Modularity, pollination systems, and interaction turnover in plant-pollinator networks across space.

Mutualistic interaction networks have been shown to be structurally conserved over space and time while pairwise interactions show high variability. In such networks, modularity is the division of species into compartments, or modules, where species within modules share more interactions with each other than they do with species from other modules. Such a modular structure is common in mutualistic networks and several evolutionary and ecological mechanisms have been proposed as underlying drivers. One prominent explanation is the existence of pollination syndromes where flowers tend to attract certain pollinators as determined by a set of traits. We investigate the modularity of seven community level plant-pollinator networks sampled in rupestrian grasslands, or campos rupestres, in SE Brazil. Defining pollination systems as corresponding groups of flower syndromes and pollinator functional groups, we test the two hypotheses that (1) interacting species from the same pollination system are more often assigned to the same module than interacting species from different pollination systems and; that (2) interactions between species from the same pollination system are more consistent across space than interactions between species from different pollination systems. Specifically we ask (1) whether networks are consistently modular across space; (2) whether interactions among species of the same pollination system occur more often inside modules, compared to interactions among species of different pollination systems, and finally; (3) whether the spatial variation in interaction identity, i.e., spatial interaction rewiring, is affected by trait complementarity among species as indicated by pollination systems. We confirm that networks are consistently modular across space and that interactions within pollination systems principally occur inside modules. Despite a strong tendency, we did not find a significant effect of pollination systems on the spatial consistency of pairwise interactions. These results indicate that the spatial rewiring of interactions could be constrained by pollination systems, resulting in conserved network structures in spite of high variation in pairwise interactions. Our findings suggest a relevant role of pollination systems in structuring plant-pollinator networks and we argue that structural patterns at the sub-network level can help us to fully understand how and why interactions vary across space and time.

The phylogenetic structure of plant-pollinator networks increases with habitat size and isolation.

Similarity among species in traits related to ecological interactions is frequently associated with common ancestry. Thus, closely related species usually interact with ecologically similar partners, which can be reinforced by diverse co-evolutionary processes. The effect of habitat fragmentation on the phylogenetic signal in interspecific interactions and correspondence between plant and animal phylogenies is, however, unknown. Here, we address to what extent phylogenetic signal and co-phylogenetic congruence of plant-animal interactions depend on habitat size and isolation by analysing the phylogenetic structure of 12 pollination webs from isolated Pampean hills. Phylogenetic signal in interspecific interactions differed among webs, being stronger for flower-visiting insects than plants. Phylogenetic signal and overall co-phylogenetic congruence increased independently with hill size and isolation. We propose that habitat fragmentation would erode the phylogenetic structure of interaction webs. A decrease in phylogenetic signal and co-phylogenetic correspondence in plant-pollinator interactions could be associated with less reliable mutualism and erratic co-evolutionary change.

Hot spots of mutualistic networks.

Incorporating interactions into a biogeographical framework may serve to understand how interactions and the services they provide are distributed in space. We begin by simulating the spatiotemporal dynamics of realistic mutualistic networks inhabiting spatial networks of habitat patches. We proceed by comparing the predicted patterns with the empirical results of a set of pollination networks in isolated hills of the Argentinian Pampas. We first find that one needs to sample up to five times as much area to record interactions as would be needed to sample the same proportion of species. Secondly, we find that peripheral patches have fewer interactions and harbour less nested networks - therefore potentially less resilient communities - compared to central patches. Our results highlight the important role played by the structure of dispersal routes on the spatial distribution of community patterns. This may help to understand the formation of biodiversity hot spots.

Beta diversity of plant-pollinator networks and the spatial turnover of pairwise interactions.

Interactions between species form complex networks that vary across space and time. Even without spatial or temporal constraints mutualistic pairwise interactions may vary, or rewire, across space but this variability is not well understood. Here, we quantify the beta diversity of species and interactions and test factors influencing the probability of turnover of pairwise interactions across space. We ask: 1) whether beta diversity of plants, pollinators, and interactions follow a similar trend across space, and 2) which interaction properties and site characteristics are related to the probability of turnover of pairwise interactions. Geographical distance was positively correlated with plant and interaction beta diversity. We find that locally frequent interactions are more consistent across space and that local flower abundance is important for the realization of pairwise interactions. While the identity of pairwise interactions is highly variable across space, some species-pairs form interactions that are locally frequent and spatially consistent. Such interactions represent cornerstones of interacting communities and deserve special attention from ecologists and conservation planners alike.

Specialization and rarity predict nonrandom loss of interactions from mutualist networks.

The loss of interactions from mutualistic networks could foreshadow both plant and animal species extinctions. Yet, the characteristics of interactions that predispose them to disruption are largely unknown. We analyzed 12 pollination webs from isolated hills ("sierras"), in Argentina, ranging from tens to thousands of hectares. We found evidence of nonrandom loss of interactions with decreasing sierra size. Low interaction frequency and high specialization between interacting partners contributed additively to increase the vulnerability of interactions to disruption. Interactions between generalists in the largest sierras were ubiquitous across sierras, but many of them lost their central structural role in the smallest sierras. Thus, particular configurations of interaction networks, along with unique ecological relations and evolutionary pathways, could be lost forever after habitat reduction.

Interactive effects of large- and small-scale sources of feral honey-bees for sunflower in the Argentine Pampas.

Pollinators for animal pollinated crops can be provided by natural and semi-natural habitats, ranging from large vegetation remnants to small areas of non-crop land in an otherwise highly modified landscape. It is unknown, however, how different small- and large-scale habitat patches interact as pollinator sources. In the intensively managed Argentine Pampas, we studied the additive and interactive effects of large expanses (up to 2200 ha) of natural habitat, represented by untilled isolated "sierras", and narrow (3-7 m wide) strips of semi-natural habitat, represented by field margins, as pollinator sources for sunflower (Helianthus annus). We estimated visitation rates by feral honey-bees, Apis mellifera, and native flower visitors (as a group) at 1, 5, 25, 50 and 100 m from a field margin in 17 sunflower fields 0-10 km distant from the nearest sierra. Honey-bees dominated the pollinator assemblage accounting for >90% of all visits to sunflower inflorescences. Honey-bee visitation was strongly affected by proximity to the sierras decreasing by about 70% in the most isolated fields. There was also a decline in honey-bee visitation with distance from the field margin, which was apparent with increasing field isolation, but undetected in fields nearby large expanses of natural habitat. The probability of observing a native visitor decreased with isolation from the sierras, but in other respects visitation by flower visitors other than honey-bees was mostly unaffected by the habitat factors assessed in this study. Overall, we found strong hierarchical and interactive effects between the study large and small-scale pollinator sources. These results emphasize the importance of preserving natural habitats and managing actively field verges in the absence of large remnants of natural habitat for improving pollinator services.

Direct effects of habitat area on interaction diversity in pollination webs.

Island biogeography theory predicts that species richness increases with habitat area and declines with isolation. We expand this framework to address changes in the number of links and species in pollination webs from 12 isolated hills, ranging in area from tens to thousands of hectares, immersed in the agriculture matrix of the Argentine Pampas. We also studied whether total interaction frequency is partitioned more evenly among individual links in richer webs. Our results reveal a direct effect of area on the number of links and species present in each pollination web. However, link richness increased twofold faster than species richness with area. These area effects were not confounded by sampling effort or correlated incidence of exotic species, despite widespread habitat disturbance. Habitat proximity, an inverse measure of isolation, had a marginally significant influence on link but not on species richness. Increased link number was associated with decreasing dominance by any particular interaction and increasing interaction evenness. Despite the strong area effect, a rich pollination web sampled from a small, protected sierra suggests that simple conservation measures, such as reduced grazing and fire suppression, may effectively preserve much local interaction diversity.