High niche partitioning promotes highly specialized, modular and non‐nested florivore–plant networks across spatial scales and reveals drivers of specialization

Abstract

Understanding patterns of ecological specialization, the processes underlying niche partitioning and how they translate into the structure of interaction networks is a persistent challenge in ecology. Advances on this regard are limited by the prevalent focus on single spatial scales, lack of tests of the mechanisms underlying specialization, and scarce investigation of some types of interactions. Here we investigated the patterns of interaction between plants and florivores (flower‐breeding drosophilids, FBD) at species‐ and community‐level and at local and regional scales, and tested the relative importance of multiple potential drivers of frequencies of interactions in a local network. First, based on a year‐round collection of 42 322 flowers belonging to 82 plant species, we investigated species specialization and network structure and tested whether frequencies of interactions were related to plant–consumer temporal overlap and resource availability. Second, we built a regional florivore–plant meta‐network for the Neotropical region and tested its structure for nestedness, modularity and overall specialization. Our findings revealed that although FBD species span a broad range of degrees of specialization, most species were highly specialized. At both local and regional scales, network structure was highly modular and non‐nested, presenting high complementary specialization. Moreover, phenological overlap between FBD and their hosts was the most influential driver of frequency of interactions, in comparison to abundance and traits. By describing the structure of FBD–plant networks, our results illustrate how a highly diverse and specialized system retains interaction patterns found in other types of interaction networks which are often driven by different processes. Furthermore, despite undersampling of interactions in the meta‐network caused by the lack of studies on this system and the high diversity of Neotropical FBD's, the high modularity and consistency of some clade–clade modules across spatial scales suggests the importance of evolutionary history and physiological constraints in shaping interactions in this system.

Morphological matching and phenological overlap promote niche partitioning and shape a mutualistic plant–hawkmoth network

“…A modular network structure often emerges as a consequence of low niche breadth in the community (Maruyama et al ., 2014; Ponisio et al ., 2017), which generally leads to high niche partitioning (Ponisio et al ., 2019; Cordeiro et al ., 2020) as detected here. In plant–hawkmoth systems, morphology importantly mediates specialisation and influences interaction frequencies (Sazatornil et al ., 2016; Johnson et al ., 2017).…”

“…Species interact with one another in communities, forming complex interaction networks (Bascompte & Jordano, 2006; Tylianakis et al ., 2010). The structure emerging in such networks is expected to reflect patterns of specialisation and niche partitioning within the community (Ponisio et al ., 2019; Cordeiro et al ., 2020; Phillips et al ., 2020), which are ultimately defined by multiple non‐exclusive processes (for a review, see Vázquez et al ., 2009a; Vizentin‐Bugoni et al ., 2018). However, how species' ecological characteristics (e.g.…”

“…traits, spatiotemporal distribution) associate with their specialisation and result in community‐level patterns of niche partitioning and the emergence of network structures remains debated (e.g. Ponisio et al ., 2019; Cordeiro et al ., 2020) and little investigated for some systems.…”

“…A nested network occurs when generalist species interact with other generalists and also with specialists, while specialists interact with subsets of the partner of the more generalist species (Krishna et al ., 2008; Almeida‐Neto & Ulrich, 2010). Such a structure is associated with low to moderate levels of resource partitioning (Ponisio et al ., 2019; Cordeiro et al ., 2020). On the other hand, modularity emerges when subsets of species interact more among themselves than with other species in the network, forming modules of interactions (Olesen et al ., 2007; Dormann & Strauss, 2014; Maruyama et al ., 2014).…”

“…On the other hand, modularity emerges when subsets of species interact more among themselves than with other species in the network, forming modules of interactions (Olesen et al ., 2007; Dormann & Strauss, 2014; Maruyama et al ., 2014). Thus, modularity may indicate higher levels of resource partitioning (Ponisio et al ., 2019; Cordeiro et al ., 2020; Phillips et al ., 2020).…”