When cheating turns into a stabilizing mechanism of plant-pollinator communities

Abstract

Mutualistic interactions, such as plant-mycorrhizal or plant-pollinator interactions, are widespread in ecological communities and frequently exploited by cheaters, species that profit from interactions without providing benefits in return. Cheating usually negatively affects the fitness of the individuals that are cheated on, but the effects of cheating at the community level remains poorly understood. Here, we describe 2 different kinds of cheating in mutualistic networks and use a generalized Lotka-Volterra model to show that they have very different consequences for the persistence of the community. Conservative cheating, where a species cheats on its mutualistic partners to escape the cost of mutualistic interactions, negatively affects community persistence. In contrast, innovative cheating occurs with species with whom legitimate interactions are not possible, because of a physiological or morphological barrier. Innovative cheating can enhance community persistence under some conditions: when cheaters have few mutualistic partners, cheat at low or intermediate frequency and the cost associated with mutualism is not too high. Under these conditions, the negative effects of cheating on partner persistence are overcompensated at the community level by the positive feedback loops that arise in diverse mutualistic communities. Using an empirical dataset of plant-bird interactions (hummingbirds and flowerpiercers), we found that observed cheating patterns are highly consistent with theoretical cheating patterns found to increase community persistence. This result suggests that the cheating patterns observed in nature could contribute to promote species coexistence in mutualistic communities, instead of necessarily destabilizing them.

Fig 1

Cheating can be conservative or innovative. (a) Above, mutualistic network built using only +/+ (legitimate) interactions and below, mutualistic networks including 1 illegitimate interaction. Cheating is conservative if it happens between partners that were already linked by a mutualistic interaction, or innovative if it happens between partners that were not linked by a mutualistic interaction. (b) Cheating characterized in a niche concept. Symbols represent the position of the legitimate (green) and illegitimate (yellow) partners of a given species of pollinator in their trait space. Conservative cheating occurs within the mutualistic niche, which is the convex hull determined by trait combination of legitimate partners. Innovative cheating occurs with plants that are outside of the mutualistic niche and lead to an expansion of the interaction niche, defined as the convex hull determined by trait combination of legitimate and illegitimate partners. (c) An example of conservative cheating, because the same species of hummingbird (Coeligena lutetiae) visits legitimately (middle) and illegitimately (right) flowers of Centropogon pichinchensis. (d) An example of innovative cheating, where Eupherusa eximia visits illegitimately a flower of Symbolanthus pulcherrimus, which has a corolla much longer than the bill of the hummingbird, preventing legitimate interaction.

Fig 2

Innovative cheating can increase network persistence. Network persistence (percentage of persisting species at equilibrium) as a function of cheating frequency (Ω), proportion of innovative cheating (Ψ), cost associated with mutualism (Λ) and scenario (cheaters are generalist or specialist species) for (a) a case with 10% and (b) a case with 50% of animal species cheating. The colors represent the average effect of cheating on network persistence relative to the corresponding case with no cheating (Ω = 0). Positive values (blue) indicate higher persistence with than without cheating, while negative values (red) indicate the opposite, a negative effect of cheating on persistence. Results are presented for initial values of connectance φ = 0.4 and averaged over the 500 initial conditions. The black boxes represent the space in which cheating has a positive effect on network persistence, which is represented in 3D for the specialist scenario in (c) for Λ = 0 (gray) and Λ = 0.3 (dark blue). (d) Represents the volume of this 3D space as a function of the scenario and the cost associated with mutualism. Note that the full parameter space presented in (c) and (d) has a volume of 1, so the absolute volume and the fraction of volume in which cheating have a positive effect are the same measure and that a cost of Λ = 0.15 or Λ = 0.3 corresponds to 10% or 20% of benefits associated with mutualism, respectively (Λ/α = 0.1 or 0.2). The data underlying this figure can be found in https://doi.org/10.5281/zenodo.10102438.

Fig 3

Effect of cheating is mediated by interference among pollinators and diversity. (a) Average effect of cheating on persistence, with or without competition among pollinators for partners (plants), when cheaters are specialists, for an initial value of connectance φ = 0.4. The case with competition is the same than in Fig 2A. (b) Effect of cheating on persistence in small communities, with 20 species (10 per guild) as a function effect of cheating on persistence in diverse communities, with 40 species (20 per guild). For Λ = 0.15 and φ = 0.4. The bottom part is a zoom on the gray area of the x-axis. The solid black line shows the first bisector while dashed black lines show the zero values (no effect of cheating on persistence). Each point is a simulation with the same parameter combination, excepting the number of species. (c) Volume of the 3D cheating parameter space in which cheating has a positive effect on network persistence, as a function of the scenario and the diversity, for Λ = 0.15 and φ = 0.4. Note that when cheaters are generalist species, there was no positive effects of cheating (when Λ = 0.15), so the volumes equal zero. The data underlying this figure can be found in https://doi.org/10.5281/zenodo.10102438.

Fig 4

Empirical cheating patterns in plant–bird interaction networks fit theoretical conditions to maximize network persistence. (a) An empirical network of interactions between plants (green) and birds (blue) from Ecuador for a given site. Thickness of the lines is proportional to interaction frequency, while color of the lines represents cheating frequency, gray (only legitimate) and from pink (low cheating frequency) to bright red (only illegitimate). (b) Frequency of cheating as a function of mutualistic partner diversity, for each combination of bird species and site, for datasets from Costa Rica (beige) and Ecuador (dark blue). The size of the point represents the total number of interactions of the species in the given site. The solid lines show predictions of a generalized linear mixed-effects model with associated 95% confidence interval. (c) Proportions of innovative cheating, where each point is a bird species who cheated at least once. The size of the point represents the total number of interactions of the species summed over sites. (d) Proportion of cheaters and (e) overall level of cheating, for each site, as a function of elevation (in meters above sea level). The solid lines show predictions of generalized linear models with associated 95% confidence intervals. The data underlying this figure can be found in https://doi.org/10.5281/zenodo.10102438.

Fig 5

Parameterizing theoretical models with empirical data reveal that cheating can increase network persistence, when cost associated with mutualism is low. Effect of cheating on network persistence when parameterizing the model with empirical patterns of cheating, as a function of per-interaction benefit (α), strength of competition for mutualistic partners (c) and cost associated with mutualism (Λ). The effect of cheating on persistence is calculated by subtracting the persistence values obtained with the simulations parameterize with all interactions (including cheating) from those obtained when only mutualistic interactions are used to parameterize the model (excluding cheating). Points represent the average effects over the 100 simulations performed over different growth rate vectors, while error bars represent the standard deviation. Black cross represents the overall average (over sites) and associated 95% confidence interval (vertical error bars, often hidden because too small). Note that the x-axis is not linear, there is a jump between Λ = 0.15 and Λ = 0.3. The data underlying this figure can be found in https://doi.org/10.5281/zenodo.10102438.