Species loss and secondary extinctions in simple and complex model communities

1. The loss of a species from an ecological community can trigger a cascade of secondary extinctions. Here we investigate how the complexity (connectance) of model communities affects their response to species loss. Using dynamic analysis based on a global criterion of persistence (permanence) and topological analysis we investigate the extent of secondary extinctions following the loss of different kinds of species. 2. We show that complex communities are, on average, more resistant to species loss than simple communities: the number of secondary extinctions decreases with increasing connectance. However, complex communities are more vulnerable to loss of top predators than simple communities. 3. The loss of highly connected species (species with many links to other species) and species at low trophic levels triggers, on average, the largest number of secondary extinctions. The effect of the connectivity of a species is strongest in webs with low connectance. 4. Most secondary extinctions are due to direct bottom-up effects: consumers go extinct when their resources are lost. Secondary extinctions due to trophic cascades and disruption of predator-mediated coexistence also occur. Secondary extinctions due to disruption of predator-mediated coexistence are more common in complex communities than in simple communities, while bottom-up and top-down extinction cascades are more common in simple communities. 5. Topological analysis of the response of communities to species loss always predicts a lower number of secondary extinctions than dynamic analysis, especially in food webs with high connectance.     

The form of direct interspecific competition modifies secondary extinction patterns in multi‐trophic food webs

Forcibly removing species from ecosystems has important consequences for the remaining assemblage, leading to changes in community structure, ecosystem functioning and secondary (cascading) extinctions. One key question that has arisen from single‐ and multi‐trophic ecosystem models is whether the secondary extinctions that occur within competitive communities (guilds) are also important in multi‐trophic ecosystems? The loss of consumer–resource links obviously causes secondary extinction of specialist consumers (topological extinctions), but the importance of secondary extinctions in multi‐trophic food webs driven by direct competitive exclusion remains unknown. Here I disentangle the effects of extinctions driven by basal competitive exclusion from those caused by trophic interactions in a multi‐trophic ecosystem (basal producers, intermediate and top consumers). I compared food webs where basal species either show diffuse (all species compete with each other identically: no within guild extinctions following primary extinction) or asymmetric competition (unequal interspecific competition: within guild extinctions are possible). Basal competitive exclusion drives extra extinction cascades across all trophic levels, with the effect amplified in larger ecosystems, though varying connectance has little impact on results. Secondary extinction patterns based on the relative abundance of the species lost in the primary extinction differ qualitatively between diffuse and asymmetric competition. Removing asymmetric basal species with low (high) abundance triggers fewer (more) secondary extinctions throughout the whole food web than removing diffuse basal species. Rare asymmetric competitors experience less pressure from consumers compared to rare diffuse competitors. Simulations revealed that diffuse basal species are never involved in extinction cascades, regardless of the trophic level of a primary extinction, while asymmetric competitors were. This work highlights important qualitative differences in extinction patterns that arise when different assumptions are made about the form of direct competition in multi‐trophic food webs.     

Cascading extinctions and community collapse in model food webs

Species loss in ecosystems can lead to secondary extinctions as a result of consumer–resource relationships and other species interactions. We compare levels of secondary extinctions in communities generated by four structural food-web models and a fifth null model in response to sequential primary species removals. We focus on various aspects of food-web structural integrity including robustness, community collapse and threshold periods, and how these features relate to assumptions underlying different models, different species loss sequences and simple measures of diversity and complexity. Hierarchical feeding, a fundamental characteristic of food-web structure, appears to impose a cost in terms of robustness and other aspects of structural integrity. However, exponential-type link distributions, also characteristic of more realistic models, generally confer greater structural robustness than the less skewed link distributions of less realistic models. In most cases for the more realistic models, increased robustness and decreased levels of web collapse are associated with increased diversity, measured as species richness S, and increased complexity, measured as connectance C. These and other results, including a surprising sensitivity of more realistic model food webs to loss of species with few links to other species, are compared with prior work based on empirical food-web data.     

Functional links and robustness in food webs

The robustness of ecosystems to species losses is a central question in ecology, given the current pace of extinctions and the many species threatened by human impacts, including habitat destruction and climate change. Robustness from the perspective of secondary extinctions has been addressed in the context of food webs to consider the complex network of species interactions that underlie responses to perturbations. In-silico removal experiments have examined the structural properties of food webs that enhance or hamper the robustness of ecosystems to species losses, with a focus on the role of hubs, the most connected species. Here we take a different approach and focus on the role of the connections themselves. We show that trophic links can be divided into functional and redundant based on their contribution to robustness. The analysis of empirical webs shows that hubs are not necessarily the most important species as they may hold many redundant links. Furthermore, the fraction of functional connections is high and constant across systems regardless of size and interconnectedness. The main consequence of this scaling pattern is that ecosystem robustness can be considerably reduced by species extinctions even when these do not result in any secondary extinctions. This introduces the possibility of tipping points in the collapse of ecosystems.     

Ecological communities are vulnerable to realistic extinction sequences

Loss of species will directly change the structure and potentially the dynamics of ecological communities, which in turn may lead to additional species loss (secondary extinctions) due to direct and/or indirect effects (e.g. loss of resources or altered population dynamics). Furthermore, the vulnerability of food webs to repeated species loss is expected to be affected by food web topology, species interactions, as well as the order in which species go extinct. Species traits such as body size, abundance and connectivity might determine a species’ vulnerability to extinction and, thus, the order in which species go primarily extinct. Yet, the sequence of primary extinctions, and their effects on the vulnerability of food webs to secondary extinctions, when species abundances are allowed to respond dynamically, has only recently become the focus of attention. Here, we analyse and compare topological and dynamical robustness to secondary extinctions of model food webs, in the face of 34 extinction sequences based on species traits. Although secondary extinctions are frequent in the dynamical approach and rare in the topological approach, topological and dynamical robustness tends to be correlated for many bottom–up directed, but not for top–down directed deletion sequences. Furthermore, removing species based on traits that are strongly positively correlated to the trophic position of species (such as large body size, low abundance, high net effect) is, under the dynamical approach, found to be as destructive as removing primary producers. Such top–down oriented removal of species are often considered to correspond to realistic extinction scenarios, but earlier studies, based on topological approaches, have found such extinction sequences to have only moderate effects on the remaining community. Thus, our result suggests that the structure of ecological communities, and therefore the integrity of important ecosystem processes could be more vulnerable to realistic extinction sequences than previously believed.     

Cascading extinctions and ecosystem functioning: contrasting effects of diversity depending on food web structure

The consequences of species loss on cascading extinctions in food webs have been the focus of several recent theoretical studies, with differing results. Changes in ecosystem properties consecutive to cascading extinctions have received far less attention even though such dramatic events might strongly alter ecosystem functioning. Here we use various food web models to investigate the effects of species loss and diversity on both secondary extinctions and their associated changes in ecosystem properties. Our analysis shows that diversity has contrasting effects depending on the presence of self-limiting terms at consumer levels and, to a lower extent, on connectance and interspecific competition. Ecosystems that lose a high proportion of species through cascading extinctions exhibit the most important changes in ecosystem properties. Linking studies on cascading extinctions in food webs with studies that investigate the effects of biodiversity on ecosystem functioning appears crucial for a better understanding of the consequences of species extinctions.     

How habitat-modifying organisms structure the food web of two coastal ecosystems

The diversity and structure of ecosystems has been found to depend both on trophic interactions in food webs and on other species interactions such as habitat modification and mutualism that form non-trophic interaction networks. However, quantification of the dependencies between these two main interaction networks has remained elusive. In this study, we assessed how habitat-modifying organisms affect basic food web properties by conducting in-depth empirical investigations of two ecosystems: North American temperate fringing marshes and West African tropical seagrass meadows. Results reveal that habitat-modifying species, through non-trophic facilitation rather than their trophic role, enhance species richness across multiple trophic levels, increase the number of interactions per species (link density), but decrease the realized fraction of all possible links within the food web (connectance). Compared to the trophic role of the most highly connected species, we found this non-trophic effects to be more important for species richness and of more or similar importance for link density and connectance. Our findings demonstrate that food webs can be fundamentally shaped by interactions outside the trophic network, yet intrinsic to the species participating in it. Better integration of non-trophic interactions in food web analyses may therefore strongly contribute to their explanatory and predictive capacity.     

Food webs robustness to biodiversity loss: the roles of connectance, expansibility and degree distribution

We analyse the robustness of food webs against species loss by considering the influence of several structural factors of the networks, such as connectance, degree distribution and expansibility. The last concept refers to the absence of structural bottlenecks in the food web, whose removal separate the network into large isolate clusters. In theory networks with identical connectance can display different expansibility characteristics. Using the spectral scaling method we studied 17 food networks and classified them as good expansion (GE) and not-GE networks. The combination of GE properties and degree distribution of species permitted the classification of food webs into six different classes. These classes characterize the differences in robustness of food webs to species loss. While the webs having uniform degree distributions and displaying GE properties are the most robust to species loss, the presence of bottlenecks and skewed distribution of the number of links per species make food webs very vulnerable to primary removal of species.     

Invasions and extinctions reshape coastal marine food webs

The biodiversity of ecosystems worldwide is changing because of species loss due to human-caused extinctions and species gain through intentional and accidental introductions. Here we show that the combined effect of these two processes is altering the trophic structure of food webs in coastal marine systems. This is because most extinctions ( approximately 70%) occur at high trophic levels (top predators and other carnivores), while most invasions are by species from lower trophic levels (70% macroplanktivores, deposit feeders, and detritivores). These opposing changes thus alter the shape of marine food webs from a trophic pyramid capped by a diverse array of predators and consumers to a shorter, squatter configuration dominated by filter feeders and scavengers. The consequences of the simultaneous loss of diversity at top trophic levels and gain at lower trophic levels is largely unknown. However, current research suggests that a better understanding of how such simultaneous changes in diversity can impact ecosystem function will be required to manage coastal ecosystems and forecast future changes.     

Connectance indicates the robustness of food webs when subjected to species loss

Many ecologists are concerned that biodiversity loss from human impact on natural ecosystems could compromise ecosystem stability. A relationship between diversity and stability was proposed by MacArthur [MacArthur, R.H., 1955. Fluctuation of animal populations and a measure of community stability. Ecology 36, 533–536.]. Current thinking (for example, McCann, K., 2000. The diversity–stability debate. Nature 405, 228–233.) acknowledges that interaction pattern among species, rather than species richness per se, is one element of this relationship. Dunne et al. [Dunne, J.A., Williams, R.J., Martinez, N.D., 2002a. Network structure and biodiversity loss in food webs: robustness increases with connectance. Ecol. Lett. 5, 558–567.] showed that the robustness of 16 food webs is correlated with their connectance. Connectance is one measure of interaction pattern. Robustness relates to the maintenance of network integrity and so has consequences for stability; the loss of integrity must have ecosystem-wide implications. This paper tests the hypothesis that changes in a food web's connectance indicate changes in its robustness. It concludes that any change in connectance with species loss, but especially large, negative changes, constitutes a decrease in robustness. Estimation of the change in connectance could support interpretation of monitoring data on species composition, acting as an indicator of food web robustness and, indirectly, of ecosystem stability. It could assist managers to understand the implications of biodiversity loss caused by human intervention in ecosystems, and could assist either choice of intervention or amelioration of impacts.     

Structural Degradation in Mediterranean Sea Food Webs: Testing Ecological Hypotheses Using Stochastic and Mass-Balance Modelling

Human-mediated disturbances such as fishing, habitat modification, and pollution have resulted in significant shifts in species composition and abundance in marine ecosystems which translate into degradation of food-web structure. Here, we used a comparative ecological modelling approach and data from two food webs (North-Central Adriatic and South Catalan Sea) and two time periods (mid-late 1970s and 1990s) in the Mediterranean Sea to evaluate how changes in species composition and biomass have affected food-web properties and the extent of ecosystem degradation. We assembled species lists and ecological information for both regions and time periods into stochastic structural and mass-balance food-web models, and compared the outcomes of 22 food-web properties. Our results show strong similarities in structural food-web properties between the North-Central Adriatic and South Catalan Seas indicating similar ecosystem structure and levels of ecological degradation between regions and time periods. In contrast, a comparison with other published marine food webs (Caribbean, Benguela, and US continental shelf) suggested that Mediterranean webs are in an advanced state of ecological degradation. This was reflected by lower trophic height, linkage density, connectance, omnivory, species involved in looping, trophic chain length and fraction of biomass at higher trophic levels, as well as higher generality and fraction of biomass at lower trophic levels. An analysis of robustness to simulated species extinction revealed lower robustness to species removals in Mediterranean webs and corroborated their advanced state of degradation. Importantly, the two modelling approaches used delivered comparable results suggesting that they both capture fundamental information about how food webs are structured. Electronic supplementary material  The online version of this article (doi:) contains supplementary material, which is available to authorized users.     

Early onset of secondary extinctions in ecological communities following the loss of top predators

The large vulnerability of top predators to human-induced disturbances on ecosystems is a matter of growing concern. Because top predators often exert strong influence on their prey populations their extinction can have far-reaching consequences for the structure and functioning of ecosystems. It has, for example, been observed that the local loss of a predator can trigger a cascade of secondary extinctions. However, the time lags involved in such secondary extinctions remain unexplored. Here we show that the loss of a top predator leads to a significantly earlier onset of secondary extinctions in model communities than does the loss of a species from other trophic levels. Moreover, in most cases time to secondary extinction increases with increasing species richness. If local secondary extinctions occur early they are less likely to be balanced by immigration of species from local communities nearby. The implications of these results for community persistence and conservation priorities are discussed.     

Trophically unique species are vulnerable to cascading extinction

Understanding which species might become extinct and the consequences of such loss is critical. One consequence is a cascade of further, secondary extinctions. While a significant amount is known about the types of communities and species that suffer secondary extinctions, little is known about the consequences of secondary extinctions for biodiversity. Here we examine the effect of these secondary extinctions on trophic diversity, the range of trophic roles played by the species in a community. Our analyses of natural and model food webs show that secondary extinctions cause loss of trophic diversity greater than that expected from chance, a result that is robust to variation in food web structure, distribution of interactions strengths, functional response, and adaptive foraging. Greater than expected loss of trophic diversity occurs because more trophically unique species are more vulnerable to secondary extinction. This is not a straightforward consequence of these species having few links with others but is a complex function of how direct and indirect interactions affect species persistence. A positive correlation between a species' extinction probability and the importance of its loss defines high-risk species and should make their conservation a priority.     

Parasites reduce food web robustness because they are sensitive to secondary extinction as illustrated by an invasive estuarine snail

A robust food web is one in which few secondary extinctions occur after removing species. We investigated how parasites affected the robustness of the Carpinteria Salt Marsh food web by conducting random species removals and a hypothetical, but plausible, species invasion. Parasites were much more likely than free-living species to suffer secondary extinctions following the removal of a free-living species from the food web. For this reason, the food web was less robust with the inclusion of parasites. Removal of the horn snail, Cerithidea californica, resulted in a disproportionate number of secondary parasite extinctions. The exotic Japanese mud snail, Batillaria attramentaria, is the ecological analogue of the native California horn snail and can completely replace it following invasion. Owing to the similarities between the two snail species, the invasion had no effect on predator–prey interactions. However, because the native snail is host for 17 host-specific parasites, and the invader is host to only one, comparison of a food web that includes parasites showed significant effects of invasion on the native community. The hypothetical invasion also significantly reduced the connectance of the web because the loss of 17 native trematode species eliminated many links.     

The Serengeti food web: empirical quantification and analysis of topological changes under increasing human impact

1.?To address effects of land use and human overexploitation on wildlife populations, it is essential to better understand how human activities have changed species composition, diversity and functioning. Theoretical studies modelled how network properties change under human-induced, non-random species loss. However, we lack data on realistic species-loss sequences in threatened, real-world food webs to parameterize these models. 2.?Here, we present a first size-structured topological food web of one of the most pristine terrestrial ecosystems in the world, the Serengeti ecosystem (Tanzania). The food web consists of 95 grouped nodes and includes both invertebrates and vertebrates ranging from body masses between 10(-7) and 10(4) kg. 3.?We study the topological changes in this food web that result from the simulated IUCN-based species-loss sequence representing current species vulnerability to human disturbances in and around this savanna ecosystem. We then compare this realistic extinction scenario with other extinction sequences based on body size and connectance and perform an analysis of robustness of this savanna food web. 4.?We demonstrate that real-world species loss in this case starts with the biggest (mega) herbivores and top predators, causing higher predator-prey mass ratios. However, unlike theoretically modelled linear species deletion sequences, this causes poor-connected species to be lost first, while more highly connected species become lost as human impact progresses. This food web shows high robustness to decreasing body size and increasing connectance deletion sequences compared with a high sensitivity to the decreasing connectance deletion scenario. 5.?Furthermore, based on the current knowledge of the Serengeti ecosystem, we discuss how the focus on food web topology alone, disregarding nontrophic interactions, may lead to an underestimation of human impacts on wildlife communities, with the number of trophic links affected by a factor of two. 6.?This study underlines the importance of integrative efforts between the development of food web theory and basic field work approaches in the quantification of the structure of interaction networks to sustain natural ecosystems in a changing world.     

Food web structure in Mediterranean streams: exploring stabilizing forces in these ecosystems

We constructed the food webs of six Mediterranean streams in order to determine ecological generalities derived from analysis of their structure and to explore stabilizing forces within these ecosystems. Fish, macroinvertebrates, primary producers and detritus are the components of the studied food webs. Analysis focused on a suite of food web properties that describe species’ trophic habits, linkage complexity and food chains. A great structural similarity was found in analyzed food webs; we therefore suggest average values for the structural properties of Mediterranean stream food webs. Percentage of omnivorous species was positively correlated with connectance, and there was a predominance of intermediate trophic level species that had established simple links with detritus. In short, our results suggest that omnivory and the weak interactions of detritivores have a stabilizing role in these food webs.     

Species loss and the structure and functioning of multitrophic aquatic systems

Experiments and theory in single trophic level systems dominate biodiversity and ecosystem functioning research and recent debates. All natural ecosystems contain communities with multiple trophic levels, however, and this can have important effects on ecosystem structure and functioning. Furthermore, many experiments compare assembled communities, rather than examining loss of species directly. We identify three questions around which to organise an investigation of how species loss affects the structure and functioning of multitrophic systems. 1) What is the distribution of species richness among trophic levels; 2) from which trophic levels are species most often lost; and 3) does loss of species from different trophic levels influence ecosystem functioning differently? Our analyses show that: 1) Relatively few high‐quality data are available concerning the distribution of species richness among trophic levels. A new data‐set provides evidence of a decrease in species richness as trophic height increases. 2) Multiple lines of evidence indicate that species are lost from higher trophic levels more frequently than lower trophic levels. 3) A theoretical model suggests that both the structure of food webs (occurrence of omnivory and the distribution of species richness among trophic levels) and the trophic level from which species are lost determines the impact of species loss on ecosystem functioning, which can even vary in the sign of the effect. These results indicate that, at least for aquatic systems, models of single trophic level ecosystems are insufficient for understanding the functional consequences of extinctions. Knowledge is required of food web structure, which species are likely to be lost, and also whether cascading extinctions will occur.     

Unravelling the complex structure of forest soil food webs: higher omnivory and more trophic levels

Food web topologies depict the community structure as distributions of feeding interactions across populations. Although the soil ecosystem provides important functions for aboveground ecosystems, data on complex soil food webs is notoriously scarce, most likely due to the difficulty of sampling and characterizing the system. To fill this gap we assembled the complex food webs of 48 forest soil communities. The food webs comprise 89 to 168 taxa and 729 to 3344 feeding interactions. The feeding links were established by combining several molecular methods (stable isotope, fatty acid and molecular gut content analyses) with feeding trials and literature data. First, we addressed whether soil food webs (n = 48) differ significantly from those of other ecosystem types (aquatic and terrestrial aboveground, n = 77) by comparing 22 food web parameters. We found that our soil food webs are characterized by many omnivorous and cannibalistic species, more trophic chains and intraguild‐predation motifs than other food webs and high average and maximum trophic levels. Despite this, we also found that soil food webs have a similar connectance as other ecosystems, but interestingly a higher link density and clustering coefficient. These differences in network structure to other ecosystem types may be a result of ecosystem specific constraints on hunting and feeding characteristics of the species that emerge as network parameters at the food‐web level. In a second analysis of land‐use effects, we found significant but only small differences of soil food web structure between different beech and coniferous forest types, which may be explained by generally strong selection effects of the soil that are independent of human land use. Overall, our study has unravelled some systematic structures of soil food‐webs, which extends our mechanistic understanding how environmental characteristics of the soil ecosystem determine patterns at the community level.     

Estimating morphologically determined connectance and structure for food webs of freshwater invertebrates

1. Connectance is a parameter of central importance in determining food-web structure, but the processes determining its value remain unclear. In evaluating possible explanations it is useful to know what patterns, and values, of connectance occur in food webs assembled at random from a set of species in a regional species pool; i.e. where the number of links is determined by the morphological features of the species present, not by the immediate effects of energetics or stability on the particular web. 2. This study examines, by means of laboratory experiments, the occurrence of potential feeding interactions among a set of freshwater invertebrate species randomly selected from different freshwater sites in a geographical region. The results from pairwise feeding trials are used to construct two ‘theoretical’ food webs, in which the patterns and values of connectance are examined. 3. Analyses of these webs indicate that their structure is consistent with the observed values in previously documented ‘real’ webs. Directed connectance values of 0.12–0.16 (or less) suggest that the assembled webs are no more connected than many freshwater webs from natural systems. The number of links per species increases curvilinearly with the number of species, during web assembly, consistent with recent hypotheses. 4. These results also indicate that quantifying, and understanding the determinants of, trophic generalism or specialism does have implications for understanding how connectance is constrained in real webs. Freshwater invertebrates seem to be relatively generalist, and freshwater food webs perhaps correspondingly highly connected. Such arguments have implications for interpreting other aspects of food-web structure in these systems, and for parameterizing models that are based on connectance.     

Relative ascendency predicts food web robustness

Threats to ecosystems globally from anthropogenic disturbance and climate change requires us to urgently identify the most sensitive biological communities to ensure they are effectively preserved. It is for this reason that understanding and predicting food web stability has been topical within ecology. Food web stability is a multi-faceted concept that represents the ability of a food web to maintain its integrity following disturbance, it includes resistance, resilience and fragility. In this study, we examine the ability of four food web metrics to predict the fragility to random species extinctions in 120 qualitative food webs. We show that three information-based indices out performed food web connectance in predicting fragility, with relative ascendency having the strongest relationship. Relative ascendency was a much stronger predictor of fragility than MacArthur’s stability metric, Average Mutual Information and connectance as it accounted for both the distribution and number of links between species. We also find that most qualitative food webs persist around a central tendency of relative ascendency.