Ecological and evolutionary processes involved in shaping microbial habitat generalists and specialists in urban park ecosystems

ABSTRACT Microbiomes are integral to ecological health and human well-being; however, their ecological and evolutionary drivers have not been systematically investigated, especially in urban park ecosystems. As microbes have different levels of tolerance to environmental changes and habitat preferences, they can be categorized into habitat generalists and specialists. Here, we explored the ecological and evolutionary characteristics of both prokaryotic and microeukaryotic habitat generalists and specialists from six urban parks across five habitat types, including moss, soil, tree hole, water, and sediment. Our results revealed that different ecological and evolutionary processes maintained and regulated microbial diversity in urban park ecosystems. Under ecological perspective, community assembly of microbial communities was mainly driven by stochastic processes; however, deterministic processes were higher for habitat specialists than generalists. Microbial interactions were highly dynamic among habitats, and habitat specialists played key roles as module hubs in intradomain networks. In aquatic interdomain networks, microeukaryotic habitat specialists and prokaryotic habitat specialists played crucial roles as module hubs and connectors, respectively. Furthermore, analyzing evolutionary characteristics, our results revealed that habitat specialists had a much higher diversification potential than generalists, while generalists showed shorter phylogenetic branch lengths as well as larger genomes than specialists. This study broadens our understanding of the ecological and evolutionary features of microbial habitat generalists and specialists in urban park ecosystems across multi-habitat. IMPORTANCE Urban parks, as an important urban greenspace, play essential roles in ecosystem services and are important hotspots for microbes. Microbial diversity is driven by different ecological and evolutionary processes, while little is currently known about the distinct roles of ecological and evolutionary features in shaping microbial diversity in urban park ecosystems. We explored the ecological and evolutionary characteristics of prokaryotic and microeukaryotic habitat generalists and specialists in urban park ecosystems based on a representative set of different habitats. We found that different ecological and evolutionary drivers jointly maintained and regulated microbial diversity in urban park microbiomes through analyzing the community assembly process, ecological roles in hierarchical interaction, and species diversification potential. These findings significantly advance our understanding regarding the mechanisms governing microbial diversity in urban park ecosystems.


Figure S5
Hierarchical cluster analysis of 90 samples based on the "ward.D2" method from five habitats (i.e., moss, sediment, soil, tree hole, and water), showing the habitat effect on the microbial community.Prokaryotes and microeukaryotes were obtained based on the high-throughput sequencing of 16S rRNA gene V3-V4 region and 18S rRNA gene V4 region, respectively.

Figure S1
Figure S1 Sampling locations in urban parks in Xiamen city.(A) Locations of sampling sites in Xiamen City.(B) The coordinates of six parks.(C) Pictures of five sampling habitats.

Figure S3
Figure S3Relative abundance (%) of identified strict generalists and specialists at phylum level of (A) prokaryotes and (B) microeukaryotes, respectively.

Figure S4
Figure S4 Taxonomic information of habitat generalists and specialists at the phylum level of (A) prokaryotes and (B) microeukaryotes, respectively.The 12 most abundant phyla are presented."Others" represents the sum of all other phyla.

Figure S6
Figure S6 Results of non-metric multidimensional scaling (NMDS) analysis based on Bray-Curtis dissimilarity of prokaryotic habitat (A) generalists and (B) specialists, and microeukaryotic habitat (C) generalists and (D) specialists in urban parks.***P < 0.001.

Figure S9
Figure S9 Relationships between weighted UniFrac dissimilarity and spatial distance of prokaryotic habitat (A) generalists and (B) specialists, as well as microeukaryotic habitat (C) generalists and (D) specialists in urban parks.A linear regression was fitted between community dissimilarity and spatial distance of pairwise samples.n is the number of sample pairs.r value is the Pearson's correlation coefficient, and star represents significance.*P < 0.05, **P < 0.01, ***P < 0.001.

Fig
Fig S10 Percentages of habitat generalists and specialists in the microbial network.(A) The percentage of habitat specialists or generalists in the intradomain network from total specialists or generalists.(B) The percentages of identified keystone species in the intradomain network from total specialists or generalists.(C) The percentage of specialists or generalists in the interdomain network from total specialists or generalists.(D) The percentage of keystone species in the interdomain network from total specialists or generalists.

Figure S11
Figure S11 Vulnerability of networks measured from the contribution of the node in network to global efficiency.The significance test was performed by the Wilcoxon test.*P < 0.05.

Figure S12
Figure S12The phylogenetic branch length of generalists and specialists.The significance test was performed by the Wilcoxon test.***P < 0.001.

Table S1
Prokaryotic zOTUs table with abundances and classifications, three niche breadth values (i.e., Levins' niche breadth, Shannon diversity, and occurrence), and assignments of generalists and specialists.See supplementary Excel file for details.

Table S2
Microeukaryotic zOTUs table with abundances and classifications, three niche breadth values (i.e., Levins' niche breadth, Shannon diversity, and occurrence), and assignments of generalists and specialists.See supplementary Excel file for details.

Table S3
Good's coverage of prokaryotic and microeukaryotic communities.Values are mean ± standard deviation.

Table S5
Network properties of prokaryotic and microeukaryotic intradomain networks.
a Mean ± standard deviation (SD) based on 100 randomized networks.b-dSignificant difference (P < 0.05) in average clustering coefficient, average path length, and modularity between empirical and randomized networks based on Student's t-test, respectively.avgCC, Average clustering coefficient; GD, Average path distance; M, Modularity.

Table S6
Network properties of prokaryotic-microeukaryotic interdomain networks.
a Mean ± standard deviation (SD) based on 100 randomized networks.b-eSignificant difference (P < 0.05) in modularity, niche overlap, robustness, and functional complementarity between empirical and randomized networks based on Student's t-test, respectively.FC, Functional complementarity; M, Modularity.

Table S7
Identified keystone species in interdomain networks and their taxonomic annotations.See supplementary Excel file for details.Table S8Prokaryotic zOTUs numbers, matched zOTUs numbers and percentages with the genome database at different levels.