Introduction
One of the most enduring challenges for community ecology is how species diversity maintains in natural communities (Sutherland et al. 2013). Many mechanisms have been proposed to explain species coexistence and community assembly, such as negative density dependence (NDD), environmental filtering, competitive exclusion, and facilitation (Chesson & Peter 2000, Webb et al. 2006, Adler et al. 2007, HilleRisLambers et al. 2012, Fichtner et al. 2017, Adler et al. 2018).
Among these various species coexistence mechanisms, NDD is a widely accepted explanation for the maintenance of forest biodiversity where seedlings suffer negative effects from host specialized natural enemies (pathogen and pest) accumulated by local conspecific or phylogenetically related neighbors (Janzen 1970, Connell 1971, Webb et al. 2006, Gripenberg et al. 2014, Zambrano et al. 2017, Forrister et al. 2019, Song et al. 2021). Multiple lineages of natural enemies, such as fungal, bacterial and viral pathogens together with herbivorous mammals and insects, may contribute to NDD (Hazelwood et al. 2021, Song & Corlett 2021). The contributions of fungal pathogens and herbivores to NDD have been documented in many studies, while the contributions of bacterial pathogens and plant virus were seldom evaluated (Bagchi et al. 2014, Gripenberg et al. 2014, Hazelwood et al. 2021, Song & Corlett 2021). Moreover, most previous researches focused mainly on single type of natural enemy, and studies simultaneously demonstrating the contribution of multiple natural enemies are still lacking (Hazelwood et al. 2021).
A recently developed community functional genomics strategy may help us to take an insight into the intrinsic mechanism of NDD from the plant side and consider multiple natural enemies simultaneously. Plants can develop multiple molecular responses to microbial pathogens and herbivores (Wiesner-Hanks & Nelson 2016, Aljbory & Chen 2018, Ramirez-Prado et al. 2018). The genomic variations involved in these defense responses of plant species may be related to plant-pathogen or plant-pest co-evolutionary history, reflect susceptibility of plants to diseases and pests, and thereby determine growth and survival of individual plant species as well as the whole plant community (Żmieńko et al. 2014, Dolatabadian et al. 2017). The community functional genomics strategy seeks to quantify the similarity of non-model species via transcriptomic data using Gene Ontology (GO) annotation of functional genes (Swenson 2012, Swenson et al. 2017). This strategy can provide information of hundreds of defense related genes and also has the potential to provide insights into how different lineages of natural enemies affect plant community structure separately and simultaneously.
Copy number variations (CNVs) and phylogenetic similarity of homologous genes are two principal methods in community functional genomics to describe the gene expressional similarity of plant species in response to biotic and abiotic environments (Han et al. 2017, Zambrano et al. 2017). CNV is defined as copy numbers of homologous genes or size of gene family (Liu 2014). CNV has been discovered for decades (Bridges 1936, Ohno 1970), which is resulting from gene duplication, gain or loss, and recently recognized as a common type of genomic variations (Żmieńko et al. 2014, Dolatabadian et al. 2017). Gene duplication may induce functional changes of genes such as subfunctionalization, pseudogenization, and neofunctionalization under rapid evolutionary processes, or keep their identical function by purifying selection or concerted evolution (Zhang 2003, Magadum et al 2013). Moreover, the functional conservation of genes with high sequence similarity, can be even across distinct phylogenetic lineages (Kater et al 2006, Liu et al 2014). For example, the expression of Arabidopsis ortholog AtWRKY40 in barley also compromised basal resistance to powdery mildew, indicating functional conservation of related proteins across dicots and monocots (Liu et al 2014). The function-conserved gene copies can be easily detected based on their high levels of sequence similarity (Magadum et al 2013), and therefore CNVs based on sequence similarity may act as a useful tool in researches on molecular functional diversity of species and natural communities. CNVs seem likely to be involved in many biological processes and can be associated with some important phenotypic variations (Manolio et al. 2009, Bai et al. 2016). Based on a functional genetics research on rice, Bai et al. (2016) suggested that CNVs might be involved in plant resistance to insects. CNVs may also differ among species and has been implicated in ecological adaptation. Besides model species, we can also identify CNVs from non-model species and it had been used in several evolutionary and ecological studies on microbes (Schirrmeisteret al. 2012, Roller et al. 2016, Nunan et al. 2020), butterflies (Seixas et al. 2021), polar bears (Rinker et al. 2019), and plant species (Zhang et al. 2019, Han et al. 2017). For example, Han et al. (2017) used CNVs of light-related GOs to test how differences of tree species in light utilization strategies affect seedling regeneration in a subtropical forest. The CNVs extracted from transcriptomic data, defined as functional CNVs, could be useful in functional ecological researches because functional CNVs could reflect not only the evolutionary history of species but also the ecological responses to biotic and abiotic environments.
There are also debates on how CNVs are involved in defense responses and plant performances (Katju & Bergthorsson 2013, Żmieńko et al. 2014, Dolatabadian et al. 2017). It is expected that high copy number of defense genes is advantageous for plants due to the enhanced resistance against natural enemies, while there might be a trade-off between the expression of defense genes and plant performance and the high cost in transcription and translation will in turn limit the copy number of resistance genes (Lin et al. 2013). Moreover, the copy number of defense genes might also be the co-evolutionary consequence with exposure chance to natural enemies, which may reflect pathogen profiles or herbivore range of host plants (Zhai et al. 2011).
In this study, we extracted CNVs of homologous genes annotated as defense response to four lineages of natural enemies (fungus, bacterium, insect and virus). We used partial linear regression analysis to reveal the effect of gene copy number on seedling survival at species- and community- level. And we also used generalized linear mixed-effects models (GLMM) to assess how the neighborhood dissimilarities in functional CNVs impact seedling survival. For these aims, we hypothesized that the high copy number of defense response to natural enemies is associated with high seedling survival (Hypothesis I), because increased gene copy number may enhance biotic resistance. As the functional CNVs of defense responses to natural enemies may be strongly linked to their pathogen profiles or herbivore range, we also hypothesized that seedlings surround by adult and seedling neighbors with high dissimilarity in defense-related CNVs should share fewer natural enemies and therefore have a high survival possibility (Hypothesis II).