0.5transformed),andbiomass(log  1transformed).ThistestandtheKruskalÐWal-lis nonparametric ANOVA also were used to evaluatedifferences in number of workers (log    1.5 trans-formed) among treatments for the six numerically Table 1. Number of sampling units (five traps) per habitat inIberá Sampling periodSavanna GrasslandTotalNongrazed Grazed Nongrazed GrazedJune 2007(late fall)2 3 3 3 11Nov. 2007(spring)3 5 3 4 15Sept. 2008(late winter)5 6 4 4 19Total 10 14 10 11 45 Table 2. Sampling effort and diversity indicators in Iberá Savanna GrasslandTotalNongrazed Grazed Nongrazed GrazedNo. stations(or samples) a 10 14 10 11 45Pitfall traps 50 70 50 55 225Expectedspecies b 44 68 27 24 75Observed species(% expected)25 (57) 35 (51) 16 (59) 13 (54) 50 (67)Max. species persample8 10 7 5 10Commonspecies c 8 11 4 5 5Rare species d 13 20 8 7 18Exclusivespecies e 5 14 2 5Invasive species  f   6 8 7 3 10 a A sample consists of 5 pitfall traps buried in the ground for 48 h. b Estimated richness value with the most stable estimator (Jack2). c Observed species in  20% of the samples. d Observed species in only one sample. e Observed species in only one treatment.  f   Invasive species or pest (Lowe et al. 2000, Pest Directory 2008). July 2010 C ALCATERRA ET AL .: H ABITAT AND  G RAZING  I NFLUENCE ON  A NTS  637  dominant species. A chi-square test was used to eval-uate the inßuence of grazing on the total number of workers in each habitat.The total numbers of species found in each habitatand grazing history was visually compared using sam-pled-based species accumulation curves (Gotelli andColwell 2001) by entering occurrence data in a two-way matrix (species  sample) and using the Estima-tesS 8.0 software (Colwell 2006). Curves were ob-tained after 500 randomizations. To quantitativelyensure that sampling effort was similar in the fourtreatments, the total number of species observed ineach situation was divided by the total number of species expected to occur in each situation accordingtothenonparametricsecond-orderJackkniferichnessestimator (henceforth Jack2) (Table 2). This estima-torwasselectedbecauseofitsbestperformanceundera wide range of sample sizes.A two-dimensional ordination of samples was car-ried out with nonmetric multidimensional scaling(NMDS) by using the So  ¨  rensen distance measure tovisuallyexaminetheantassemblagecomposition;thiswas tested using an Analysis of Similarity (ANOSIMandposthocBonferronipairwisecomparisons;ClarkeandGreen,1988)basedon1,000permutations.NMDSiswellsuitedfornon-normaldataanddoesnotassumelinear relationships among variables (McCune andGrace 2002). The So  ¨  rensen similarity index also wasused to evaluate the similarity in species composition(Magurran1988).Becauserarespeciesmaydistorttheanalyses,onlyspecieswithmorethanfourindividualsoccurring in more than one sample were consideredfor the analyses, which were performed using PC-ORD 4.0 software (McCune and Mefford 1999).Functionalcompositionwascomparedbyassigningspeciestofunctionalgroupsbasedonhabitatrequire-ments, foraging mode, competitive interactions, andresponses to environmental stress and disturbance(Andersen 1995, 1997; Bestelmeyer and Wiens 1996; King et al. 1998; Andersen et al. 2007; Hill et al. 2008;van Ingen et al. 2008; Calcaterra et al. 2010). Wefollowed the classiÞcation of functional groups pro-posed by Bestelmeyer and Wiens (1996) for Argen-tinean Chaco ants and previously adopted by Calca-terra et al. (2010) for the INR. This classiÞcation wasconsidered more appropriate for our study than thatproposedbyAndersen(1995)forAustralianantcom-munities. The following groups were considered inthis study: opportunists (unspecialized species thatare weak competitors), specialist predators (large-body species with small colony size and specialistdiet),subordinateCamponotini(large-bodyandoftennocturnal species, submissive to the dominant spe-cies),Attini(fungus-culturingantsincludingleaf-cut-ting and detritus-collecting ants), cryptics (minutespecies that live mostly in the litter, rotting logs, orsoil), generalized Myrmicinae (small-bodied specieswith a variety of behaviors and habitat requirementthat defend resources by recruitment), arboreals(species that nest and forage almost exclusively intrees or shrubs), and climate-specialist species (spe-cies adapted to speciÞc temperature and humidityconditions). Dominant Dolichoderinae (active anddominant ants) was not recorded in this study. Antspecies were assigned to these groups based uponhabitat and strata occurrence, presence, abundance,and behavior observed at baits; also, other works con-ducted mainly in South America on these and similarspecies with known natural history were considered(Kusnezov 1978; Bestelmeyer and Wiens 1996;Andersen 1997; Silvestre et al. 2003; Andersen et al.2007; Calcaterra et al. 2008, 2010; Hill et al. 2008; van Ingen et al. 2008; Cuezzo and Gonza  ´  lez Campero2010). A two-way ANOVA was used to examine theinßuence of habitat type, grazing history, and theirinteraction on the number of functional groups(square-root  0.5 transformed) and the number of species(square-root  0.5transformed)ineachfunc-tionalgroup.Aone-wayANOVAwasusedtoexaminewhether different grazing intensities in the savanna (nongrazed,Rinco ´  ndelSocorroRanch;heavygrazed,El Porvenir Ranch; and very heavy, Paraje Uguay)affect the richness, abundance, diversity, dominance,biomass, and functional group number of ants.Means  SE are reported. ResultsSpeciesRichness. Intotal,1,163antsbelongingto50species or morphospecies in 16 genera were captured(  Appendix  1). The most represented genera were Pheidole  Westwood with 16 species (32%) and  So-lenopsis  Westwood with seven (14%). Forty-threespecieswerecapturedinthesavannaand24speciesinthe grassland. Grazed and nongrazed sites showed 41and 30 species, respectively. Because none of thecurvesapproachedtheasymptoticlevel,morespecies(mainly rare species) should be expected to occur inallsituations(Fig.2).However,accordingtothemoststable richness estimator (Jack2), the percentages of expected species were relatively similar in all treat-ments, ranging from 51% for the grazed savanna (therichest treatment) to 59% for nongrazed grassland Fig. 2.  Cumulative observed species richness as a func-tion of sampling effort for each treatment: savanna non-grazed (  ), savanna grazed ( f ), grassland nongrazed ( ‚ ),and grassland grazed ( Œ ). Each sample consists of Þve un-baited pitfall traps 10 m distant from each other. Curves aregenerated from 500 randomizations. 638 A NNALS OF THE  E NTOMOLOGICAL  S OCIETY OF  A MERICA  Vol. 103, no. 4  (the poorest treatment). A high number of rare andexclusive species was observed in the grazed savanna (Table 2).The mean number of species captured per stationwas6  0.4inthesavannaand2.7  0.4inthegrassland( R 2   0.48;  F  1, 41    32.96;  P     0.001) (Table 3). Incontrast, grazing history ( F  1, 41  0.31;  P   0.582) andthe interaction habitat-grazing history ( F  1, 41  1.57; P     0.218) were not signiÞcant. In the savanna, themean ant richness was similar under different grazingintensities ( F  2, 21  1.44,  P   0.26) (Table 4).Although the Shannon ( F  1, 41  17.25;  P   0.001)and dominance ( F  1, 41    9.65;  P     0.004) indicesindicated a signiÞcant difference in the diversity anddominance of ant species between habitats (Table 3),the variance explained by the whole model were rel-atively low ( R 2  0.32 and 0.21, respectively). Again,grazing history and the interaction were not signiÞ-cant for these two indices ( P   0.35 in all the cases).Shannonanddominanceindiceswerealsosimilarunderdifferentgrazingintensitiesinthesavanna( F  2, 21  0.23; P   0.8 and  F  2, 21  0.25;  P   0.78, respectively). Spatial Abundance.  S. invicta  was the most widelydistributed ant, being captured in 64% (29/45) of thestations,followedby Paratrechinafulva (Mayr)(31%)and  Pheidole laevinota  Forel (29%). Also,  S. invicta dominated all habitats except the nongrazed grass-land, which were codominated by  P. fulva  and  P.laevinota  (40%) (Table 5).  Pheidole aberrans  Mayr(29%) and  Pheidole obscurithorax  Naves (22%) oc-curred exclusively in the savanna. Only Þve (10%) of the 50 species captured occurred at  20% of the sta-tions (Table 5). Numerical Abundance.  S. invicta  was the mostabundant with 309 workers (27%), followed by  P.laevinota  with 148 (13%), whereas  Camponotuspunctulatus punctulatus  Mayr showed the highest bio-mass with 95.9 mg (32%), followed by  Camponotuspunctulatus cruentus  Santschi with 55.4 mg (19%) and S. invicta  with 45.7 mg (15%).  S. invicta  and other 11species represented   80% of the total workers cap-tured and almost 90% of the total biomass (Table 6).Nongrazed savanna sites were dominated by  S. in-victa  with 46% of the workers and 36% of the biomass(Table 6). Grazed savanna was codominated by  Do-rymyrmex steigeri platensis  Santschi (17%, it occurredonlyinthissituation), P.laevinota (16%),and S.invicta (15%), but only by  C. punctulatus cruentus  (40%) interms of biomass. Nongrazed grassland sites werecodominatedby P.fulva (21%)and P.laevinota (20%),and by  Camponotus rufipes  (F.) in biomass (51%). Solenopsisinvicta (37%)and C.punctulatuspunctulatus (32%) codominated in the grazed grassland; in addi-tion,  C. punctulatus punctulatus  was clearly the dom-inant species in biomass (80%).Overall abundance of   S. invicta  was positively re-lated to  P. aberrans  and negatively to  C. punctulatuspunctulatus  ( r  s  0.36;  P   0.014 and  r  s  0.41;  P   0.006, respectively). Abundance of   P. fulva  was neg-atively related to this latter species ( r  s  0.33;  P   0.029).The mean number of workers per station was34.3    4.9 in the savanna and 16.1    4.1 in thegrassland ( F  1, 41    10.99;  P   0.002) (Table 3), butgrazing history was only marginally signiÞcant ( F  1, 41  3.85; P   0.057)andtheinteractionwasnotsigniÞcant. Table 3. Mean ant diversity estimators at the station level and vegetation variables Estimator/variableSavanna Grassland P   valueNongrazed( n  10)Grazed( n  14)Nongrazed( n  10)Grazed( n  11)AntsRichness 5.3  0.6a 6.5  0.5a 3  0.7b 2.5  0.5b   0.001Diversity (H  ) 1.17  0.1a 1.25  0.1a 0.73  0.22b 0.55  0.11b   0.001Dominance (D) 0.41  0.05a 0.41  0.05a 0.61  0.11b 0.67  0.07b   0.01No. workers 27.4  7.4a 39.3  6.4a 9.7  2.9b 22  7.2b   0.01Biomass (mg) 5.1  1.3a 7.5  2.7a 3.8  1.6a 9.4  4.7a N.S.VegetationRichness 28.8  2.4a 25.5  1.9a 21.3  2.2a 26.6  1.6a N.S.Ht (cm) 66.1  16.1a 11.1  4.1b 130  25.2 44.2  24.4   0.05Cover (%) 84.5  5.3a 71.5  4.7b 85  15 56  5.8   0.05Lowercase letters within columns indicate that a variable differed signiÞcantly among treatments. Table 4. Mean ant diversity estimators under different grazing intensities (nongrazed, Rincón del Socorro Ranch; heavy, El PorvenirRanch; and very heavy, Paraje Uguay) in the savanna EstimatorGrazing intensitySigniÞcanceNongrazed ( n  10) Heavy ( n  6) Very heavy ( n  8)Richness 5.3  0.6 7  0.8 6.1  0.7 N.S.Diversity 1.2  0.1 1.2  0.2 1.3  0.1 N.S.Dominance 0.41  0.05 0.45  0.09 0.37  0.05 N.S.No. workers 27.4  7.4 44.5  11.8 35.4  7.3 N.S.Biomass 5.1  1.3 6.3  2.1 8.4  4.6 N.S.No. functional groups 2.6  0.3 3.3  0.5 3  0.4 N.S. July 2010 C ALCATERRA ET AL .: H ABITAT AND  G RAZING  I NFLUENCE ON  A NTS  639
Related Documents
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!