Species Abundance Distribution Pattern of Microarthropod Communities in SW Canada.
It is still unclear whether simple niche-derived or neutrality-derived statistical models is better to quantify the experimental species-abundance distribution pattern (SAD) for microarthropod communities. In the present study by utilizing the sampling diversity data of three microarthropod taxonomic groups (oribatids collembolans and mesostigmatids) my objectives are to test and compare five alternative statistical models for fitting empirical microarthropod SAD curves including neutral Zipf broken stick niche preemption and geometric models. Fitting power of the models were evaluated using 2 test Kolmogorov-Smirnov (KS) test and Akaike Information Criterion (AIC). my results showed that for the SAD of the whole microarthropod community and mesostigmatid group Zipf model is the best model identified by AIC criteria. For oribatid and collembolan SAD curves geometric model is the most favored one. However all the models yielded significant difference between the expected and observed SAD patterns over different taxonomic groups as indicated by both 2 and K-S tests. Thus either neutral and niche models could explain SAD patterns of microarthropod communities perfectly. In summary the synergy of different mechanisms and the development of hybrid models and the proper transformation might be of some helps to remove the observed significant difference for microarthropod communities.
Keywords: Dominance-rank relationship neutral theory least-square minimization microcosm model system
Over the past ten years there was a hot debate on the deterministic role of niche and neutrality processes on influencing biological communities (Hubbell 2001; Nee and Stone 2003; Rosindell et al. 2011 2012; Munoz et al. 2012; Ricklefs and Renner 2012). Neutrality theory assumed no differentiation between species thus each species is functionally and physiologically equivalent (Hubbell 2001 2006). Species under neutral theory have identical birth death and mutation rates. Moreover neutral theory emphasizes the importance of stochasticity which has been usually overlooked in niche theory (Cheng et al.
The most striking support for neutral theory is the fitting of species abundance curve (SAD) which might yield the highest fit for the neutral model (Volkov et al. 2003 2007). However other simple models whether they were directly derived from niche theory or not also have remarabkly high fitting powers. Also even when neutral theory could
predict empirical SAD perfectly the underpinning mechanisms driving species assemblages are still niche-based (Cheng et al. 2011). SAD could not fully reflect the mechanisms structuring species communities.
The comparison of different statistical models on their powers for fitting SAD has been well quantified in recent studies (Walker and Cyr 2007; Du and Zhou 2008; Gao et al. 2011; Yan et al.
2012). In most cases simple statistical models are powerful enough to quantify SAD (Cheng et al. 2011) and the neutral model didn't have a remarkably better fit.
It is still not fully clear whether a simple niche model a sophisticated neutral model or none could be applied to the SAD of microarthropod communities although the universality of some common SAD statistical models is expected to prevail across species assemblage and sampling spatial scales. There are indeed some historic literature describing the species-abundance rank relationship (Cepeda-Pizarro and Whitford 1989; Coulson et al. 2003; Hoyle and Harborne 2005; Perdomo et al. 2012). However up to date there is no literature being found to compare the SAD patterns for different taxonomic groups of microarthropod and fit alternative statistical SAD theoretical modelsIn the present study by utilizing
the sampling data of microarthropod communities in SW Canada I evaluated the most suitable SAD models for major microarthropod taxonomic groups including oribatids collembolans and mesostigmatids and all species as a whole.
Dispersal ability is one of the major important mechanisms structuring the distribution and diversity and biogeographic patterns of species (Morrone and Crisci 1995; Potthoff et al. 2006; Gove et al. 2009; Chen 2013abc). Thus if some species have long-distance dispersal ability then the resultant species abundance and presence will be expected to be high because they can occur over different areas with high likelihood. Accordingly
long-distance dispersers can escape the influence of environmental filtering stresses and thus the corresponding SAD pattern should be neutrality
favored (Chen 2013a). In contrast when species assemblage has very restricted dispersal ability it is expected that these species are subjected to strong
environmental influence and correspondingly the niche-related SAD models might better fit the observed SAD pattern for short-distance dispersers.
Based on the above arguments I thus set up the following hypotheses to be tested in the study: oribatids are soil-dwelled and wingless therefore their abundances are presumed to be largely
influenced by dispersal limitation (Lindo and
Winchester 2007 2009; Chen 2013c). As such I predict that neutral model should be of highest power among the models to fit SAD for oribatids. In
contrast collembolans could disperse actively and passively via different vectors (Binns 1982; Szymkowiak et al. 2007; Lindo and Winchester
2009). I predicted this taxonomic group should be environmentally structured and therefore nice- based models might be better to fit the SAD. At last the corresponding model for the SAD of mesostigmatids is undetermined given that they have intermediate dispersal abilities.
MATERIALS AND METHODS
Sampling locations 32 moss field plots were surveyed across SW Canada based on the following standards of site selection: (1) they should be contiguous with the mainland (islands excluded); (2) they should be flattened large rocky outcrops with greater than 4m2 of moss carpets; (3) they should be accessed easily being adjacent to highway roads. 353 morphospecies were identified and a total number of 13260 individuals were counted. The abundance of each species was calculated and utilized in the subsequent analyses.
SAD statistical models
The existence of a new species in the community is influenced by the species arrived earlier (Cheng et al. 2011) thus the Zipf model (Frontier 1985) has the formula as Equation
Where Ni denoted the predicted abundance for the i- th species in the SAD N is the total individual number in the community q is the predicted relative abundance of the species with highest abundance in the community indicated the influence of priority effect.
Broken stick model (BSM)
BSM model (MacArthur 1957) has the expected abundance for the i-th species based as below Equation
Where a is the estimated scale parameter S is the total species number in the community.
Niche preemption model (NPM)
NPM (Motomura 1932) assumed that the percentage of the total niche occupied by the first species is the second one occupied a percentage a of the reminder being a(1- a) and so on As such the expected abundance for the i-th species is Equation
Geometric model (GEOM)
GEOM (Bastow 1991) is another form of niche preemption model but the formula is different since it has two independent parameters Equation
Neutral model (NM)
NM sampling formula is complex (Etienne 2005) for simplicity it is not present here. Two important parameters fundamental biodiversity index and migration rate m are fitted using the program Tetame version 2.1 (Jabot et al. 2008). Expected abundance of species were estimated by taking the means of 1000 simulations of neutral communities using the estimated m and total individual number J (for the whole dataset J=13260) as the input in urn.gp" program (Etienne 2005) under PARI computational algebra system (http://pari.math.u-bordeaux.fr/).
I implemented both 2 test and Kolmogorov- Smirnov (KS) test for comparing the discrepancy of the fitted and observed SAD patterns. The Akaike Information Criterion (AIC) method is used as well to compare the models and identify the bets model by using log-likelihoods (logL) of the fitted models as the input. The calculation of AIC formula is given byEquation
Where k is the parameter number in the fitted model.
Most of the computations (except for the NM model) were done using ad-hoc scripts under R computing environment (R Development Core Team 2013) the codes are available upon request.
SAD for microarthropod species as a whole
As showed in Figure 1 and Table I all the models could fit the whole-microarthropod SAD quite well but the difference between the expected and observed SAD still have a large significant discrepancy (indicated by 2 and K-S tests). Among the models Zipf model has the lowest AIC value indicating the most favored model for whole- microarthropod SAD pattern.
SAD for oribatids
As showed in Figure 2 and Table I all the models could fit the oribatid SAD quite well but the difference between the expected and observed SAD still have a large significant discrepancy (indicated by Chi-square and K-S tests). Among the models geometric model has the lowest AIC value indicating the most favored model for oribatid SAD pattern.
SAD for mesostigmatids
As showed in Figure 3 and Table I all the models could fit the mesostigmatid SAD quite well but the difference between the expected and observed SAD still have a large significant discrepancy (indicated by Chi-square and K-S tests). Among the models Zipf model has the lowest AIC value indicating the most favored model for mesostigmatid SAD pattern.
SAD for collembolans
As showed in Figure 4 and Table I all the models could fit the collembolan SAD quite well but the difference between the expected and observed SAD still have a large significant discrepancy (indicated by Chi-square and K-S tests). Among the models geometric model has the lowest
AIC value indicating the most favored model for collembolan SAD pattern.
Previous studies have showed the importance of spatial scales (Cheng et al. 2011) on influencing the selection of the best-fit model. At local scales typically niche-derived models were found to have highest powers for plant communities (Cheng et al. 2011; Gao et al. 2011; Yan et al. 2012). In contrast at large spatial scales Hubbell's neutral model was found to be of highest power in many cases (Cheng et al. 2011).
For microarthropod communities I would originally expect that sampling area of my study (130km 60km) is large enough for microarthropod species. As a consequence neutrality might be prevailing to influence SAD patterns. However based on the results (Table I) it is broadly supported that niche-based Zipf and geometric models are the best models over different taxonomic groups based on AIC standard. Thus my hypotheses were falsified.
One important reason for the contradictory predicted and fitted patterns above should be related to the mechanisms structuring SAD. Many studies have showed that it is not safe to draw relevant conclusions on the importance of niche and neutrality processes based on the fitting ability of exclusive models on SAD patterns. As mentioned earlier in the introduction different mechanisms could result into similar SAD patterns (Harpole and Tilman 2006; Cheng et al. 2011) and therefore the corresponding mechanisms could be not inferred.
My results are remarkably different from any previous empirical comparisons of different methods for fitting plant SADs (Du and Zhou 2008; Cheng et al. 2011; Yan et al. 2012) because I found either neutral or niche-relevant models could fit the microarthropod SADs models satisfying the requirement of statistical significance based on 2 and K-S tests. A key possible reason of the present observations might be because I fitted different
Table I.- Evaluation of different models for fitting SADs of various microarthropod taxonomic groups. Codes: Zipf-Zipf model; BSM-broken stick model; NPM-niche preemption model; GEOM-geometric model; NM-neutral model. The best AIC model for each taxonomic group is marked in boldface. Asterisk denotes significant difference between expected and observed SAD with Pless than 0.05. NA depicts non-applicable results without numeric information.
Models###Evaluation methods###Oribatids###Mesostigmatids###Collembolans###Whole community
SAD models on the raw data without any transformation. Typically log-transformation is a common practice before fitting SAD models (McGill et al. 2007). Thus the present results might be altered to some extents if log-transformation of the raw abundance data of species is utilized.
Conflict of interests
The author declares that there is no conflict of interests regarding the publication of this article.
BASTOW W. 1991. methods for fitting dominance diversity curves. J. Veg. Sci. 2: 3546.
BINNS E. 1982. Phoresy as migration-some functional aspects of phoresy in mites. Biol. Rev. 57: 571620.
CEPEDA-PIZARRO J. AND WHITFORD W. 1989. Species abundance distribution patterns of microarthropods in surface decomposing leaf-litter and mineral soil on a desert watershed. Pedobiologia (Jena). 33: 254268.
CHEN Y. 2013a. An autoregressive model for global vertebrate richness rankings: long-distance dispersers could have stronger spatial structures. Zool. Stud. 52:
CHEN Y. 2013b. Biotic element analysis of reptiles of China:
a test of vicariance model. Curr. Zool. 59: 447455.
CHEN Y. 2013c. Microarthropod diversity and distribution in Southwestern Canada. MSc thesis University of British Columbia Vancouver Canada.
CHENG J. MI X. MA K. AND ZHANG J. 2011. Responses of species-abundance distriubtion to varying sampling scales in a subtropical broad-leaved forest. Biodivers. Sci. 19: 168177.
COULSON S. HODKINSON I. AND WEBB N. 2003.
Microscale distribution patterns in high Arctic soil microarthropod communities: the influence of plant
species within the vegetation mosaic. Ecography 26:
DU X. AND ZHOU S. 2008. Testing the neutral theory of plant communities in subalpine meadow. J. Plant Ecol. (Chinese Version) 32: 347354.
ETIENNE R.S. 2005. A new sampling formula for neutral biodiversity. Ecol. Lett. 8: 253260.
FRONTIER S. 1985. Diversity and structure in aquatic ecosystems. In: Oceanography and marine biology: An annual review (ed. M. Rnes) Aberdeen University Press Aberdeen pp. 253312.
GAO L. BI R. AND YAN M. 2011. Species abundance distribution patterns of Pinus tabulaeformis forest in Huoshan Mountain of Shanxi Province China. Chinese
J. Pl. Ecol. 35: 12561270.
GOVE A. FITZPATRICK M. MAJER J. AND DUNN R.
2009. Dispersal traits linked to range size through range location not dispersal ability in Western Australian angiosperms. Glob. Ecol. Biogeogr. 18: 595606.
HARPOLE W. AND TILMAN D. 2006. Non-neutral patterns of species abundance in grassland communities. Ecol. Lett. 9: 1523.
HOYLE M. AND HARBORNE A. 2005. Mixed effects of habitat fragmentation on species richness and community structure in a microarthropod microecosystem. Ecol. Ent. 30: 684691.
HUBBELL S. 2006. Neutral theory and the evolution of ecological equivalence. Ecology 87: 13871398.
HUBBELL S.P. 2001. The unified neutral theory of biodiversity and biogeography (MPB-32) (Monographs in population biology) Princeton University Press Princeton USA.
JABOT F. ETIENNE R. AND CHAVE J. 2008. Reconciling neutral community models and environmental filtering: theory and an empirical test. Oikos 117: 13081320.
LINDO Z. AND WINCHESTER N.N. 2007. Resident corticolous oribatid mites (Acari: Oribatida): Decay in community similarity with vertical distance from the ground. Ecoscience 14: 223229.
LINDO Z. AND WINCHESTER N.N. 2009. Spatial and environmental factors contributing to patterns in arboreal and terrestrial oribatid mite diversity across spatial scales. Oecologia 160: 817825.
MACARTHUR R. 1957. On the relative abundance of bird species. Proc. natl. Acad. Sci. 43: 283295.
MCGILL B. ETIENNE R. GRAY J. ALONSO D.
ANDERSON M. BENECHA H. DORNELAS M. ENQUIST B. GREEN J. HE H. HURLBERT A.H. MAGURRAN A.E. MARQUET P.A. MAURER B.A. OSTLING A. SOYKAN C.U. UGLAND K.I. WHITE E.P. 2007. Species abundance distributions: moving beyond singla prediciton theories to integration
within an ecological framework. Ecol. Lett. 10: 995
MORRONE J.J. AND CRISCI J.V. 1995. Historical Biogeography: Introduction to Methods. Annu. Rev. Ecol. Syst. 26: 373401.
MOTOMURA I. 1932. On the statistical treatment of communities. Zool. Mag. 44: 379383.
MUNOZ F. COUTERON P. AND HUBBELL S. 2012.
Comment on Global correlations int ropical tree species richness and abundance reject neutrality". Science 336: 1639.
NEE S. AND STONE G. 2003. The end of the beginning for neutral theory. Trends Ecol. Evol. 18: 433434.
PERDOMO G. SUNNUCKS P. AND THOMPSON R. 2012.
The role of temperature and dispersal in moss- microarthropod community assembly after a catastrophic event. Proc. R. Soc. B 367: 30423049.
POTTHOFF M. JOHST K. GUTT J. AND WISSEL C. 2006.
Clumped dispersal and species coexistence. Ecol. Modell. 198: 247254.
R DEVELOPMENT CORE TEAM 2013. R: A language and environment for statistical computing Vienna Austria. ISBN 3-900051-07-0 URL http://www.R-project.org.
RICKLEFS R. AND RENNER S. 2012. Global correlations in tropical tree species richness and abundance reject neutrality. Science 335: 464467.
ROSINDELL J. HUBBELL S. AND ETIENNE R. 2011. The unified neutral theory of biodiversity and biogeography at age 10. Trends Ecol. Evol. 26: 340348.
ROSINDELL J. HUBBELL S.P. HE F. HARMON L.J.
AND ETIENNE R.S. 2012. The case for ecological neutral theory. Trends Ecol. Evol. 27: 203208.
SZYMKOWIAK P. GORSKI G. AND BAJERLEIN D. 2007.
Passive dispersal in arachnids. Biol. Lett. 44: 75101.
VOLKOV I. BANAVAR J. HUBBELL S. AND MARITAN A. 2003. Neutral theory and relative species abundance in ecology. Nature 424: 10351037.
VOLKOV I. BANAVAR J. HUBBELL S. AND MARITAN A. 2007. Patterns of relative species abundance in rainforests and coral reefs. Nature 450: 4549.
WALKER S. AND CYR H. 2007. Testing the standard neutral model of biodiversity in lake communities. Oikos 116:
YAN Y. ZHANG C. AND ZHAO X. 2012. Species- abundance distribution patterns at different succesional stages of conifer and broad-leaved mixed forest communities in Changbai Mountains China. Chinese J. Pl. Ecol. 36: 923934.
|Printer friendly Cite/link Email Feedback|
|Publication:||Pakistan Journal of Zoology|
|Date:||Aug 31, 2014|
|Previous Article:||Dietary Distinctive Features of Tawny Owl Strix aluco (Linn 1758) and Barn Owl Tyto alba (Scopoli 1759) in Gardens of Algerian Sahel El Harrach...|
|Next Article:||A Check-List of Tintinnids (Protozoa: Ciliophora) in the Coastal Zone of Turkey.|