Background: Understanding the patterns of species richness across elevational gradients is a key concept for contemporary research in ecology and evolution, and critical to understanding large-scale trends in biodiversity, global change and conservation. However, patterns of elevational species richness between taxonomic groups, regions and latitudes are inconsistent, so that various, sometimes conflicting hypotheses exist. Several scholars have pointed out that research on elevational distribution patterns is often biased by the sampling design employed. To test this hypothesis, we analyzed species richness of Nematode-Trapping Fungi (NTF) across an elevation gradient at two mountainous sites in western Yunnan Province, P.R. China. We tested for potential differences in the results when using different sampling designs.Results: A total of 3 genera, 17 species, 222 strains of NTF were isolated and identified from Gaoligongshan and Cangshan. Species accumulation curves for both sites and sampling modes had acceptable leveling, demonstrating sufficient sampling effort. At Gaoligongshan, the elevation distribution patterns of NTF were different under two sampling patterns. When reducing the analyzed altitude range in Gaoligongshan, the elevation distribution pattern of the NTF changed. A similar elevation distribution pattern was observed in Cangshan when testing the same altitude range. In general, when treating the same dataset using different sampling designs, the resulting distribution patterns of species richness and occurrence frequencies were clearly different. Moreover, after removal of the samples located within lower-altitude zones affected by anthropogenic interferences, the distribution pattern of NTF in the two sites tended to become uniform.Conclusion: The sampling design, and in particular the elevation interval between plots, has a significant effect on the assessment of species distribution in mountainous regions. Other factors such as human activities and the multi-dimensionality of biodiversity also contribute to result biases. It is recommended that sampling design is given careful consideration in future studies on the elevational gradients of species richness, using stratified approaches according to the most relevant factors.

Studies of biogeographic patterns of fungi have long been behind those of plants and animals. The presence of worldwide species, the lack of systematic sampling design and adequate sampling effort, and the lack of research units are responsible for this status. This study investigates the biogeographical patterns of Arthrobotrys oligospora, the most widespread globally distributed nematode-trapping fungi (NTF), by stratified collecting and analyzing 2,250 samples from 228 sites in Yunnan Province, China. The A. oligospora was isolated, and 149 strains were subjected to ITS, TUB, TEF and RPB2 gene sequencing and multi-gene association phylogeographic analysis. The results show that at population level A. oligospora is randomly distributed throughout Yunnan Province and has no biogeographical distribution pattern. At the genetic level, the phylogenetic tree of A. oligospora diverges into five major evolutionary clades, with a low degree of gene flow between the five clades. However, the correlation between the phylogenetic diversity of A. oligospora and geographical factors was low. There was no clear pattern in the phylogenetic clades distribution of A. oligospora either without dividing the study unit or when the grid was used as the study unit. When watersheds were used as the study unit, 67.4%, 63.3%, 65.9%, 83.3%, and 66.7% of clade 1–5 strains were distributed in the Jinsha river, Red river, Peal river, Lancang river, and Nujiang-Irawaddy river watersheds, respectively. The clades distribution of A. oligospora was highly consistent with the watersheds distribution. Training predictions of the clades distributions using randomly generated polygons were also less accurate than watersheds. These results suggest that watersheds are key to discovering the biogeographic distribution patterns of A. oligospora. The A. oligospora populations are blocked by mountains in the watershed, and gene flow barriers have occurred, which may have resulted in the formation of multiple cryptic species. Watersheds are also ideal for understanding such speciation processes, explaining factors affecting biodiversity distribution and coupling studies of plant and animal and microbial diversity.

The mechanisms that underpin the species–area relationship (SAR) are crucial for both the development of biogeographic theory and the application of biodiversity conservation. Since its origin, the resource hypothesis, which proposes that rich resources in vast ecosystems will lower extinction rates and shape the SAR, has not been tested. The impossibility to quantify resources and extinction rates using plants and animals as research subjects, as well as the inability to rule out the influences of the area per se, habitat diversity, dispersal, and the historical background of biodiversity, make testing this hypothesis problematic. To address these challenges and test this hypothesis, two sets of microbial microcosm experimental systems with positive and negative correlated resources and volumes were created in this work. The results of 157 high-throughput sequencing monitoring sessions at 11 time points over 30 consecutive days showed that neither of the experimental groups with positive or negative correlations between total resources and microcosm volume had a significant SAR, and there were no negative correlations between extinction rates and resources. Therefore, in our microcosmic system, resources do not influence extinction rates or shape the SAR. Dispersal should be the principal mode of action if the resource theory is correct.

Dispersal is one of the key processes determining biodiversity. The passive sampling hypothesis, which emphasizes dispersal processes, suggests that larger habitats receive more species from the species pool as the main mechanism leading to more species in larger habitats than in smaller habitats (i.e., species-area relationships). However, the specific mechanisms by which dispersion shapes biodiversity still need to be discovered due to the difficulties of quantifying dispersal and the influence of multiple factors. Solving the above problem with a designed experiment is necessary to test the passive sampling hypothesis. This study designed a passive sampling experiment using sterile filter paper to quantify the microbial diffusion process, excluding the effects of pure sampling effects, habitat heterogeneity, and extinction processes. The results of high-throughput sequencing showed that a larger filter paper could receive more colonists, and the passive sampling hypothesis of SAR was confirmed. Dispersal shaped SAR by increasing species richness, especially rare species, and increasing the species replacement rate between habitats. These two processes are the mechanisms by which dispersal shapes biodiversity patterns. Compared with the results of this study, the commonly used mathematical model of passive sampling was able to predict the richness of non-rare species accurately but underestimated the richness of rare species. Underestimating rare species by mathematical models of passive sampling is more severe in small habitats. These findings provide new insights into the study of dispersal processes and the mechanism of species-area relationships.