Past and potential future effects of habitat fragmentation on structure and stability of plant–pollinator and host–parasitoid networks

Abstract

Habitat fragmentation is a primary threat to biodiversity, but how it affects the structure and stability of ecological networks is poorly understood. Here, we studied plant-pollinator and host-parasitoid networks on 32 calcareous grassland fragments covering a size gradient of several orders of magnitude and with amounts of additional habitat availability in the surrounding landscape that varied independent of fragment size. We find that additive and interactive effects of habitat fragmentation at local (fragment size) and landscape scales (1,750 m radius) directly shape species communities by altering the number of interacting species and, indirectly, their body size composition. These, in turn, affect plant-pollinator, but not host-parasitoid, network structure: the nestedness and modularity of plant-pollinator networks increase with pollinator body size. Moreover, pollinator richness increases modularity. In contrast, the modularity of host-parasitoid networks decreases with host richness, whereas neither parasitoid richness nor body size affects network structure. Simulating species coextinctions also reveals that the structure-stability relationship depends on species' sensitivity to coextinctions and their capacity for adaptive partner switches, which differ between mutualistic and antagonistic interaction partners. While plant-pollinator communities may cope with future habitat fragmentation by responding to species loss with opportunistic partner switches, past effects of fragmentation on the current structure of host-parasitoid networks may strongly affect their robustness to coextinctions under future habitat fragmentation.

Agricultural intensification drives changes in hybrid network robustness by modifying network structure

“…Few studies have empirically examined the response of network structure and robustness to agricultural intensification, a predominant environmental disturbance of the Anthropocene (Tylianakis et al ; Grass et al ), and fewer still have attempted to characterise this relationship specifically for networks that include both mutualistic and antagonistic interactions of animals and plants (García‐Callejas et al ). We present evidence for three main conclusions: First, anthropogenic disturbance, in the form of agricultural intensification, drove changes in hybrid network structure and network robustness.…”

“…Because most network metrics are dependent on network size and shape (Olesen et al ; Dormann et al ), in order to directly compare hybrid network structure along the agricultural intensification gradient, we standardised all network metrics relative to a null expectation by calculating z ‐scores (Tylianakis & Morris ; Grass et al ) using the equation z = ( – )/() where and are, respectively, the mean and standard deviation of the null distribution for metric . One thousand null networks were created for each bipartite sub‐network separately, then joined together before calculating the null network metrics.…”

“…The co‐extinction threshold attempts to capture the reality of functional extinctions within natural communities (Säterberg et al ). In addition, in real‐world extinction scenarios, species likely re‐allocate some of their interactions to compensate for the loss of an interaction partner (Kaiser‐Bunbury et al ; Ramos‐Jiliberto et al ; Schleuning et al ; Tylianakis & Morris ; Grass et al ; Simmons et al ). Therefore, at each step in the extinction scenario we incorporated the capacity for species to re‐allocate 50% of their total lost interactions to their remaining interaction partners within the network (constrained rewiring, as in Schleuning et al ), which granted greater flexibility to generalists than specialists.…”

“…Robustness is directly linked to network structure (Dunne et al ; Colwell et al ; Evans et al ) and in the midst of the current global extinction crisis, identifying community structures that are more vulnerable to species extinction could help guide conservation efforts to prevent or mitigate the impacts of such species losses (Tylianakis et al ; Kaiser‐Bunbury & Blüthgen ; Harvey et al ). Furthermore, anthropogenic disturbances are rarely isolated (Vinebrooke et al ) – for instance, habitat fragmentation is often followed by increases in hunting and extraction (Peres ; Kosydar et al ) – and changing network structure via one disturbance can alter its robustness to subsequent disturbances (Grass et al ).…”

Mechanisms structuring host–parasitoid networks in a global warming context: a review

“…Species' responses to global warming, such as shifts in phenology, physiological changes and range shifts, have been widely studied (Walther et al, 2002). However, responses of individual species to global warming can have cascading effects via their interactions with other species (Valiente-Banuet et al, 2015), and influence the structure of ecological networks linking the wider community (Grass et al, 2018). Despite this, most previous studies on the effect of global warming on terrestrial ecosystems have focused on individual species in isolation, or at most have considered pairwise interactions between species (Tylianakis et al, 2008;Walther, 2010), overlooking important effects on ecological network structure (Tylianakis et al, 2007) and ecosystem functioning (Goudard & Loreau, 2008;Miele et al, 2018).…”