Hidden effects of habitat restoration on the persistence of pollination networks

Abstract Past and recent studies have focused on the effects of global change drivers such as species invasions on species extinction. However, as we enter the United Nations Decade of Ecosystem Restoration the aim must switch to understanding how invasive‐species management affects the persistence of the remaining species in a community. Focusing on plant‐pollinator interactions, we test how species persistence is affected by restoration via the removal of invasive plant species. Restoration had a clear positive effect on plant persistence, whereas there was no difference between across treatments for pollinator persistence in the early season, but a clear effect in late season, with higher persistence in unrestored sites. Network structure affected only pollinator persistence, while centrality had a strong positive effect on both plants and pollinators. Our results suggest a hidden effect of invasive plants—although they may compete with native plant species, invasive plants may provide important resources for pollinators, at least in the short term.

).  Figure S1: Invasive plant species removal had no effect on population persistence between restored (R) and unrestored (U) sites, neither for plants (a), nor pollinators (b), at each different sampling month (numbers on the top). Box plot represent median (mid line), interquartile range (box edges), and 1.5 x interquartile range (whiskers). Equivalent to Fig. 1 in main text but with the monthly-level data (as opposed to early and late season). Bee image by Melissa Broussard, CC BY 3.0, courtesy of PhyloPic.org. 7 For the month-level data we found that specialization had a clear, positive effect on the persis- , and this effect was stronger in restored (green) than in unrestored (orange) sites. Specialization is scaled, solid line indicates the mean slope estimate from the GLMM model, and the shaded area represents the 95% confidence intervals around the estimate. Results shown for month-level data.

2) Effect of specialization (d') on species persistence
We found similar patterns for the season-level data, with specialization having a positive

30
As in the analyses in the main text, we used generalized linear mixed-effects models to test 31 the effect of restoration status on species persistence. We included month, restoration status, and 32 their interaction as fixed effects and included by-site random intercepts. We assumed a beta-error 33 distribution. We used a similar model to test for effects of restoration status and seasonality on 34 species persistence, swapping the fixed effects of month and month-restoration interaction for 35 fixed effects of season and a season-restoration interaction. 36 We expected that persistence of both plants and pollinators would be higher at restored sites, Pollinators Persistence b Figure S3: Invasive plant species removal had no effect on (a) plants, nor (b) pollinators persistence. Box plot represent median (mid line), interquartile range (box edges), and 1.5 x interquartile range (whiskers). Equivalent to Fig. S1 Figure S4: (a) Plant persistence was higher in restored sites, especially in late season whereas (b) pollinator persistence was higher in the late season, especially in unrestored sites. Box plot represent median (mid line), interquartile range (box edges), and 1.5 x interquartile range (whiskers). Equivalent to Fig. 1 in main text but including vertebrates. Bee image by Melissa Broussard, CC BY 3.0, courtesy of PhyloPic.org.
To test whether nestedness and modularity contribute to species persistence, we used a general-46 ized linear mixed-effects model with a beta-error distribution, including fixed effects of restoration status, network structure (either nestedness or modularity), and their interaction. We fit 48 by-site random intercepts. To account for the fact that species richness is correlated with nested-49 ness and modularity, we also included a quadratic fixed effect for the total number of species.

50
For the month-level data, we found no effect of network structure on plant persistence for  Figure S5: Nestedness (a) and modularity (b) and no effect on plant persistence. In contrast, nestedness had a strong positive effect on pollinator persistence (c) whereas modularity had no effect on pollinator persistence (d). Solid lines indicate the mean slope estimate from the GLMM model, and the shaded area represents the 95% confidence intervals around the estimate. Equivalent to Fig. 2

3.3) Effect of centrality on species persistence 60
To test whether species that are more central in their interaction networks have greater persis-61 tence than more peripheral species, we fitted a binomial linear mixed-effects model to the num-