Experimental Verification Of Limiting Similarity Hypothesis And Habitat Filtering Hypothesis

The limiting similarity hypothesis and the habitat filtering hypothesis are always considered as major of processes of community assembly in community ecology. Specifically, the limiting similarity hypothesis states that strength of interspecific competition highly depends on ecological similarity among species, and there is some maximum level of similarity between species for avoiding too much intensive competition that can drive competitive exclusion. The habitat filtering hypothesis emphasizes the role of habitat conditions in selecting and persisting of local species from the regional species pool. A fully understanding of the two hypotheses in natural community can be very helpful in dealing with the practical issues of biodiversity loss, protection of endangered or rare species, natural community recovery and invasive species control. With the flourishing development of phylogenetic community ecology and functional trait ecology in recent decade, the hypotheses have become two of the main focus of contemporary community ecology.A key assumption of the limiting similarity hypothesis is that the intensity of interspecific competition will be positively correlated with their ecological similarity, as well as negatively correlated with their phylogenetic relatedness. A growing body of studies, however, suggest that this competition-relatedness relationship might be weak, or invalid and even opposite in natural communities. Some recent studies also pointed out that the effect of habitat filtering on community structures might be not as strong as widely acknowledged before. The two predictions of the habitat filtering hypothesis, similarity of species composition between two communities will increase and decrease under distinct and similar habitat respectively, could not be observed under the effect of other potentially important ecological processes. As results, inferences of ecological processes based on these two predictions become inappropriate and inaccurate.In light of the importance of the two hypotheses in community ecology and the current gaps in this field discussed above, this study aimed to examine the following three unsolved scientific questions about the hypotheses by a wood seedling experiment and an annual weeds experiment.(1) Do the interspecific competitive intensity and interspecific phylogenetic distance have significant negative correction?(2) Do the different levels of habitat filtering significantly increase the species composition difference among communities?(3) Does the same levels of habitat significantly decrease the species composition difference among communities?To this end, the experiment of woody seedlings was setup in Ningbo, Zhejiang Province, near the Tiantong National field observation station. Six canopy species, Cyclobalanopsis glauca, Castanopsis carlesii, Cinnamomum camphora, Castanopsis sclerophylla, Liquidambar formosana, Lithocarpus glaber, were selected to test the 1st question. Seedlings of these species were monoculture and bioculture in the field, base diameter, height and dry biomass of each seedling were collected to calculate their pairwise competition intensity. Phylogenetic distances among the species were estimated from APG III to approximate their ecological similarity. The second weed experiment was carried out near the Heishiding natural protection area, Zhaoqing, Guangdong Province. The experiment designed four different light shading treatments on the local natural herb communities, and monitored their dynamic of inter-community similarity to test the 2nd and 3rdquestions. Our main results are below.1、In the pairwise competition experiment of six canopy species, the competition intensity of all experimental species, except the Cinnamomum camphora and Liquidambar formosana with highest competitive ability, increases with their phylogenetic distances. This result is contrary to the key assumption of the limiting similarity hypothesis, but consistent with new species coexistence theory proposed by Mayfield. Further analysis revealed that this result could be mainly attributed to species competitive ability differences linked to difference of seedling height.2、In the herbaceous experiment, the inter-community difference of species composition significantly increased after the different light-shading treatments. Meanwhile, the bigger the light intensity difference among communities is, the larger the species composition difference among community in the final of experiment periods. Therefore, our results confirmed one of the habitat filtering hypothesis that the difference types of habitat filtering can increase the inter-community difference of the species composition.3、Unexpectedly, the inter-community difference of species composition in the same light-shading treatments also increased significantly at the end of experiment. This result is not consistent with the conventional expectation of the habitat filtering hypothesis in the same environment. Our results also hint that the difference of initial species composition among communities may be an important factor leading to the increased inter-community differences of species composition even under the same shading treatment.In summary, results of only one of the three questions mentioned before are confirmed by the limiting similarity hypothesis and the habitat filtering hypothesis; other are not, and even opposite to the expectations of the two hypothesis. These outcomes, on one hand, support the possibility that the assembly of natural community is far more complicated than the picture described by the limiting similarity hypothesis and the habitat filtering hypothesis, and there are other factors, such as interspecific competitive ability difference and initial species composition, that cannot be ignored in our simplified framework, and are essential for fully understanding various community patterns. On the other hand, these results highlight the importance of field experiments in rigorously testing ecological theories and hypotheses. This type of experiment can control and even isolate effects of other non-interesting factors, and simulate environment of the experiment as similar as the natural community. Despite the unavoidable physical and fanatical limitations of the experiment, our conclusion may only valid for limited communities in certain life stages, the inconsistence between our results and the predictions of the two hypotheses is worth enough for us to reconsider the application scope and importance of the limiting similarity hypothesis and the habitat filtering hypothesis.