Towards a mechanistic understanding of variation in aquatic food chain length.

Ecologists have long sought to understand variation in food chain length (FCL) among natural ecosystems. Various drivers of FCL, including ecosystem size, resource productivity and disturbance, have been hypothesised. However, when results are aggregated across existing empirical studies from aquatic ecosystems, we observe mixed FCL responses to these drivers. To understand this variability, we develop a unified competition-colonisation framework for complex food webs incorporating all of these drivers. With competition-colonisation tradeoffs among basal species, our model predicts that increasing ecosystem size generally results in a monotonic increase in FCL, while FCL displays non-linear, oscillatory responses to resource productivity or disturbance in large ecosystems featuring little disturbance or high productivity. Interestingly, such complex responses mirror patterns in empirical data. Therefore, this study offers a novel mechanistic explanation for observed variations in aquatic FCL driven by multiple environmental factors.

Dispersal network heterogeneity promotes species coexistence in hierarchical competitive communities.

Understanding the mechanisms of biodiversity maintenance is a fundamental issue in ecology. The possibility that species disperse within the landscape along differing paths presents a relatively unexplored mechanism by which diversity could emerge. By embedding a classical metapopulation model within a network framework, we explore how access to different dispersal networks can promote species coexistence. While it is clear that species with the same demography cannot coexist stably on shared dispersal networks, we find that coexistence is possible on unshared networks, as species can surprisingly form self-organised clusters of occupied patches with the most connected patches at the core. Furthermore, a unimodal biodiversity response to an increase in species colonisation rates or average patch connectivity emerges in unshared networks. Increasing network size also increases species richness monotonically, producing characteristic species-area curves. This suggests that, in contrast to previous predictions, many more species can co-occur than the number of limiting resources.

Habitat loss alters effects of intransitive higher-order competition on biodiversity: a new metapopulation framework.

Recent studies have suggested that intransitive competition, as opposed to hierarchical competition, allows more species to coexist. Furthermore, it is recognized that the prevalent paradigm, which assumes that species interactions are exclusively pairwise, may be insufficient. More importantly, whether and how habitat loss, a key driver of biodiversity loss, can alter these complex competition structures (and therefore species coexistence) remain unclear. We thus present a new, simple yet comprehensive metapopulation framework that can account for any competition pattern and more complex higher-order interactions (HOIs) among species. We find that competitive intransitivity increases community diversity and that HOIs generally enhance this effect. Essentially, intransitivity promotes species richness by preventing the dominance of a few species, unlike the hierarchical competition, while HOIs facilitate species coexistence through stabilizing community fluctuations. However, variation in species' vital rates and habitat loss can weaken or even reverse such higher-order effects, as their interaction can lead to a more rapid decline in competitive intransitivity under HOIs. Thus, it is essential to correctly identify the most appropriate interaction model for a given system before models are used to inform conservation efforts. Overall, our simple model framework provides a more parsimonious explanation for biodiversity maintenance than the existing theory.

Habitat heterogeneity mediates effects of individual variation on spatial species coexistence.

Numerous studies have documented the importance of individual variation (IV) in determining the outcome of competition between species. However, little is known about how the interplay between IV and habitat heterogeneity (i.e. variation and spatial autocorrelation in habitat quality) affects species coexistence at the landscape scale. Here, we incorporate habitat heterogeneity into a competition model with IV, in order to explore the mechanism of spatial species coexistence. We find that individual-level variation and habitat heterogeneity interact to promote species coexistence, more obviously at lower dispersal rates. This is in stark contrast to early non-spatial models, which predicted that IV reinforces competitive hierarchies and therefore speeds up species exclusion. In essence, increasing variation in patch quality and/or spatial habitat autocorrelation moderates differences in the competitive ability of species, thereby allowing species to coexist both locally and globally. Overall, our theoretical study offers a mechanistic explanation for emerging empirical evidence that both habitat heterogeneity and IV promote species coexistence and therefore biodiversity maintenance.

Food web persistence in fragmented landscapes.

Habitat destruction, characterized by patch loss and fragmentation, is a key driver of biodiversity loss. There has been some progress in the theory of spatial food webs; however, to date, practically nothing is known about how patch configurational fragmentation influences multi-trophic food web dynamics. We develop a spatially extended patch-dynamic model for different food webs by linking patch connectivity with trophic-dependent dispersal (i.e. higher trophic levels displaying longer-range dispersal). Using this model, we find that species display different sensitivities to patch loss and fragmentation, depending on their trophic position and the overall food web structure. Relative to other food webs, omnivory structure significantly increases system robustness to habitat destruction, as feeding on different trophic levels increases the omnivore's persistence. Additionally, in food webs with a dispersal-competition trade-off between species, intermediate levels of habitat destruction can enhance biodiversity by creating refuges for the weaker competitor. This demonstrates that maximizing patch connectivity is not always effective for biodiversity maintenance, as in food webs containing indirect competition, doing so may lead to further species loss.

Robustness of metacommunities with omnivory to habitat destruction: disentangling patch fragmentation from patch loss.

Habitat destruction, characterized by patch loss and fragmentation, is a major driving force of species extinction, and understanding its mechanisms has become a central issue in biodiversity conservation. Numerous studies have explored the effect of patch loss on food web dynamics, but ignored the critical role of patch fragmentation. Here we develop an extended patch-dynamic model for a tri-trophic omnivory system with trophic-dependent dispersal in fragmented landscapes. We found that species display different vulnerabilities to both patch loss and fragmentation, depending on their dispersal range and trophic position. The resulting trophic structure varies depending on the degree of habitat loss and fragmentation, due to a tradeoff between bottom-up control on omnivores (dominated by patch loss) and dispersal limitation on intermediate consumers (dominated by patch fragmentation). Overall, we find that omnivory increases system robustness to habitat destruction relative to a simple food chain.

Diverse responses of species to landscape fragmentation in a simple food chain.

Habitat destruction, characterized by habitat loss and fragmentation, is a key driver of species extinction in spatial extended communities. Recently, there has been some progress in the theory of spatial food webs, however to date practically little is known about how habitat configurational fragmentation influences multi-trophic food web dynamics. To explore how habitat fragmentation affects species persistence in food webs, we introduce a modelling framework that describes the site occupancy of species in a tri-trophic system. We assume that species dispersal range increases with trophic level, exploiting pair-approximation techniques to describe the effect of habitat clustering. In accordance with the trophic rank hypothesis, both habitat loss and fragmentation generally cause species extinction, with stronger effects occurring at higher trophic levels. However, species display diverse responses (negative, neutral or positive) to habitat loss and fragmentation separately, depending on their dispersal range and trophic position. Counter-intuitively, prey species may benefit from habitat loss due to a release in top-down control. Similarly, habitat fragmentation has almost no influence on the site occupancy of the intermediate consumer in the tri-trophic system, though it decreases those of both basal species and top predator. Consequently, species' responses to habitat destruction vary as other species become extinct. Our results reiterate the importance of the interplay between bottom-up and top-down control in trophically linked communities, and highlight the complex responses occurring in even a simple food chain.

Coexistence of species with different dispersal across landscapes: a critical role of spatial correlation in disturbance.

Disturbance is key to maintaining species diversity in plant communities. Although the effects of disturbance frequency and extent on species diversity have been studied, we do not yet have a mechanistic understanding of how these aspects of disturbance interact with spatial structure of disturbance to influence species diversity. Here we derive a novel pair approximation model to explore competitive outcomes in a two-species system subject to spatially correlated disturbance. Generally, spatial correlation in disturbance favoured long-range dispersers, while distance-limited dispersers were greatly suppressed. Interestingly, high levels of spatial aggregation of disturbance promoted long-term species coexistence that is not possible in the absence of disturbance, but only when the local disperser was intrinsically competitively superior. However, spatial correlation in disturbance led to different competitive outcomes, depending on the disturbed area. Concerning ecological conservation and management, we theoretically demonstrate that introducing a spatially correlated disturbance to the system or altering an existing disturbance regime can be a useful strategy either to control species invasion or to promote species coexistence. Disturbance pattern analysis may therefore provide new insights into biodiversity conservation.

Species coexistence in a lattice-structured habitat: effects of species dispersal and interactions.

Opinions differ on how the spatial distribution of species over space affects species coexistence. Here, we constructed both mean-field and pair approximation (PA) models to explore the effects of interspecific and intraspecific interactions and dispersal modes on species coexistence. We found that spatial structure resulting from species dispersal traits and neighboring interactions in PA model did not promote coexistence if two species had the same traits, though it might intensify the contact frequency of intraspecific competition. If two species adopt different dispersal modes, the spatial structure in PA would make the coexistence or founder control less likely since it alters the species effective birth rate. This suggests that the spatial distribution caused by neighboring interactions and local dispersal does not affect species coexistence unless it adequately alters the effective birth rate for two species. Besides, we modeled how the initial densities and patterns affected population dynamics and revealed how the final spatial pattern was generated.

Habitat amount modulates biodiversity responses to fragmentation.

Anthropogenic habitat destruction leads to habitat loss and fragmentation, both of which interact to determine how biodiversity changes at the landscape level. While the detrimental effects of habitat loss are clear, there is a long-standing debate about the role of habitat fragmentation per se. We identify the influence of the total habitat amount lost as a modulator of the relationship between habitat fragmentation and biodiversity. Using a simple metacommunity model characterized by colonization-competition (C-C) trade-offs, we show that the magnitude of habitat loss can induce a unimodal response of biodiversity to habitat fragmentation. When habitat loss is low, habitat fragmentation promotes coexistence by suppressing competitively dominant species, while habitat fragmentation at high levels of habitat loss can shape many smaller isolated patches that drive extinctions of superior competitors. While the C-C trade-off is not the only mechanism for biodiversity maintenance, the modulation of habitat fragmentation effects by habitat loss is probably common. Reanalysis of a globally distributed dataset of fragmented animal and plant metacommunities shows an overall pattern that supports this hypothesis, suggesting a resolution to the debate regarding the relative importance of positive versus negative fragmentation effects.

Species persistence in landscapes with spatial variation in habitat quality: a pair approximation model.

Habitat degradation has become a major threat to species persistence. Although several models have explicitly integrated habitat quality into metapopulation dynamics, we still lack knowledge of the spatial variability of species persistence which may result from the clustering of habitat patches of differing quality. Here we construct both pair approximation (PA) and cellular automaton (CA) models for species persistence in homogeneous versus heterogeneous landscapes. Heterogeneous landscapes are generated by varying the orthogonal-neighbour correlation between two different-quality habitats. In our simulations, the PA model exhibits similar population dynamics to the CA model, though it overestimates species persistence due to the doublet approximation neglecting correlation beyond nearest neighbours. Generally, landscape heterogeneity enhances species persistence relative to landscape homogeneity, especially with enlarging habitat-quality difference. This indicates that models based on homogeneous landscapes may overestimate species extinction rate. In heterogeneous landscapes, habitat clumping does not influence global dispersers because of random establishment, although it does promote the persistence of local dispersers, especially under severe habitat degradation. However, habitat configurational fragmentation improves the persistence of global dispersers that are highly sensitive to local crowding, probably by reducing density dependence, but this positive fragmentation effect on local dispersers is overshadowed by the stronger negative border effect on impeding local extension. Furthermore, increasing density dependence promotes the extinction risk of local dispersers, while global dispersers are not influenced. For conservation and habitat management, our results suggest that minimising random anthropogenic disturbance should take priority over increasing the connectivity of good-quality habitat, as random habitat degradation poses a more serious threat to species persistence than clustered habitat degradation. Owing to species' diverse responses to habitat configurational fragmentation, landscapes with different properties may accommodate different species.

Colonization-competition dynamics of basal species shape food web complexity in island metacommunities.

Exploring how food web complexity emerges and evolves in island ecosystems remains a major challenge in ecology. Food webs assembled from multiple islands are commonly recognized as highly complex trophic networks that are dynamic in both space and time. In the context of global climate change, it remains unclear whether food web complexity will decrease in a monotonic fashion when undergoing habitat destruction (e.g., the inundation of islands due to sea-level rise). Here, we develop a simple yet comprehensive patch-dynamic framework for complex food web metacommunities subject to the competition-colonization tradeoff between basal species. We found that oscillations in food web topological complexity (characterized by species diversity, mean food chain length and the degree of omnivory) emerge along the habitat destruction gradient. This outcome is robust to changing parameters or relaxing the assumption of a strict competitive hierarchy. Having oscillations in food web complexity indicates that small habitat changes could have disproportionate negative effects on species diversity, thus the success of conservation actions should be evaluated not only on changes in biodiversity, but also on system robustness to habitat alteration. Overall, this study provides a parsimonious mechanistic explanation for the emergence of food web complexity in island ecosystems, further enriching our understanding of metacommunity assembly. Supplementary Information: The online version contains supplementary material available at 10.1007/s42995-023-00167-0.

Complex nonmonotonic responses of biodiversity to habitat destruction.

It has typically been assumed that habitat destruction, characterized by habitat loss and fragmentation, has consistently negative effects on biodiversity. While numerous empirical studies have shown the detrimental effects of habitat loss, debate continues as to whether habitat fragmentation has universally negative effects. To explore the effects of habitat fragmentation, we developed a simple model for site-occupancy dynamics in fragmented landscapes. With the model, we demonstrate that a competition-colonization trade-off can result in nonlinear oscillatory responses in biodiversity to both habitat loss and fragmentation. However, the overall pattern of habitat loss reducing species richness is still established, in line with empirical observations. Interestingly, the existence of localized oscillations in biodiversity can explain the mixed responses of species richness to habitat fragmentation per se observed in nature, thereby reconciling the debate on the fragmentation-diversity relationship. Therefore, this study offers a parsimonious mechanistic explanation for empirically observed biodiversity patterns in response to habitat destruction.

Competition-colonization dynamics and multimodality in diversity-disturbance relationships.

Disturbance has long been recognized as a critical driver of species diversity in community ecology. Recently, it has been found that the well-known intermediate disturbance hypothesis, which predicts a unimodal diversity-disturbance relationship (DDR), fails to describe numerous experimental observations, as empirical DDRs are diverse. Consequently, the precise form of the DDR remains a topic of debate. Here we develop a simple yet comprehensive metacommunity framework that can account for complex competition patterns. Using both numerical simulations and analytical arguments, we show that strongly multimodal DDRs arise naturally, and this multimodality is quite robust to changing parameters or relaxing the assumption of a strict competitive hierarchy. Having multimodality as a robust property of DDRs in competition models suggests that much of the noise observed in empirical DDRs could be a critical signature of the underlying competitive dynamics.

A patch-dynamic metacommunity perspective on the persistence of mutualistic and antagonistic bipartite networks.

The structure of interactions between species within a community plays a key role in maintaining biodiversity. Previous studies found that the effects of these structures might vary substantially depending on interaction type, for example, a highly connected and nested architecture stabilizes mutualistic communities, while the stability of antagonistic communities is enhanced in modular and weakly connected structures. Here we show that, when network dynamics are modeled using a patch-dynamic metacommunity framework, the qualitative differences between antagonistic and mutualistic systems disappear, with nestedness and modularity interacting to promote metacommunity persistence. However, the interactive effects are significantly weaker in antagonistic metacommunities. Our model also predicts an increase in connectance, nestedness, and modularity over time in both types of interaction, except in antagonistic networks, where nestedness declines. At steady state, we find a strong negative correlation between nestedness and modularity in both mutualistic and antagonistic metacommunities. These predictions are consistent with the structural trends found in a large data set of real-world antagonistic and mutualistic communities.

Multilayer network structure enhances the coexistence of competitive species.

The concept of a multiplex network can be used to characterize the dispersal paths and states of different species in a patch habitat system. The multiplex network is one of three types of multilayer networks. In this study, the effect of a multiplex network on the long-term stable coexistence of species is investigated using the concept of metapopulation. Based on the mean field approximation, the stability analysis of a two-species system shows that, compared to the single layer network, the multiplex network is more conducive to the stable coexistence of species when one species has a stronger colonization ability. That is, in such a patch habitat system, if the dispersal paths of the stronger species are different than those of the weaker species, then the larger the heterogeneity of the dispersal network of the stronger species is, the more likely the long-term stable coexistence of species. This result provides a different perspective for understanding the biodiversity in heterogeneous habitats.

The role of omnivory in mediating metacommunity robustness to habitat destruction.

Omnivores have long been known to play an important role in determining the stability of ecological communities. Recent theoretical studies have suggested that they may also increase the resilience of their communities to habitat destruction, one of the major drivers of species extinctions globally. However, these outcomes were obtained for minimal food webs consisting of only a single omnivore and its prey species, while much more complex communities can be anticipated in nature. In this study, we undertake a systematic comparative analysis of the robustness of metacommunities containing various omnivory structures to habitat loss and fragmentation using a mathematical model. We observe that, in general, omnivores are better able to survive facing habitat destruction than specialist predators of similar trophic level. However, the community as a whole does not always benefit from the presence of omnivores, as they may drive their intraguild prey to extinction. We also analyze the frequency with which these modules occur in a set of empirical food webs, and demonstrate that variation in their rate of occurrence is consistent with our model predictions. Our findings demonstrate the importance of considering the complete food web in which an omnivore is embedded, suggesting that future study should focus on more holistic community analysis.

Metacommunity robustness of plant-fly-wasp tripartite networks with specialization to habitat loss.

Recent observations have found plant-species-specific fly-host selection (i.e., specialization) of wasp parasitoids (wasps) in plant-fly-wasp (P-F-W) tripartite networks, yet no study has explored the dynamical implications of such high-order specialization for the persistence of this network. Here we develop a patch-dynamic framework for a unique P-F-W tripartite network with specialization observed in eastern Tibetan Plateau and explore its metacommunity robustness to habitat loss. We show that specialization in parasitoidism promotes fly species diversity, while the richness of both plant and wasp decreases. Compared to other two null models, real network structure favors plant species coexistence but increases the extinction risk for both flies and wasps. However, these effects of specialization and network structure would be weakened and ultimately disappear with increasing habitat loss. Interestingly, intermediate levels of habitat loss can maximize the diversity of flies and wasps, while increasing or decreasing habitat loss results in more species losses, supporting intermediate disturbance hypothesis. Finally, we observe that high levels of habitat loss initiate a bottom-up cascade of species extinction from plants to both flies and wasps, resulting in a rapid collapse of the whole tripartite networks. Overall, this theoretical framework is the first attempt to characterize the dynamics of whole tripartite metacommunities interacting in realistic high-order ways, offering new insights into complex multipartite networks.

Modeling the impact of reproductive mode on masting.

Masting is defined as the intermittent highly variable production of seed in a plant population. According to reproductive modes, that is, sexual and asexual reproduction, masting species can be separated into three groups, that is, (1) species, for example, bamboo, flower only once before they die; (2) species, for example, Fagus, reproduce sexually; and (3) species, for example, Stipa tenacissima, reproduce both sexually and asexually. Several theories have been proposed to explore the underlying mechanisms of masting. However, to our knowledge, no theory has been found to explain the mechanism of masting species that reproduce both sexually and asexually. Here we refine the Resource Budget Model by considering a trade-off between sexual and asexual reproduction. Besides the depletion efficient (i.e., the ratio of the cost of seed setting and the cost of flowering), other factors, such as the annual remaining resource (i.e., the rest of the resource from the photosynthetic activity after allocating to growth and maintenance), the trade-off between sexual and asexual reproduction, and the reproductive thresholds, also affect masting. Moreover, two potential reproductive strategies are found to explain the mechanisms: (1) When the annual remaining resource is relatively low, plants reproduce asexually and a part of the resource is accumulated as the cost of asexual reproduction is less than the annual remaining resource. Plants flower and set fruits once the accumulated resource exceeds the threshold of sexual reproduction; (2) when the annual remaining resource is relatively high, and the accumulated resource surpasses the threshold of sexual reproduction, masting occurs. Remarkably, under certain depletion efficient, more investigation in sexual reproduction will lead plants to reproduce periodically. Additionally, plants investigate less resource to reproduce periodically when depletion efficient keeps increasing as plants can reproduce efficiently. Overall, our study provides new insights into the interpretation of masting, especially for species that reproduce both sexually and asexually.

An extended patch-dynamic framework for food chains in fragmented landscapes.

Habitat destruction, a key determinant of species loss, can be characterized by two components, patch loss and patch fragmentation, where the former refers to the reduction in patch availability, and the latter to the division of the remaining patches. Classical metacommunity models have recently explored how food web dynamics respond to patch loss, but the effects of patch fragmentation have largely been overlooked. Here we develop an extended patch-dynamic model that tracks the patch occupancy of the various trophic links subject to colonization-extinction-predation dynamics by incorporating species dispersal with patch connectivity. We found that, in a simple food chain, species at higher trophic level become extinct sooner with increasing patch loss and fragmentation due to the constraint in resource availability, confirming the trophic rank hypothesis. Yet, effects of fragmentation on species occupancy are largely determined by patch loss, with maximal fragmentation effects occurring at intermediate patch loss. Compared to the spatially explicit simulations that we also performed, the current model with pair approximation generates similar community patterns especially in spatially clustered landscapes. Overall, our extended framework can be applied to model more complex food webs in fragmented landscapes, broadening the scope of existing metacommunity theory.