Emergent Rarity Properties in Carabid Communities From Chinese Steppes With Different Climatic Conditions

Tsafack, Noelline and Borges, Paulo A. V. and Xie, Yingzhong and Wang, Xinpu and Fattorini, Simone (2021) Emergent Rarity Properties in Carabid Communities From Chinese Steppes With Different Climatic Conditions. Frontiers in Ecology and Evolution, 9. ISSN 2296-701X

[thumbnail of pubmed-zip/versions/1/package-entries/fevo-09-603436/fevo-09-603436.pdf] Text
pubmed-zip/versions/1/package-entries/fevo-09-603436/fevo-09-603436.pdf - Published Version

Download (676kB)

Abstract

Species abundance distributions (SADs) are increasingly used to investigate how species community structure changes in response to environmental variations. SAD models depict the relative abundance of species recorded in a community and express fundamental aspects of the community structure, namely patterns of commonness and rarity. However, the influence of differences in environmental conditions on SAD characteristics is still poorly understood. In this study we used SAD models of carabid beetles (Coleoptera: Carabidae) in three grassland ecosystems (desert, typical, and meadow steppes) in China. These ecosystems are characterized by different aridity conditions, thus offering an opportunity to investigate how SADs are influenced by differences in environmental conditions (mainly aridity and vegetation cover, and hence productivity). We used various SAD models, including the meta-community zero sum multinomial (mZSM), the lognormal (PLN) and Fisher’s logseries (LS), and uni- and multimodal gambin models. Analyses were done at the level of steppe type (coarse scale) and for different sectors within the same steppe (fine scale). We found that the mZSM model provided, in general, the best fit at both analysis scales. Model parameters were influenced by the scale of analysis. Moreover, the LS was the best fit in desert steppe SAD. If abundances are rarefied to the smallest sample, results are similar to those without rarefaction, but differences in models estimates become more evident. Gambin unimodal provided the best fit with the lowest α-value observed in desert steppe and higher values in typical and meadow steppes, with results which were strongly affected by the scale of analysis and the use of rarefaction. Our results indicate that all investigated communities are adequately modeled by two similar distributions, the mZSM and the LS, at both scales of analyses. This indicates (1) that all communities are characterized by a relatively small number of species, most of which are rare, and (2) that the meta-communities at the large scale maintain the basic SAD shape of the local communities. The gambin multimodal models produced exaggerated α-values, which indicates that they overfit simple communities. Overall, Fisher’s α, mZSM θ, and gambin α-values were substantially lower in the desert steppe and higher in the typical and meadow steppes, which implies a decreasing influence of environmental harshness (aridity) from the desert steppe to the typical and meadow steppes.

Item Type: Article
Subjects: Science Repository > Multidisciplinary
Depositing User: Managing Editor
Date Deposited: 12 Jul 2023 03:34
Last Modified: 07 Oct 2023 09:20
URI: http://research.manuscritpub.com/id/eprint/2620

Actions (login required)

View Item
View Item