Inferring species interactions in ecological communities: a comparison of methods at different levels of complexity

Natural communities commonly contain many different species and functional groups, and multiple types of species interactions act simultaneously, such as competition, predation, commensalism or mutualism. However, experimental and theoretical investigations have generally been limited by focusing on one type of interaction at a time or by a lack of a common methodological and conceptual approach to measure species interactions. 2. We compared four methods to measure and express species interactions. These approaches are, with increasing degree of model complexity, an extinction-based model, a relative yield model and two generalized Lotka-Volterra (LV) models. All four approaches have been individually applied in different fields of community ecology, but rarely integrated. We provide an overview of the definitions, assumptions and data needed for the specific methods and apply them to empirical data by experimentally deriving the interaction matrices among 11 protist and rotifer species, belonging to three functional groups. Furthermore, we compare their advantages and limitations to predict multispecies community dynamics and ecosystem functioning. 3. The relative yield method is, in terms of final biomass production, the best method in predicting the 11-species community dynamics from the pairwise competition experiments. The LV model, which is considering equilibrium among the species, suffers from experimental constraints given the strict equilibrium assumption, and this may be rarely satisfied in ecological communities. 4. We show how simulations of a LV stochastic community model, derived from an empirical interaction matrix, can be used to predict multispecies community dynamics across multiple functional groups. 5.Our work unites available tools to measure species interactions under one framework. This improves our ability to make management-oriented predictions of species coexistence/extinction and to compare ecosystem processes across study systems.

Resistance of plant communities to invasive species

Invasive species are nowadays considered as one of the most important threat to biodiversity. By displacing native species, modifying ecosystem functioning and causing substantial losses to agricultural production, they represent a menace to natural and managed ecosystems. Although ecology of invasions has become an important research topic since the last decades, the mechanisms that determine why a given species may invade a given ecosystem and why some biomes are less resistant to invasion are still not clarified. Ecology of invasions is divided into 2 main topics: invasiveness and invasibility. While invasiveness refers to species ability to invade a community, invasibility focuses on the resistance of a community to invasion. Invasiveness may be the result of ecological processes, such as release from biotic constraints or human alteration of the environment (disturbance, stress…) or the consequence of evolutionary processes, such as hybridization or polyploidization that may increase genetic variation and therefore, enhance niche breadth. Invasibility has been said to be influenced by disturbance and biotic factors such as community diversity, dominant species identity, biotic interactions or community compositional stability. The invasion success is the consequence of the interaction between species invasiveness and community invasibility. Most studies in ecological invasions have focused on either invasiveness or invasibility, but hardly both together. By working at the same time and in the same conditions with native and introduced genotypes and by comparing their ecological performances, this thesis aims at a better understanding of both invasiveness and invasibility mechanisms. Two worldwide invasive species, Centaurea maculosa and Senecio inaequidens, were used in several experiments (pot, microcosm, field) to disentangle the importance of invasiveness and community invasibility in their invasion success. Both species encountered polyploidisation in their native range, leading to the presence of diploid and tetraploid populations, whereas only tetraploid populations have been found in the introduced range. Using native diploid, native tetraploid and introduced tetraploid genotypes of the two model-species, allows assessing the effects of genetic variation (diploid vs. tetraploid genotypes) and environmental variation (genotype from native vs. introduced range) on species phenotypic traits variations and consequently on fitness variation and invasiveness. In the community context, studying response of different genotypes to experimental factors and community change gives information on the interaction between invasiveness and invasibility. Plants were grown in pot, in field or in artificially built communities where (i) the management treatment, (ii) the community diversity and (iii) the spatial organisation of resident species were manipulated. In addition, (iv) community species composition, (v) community competitive ability and (vi) compositional and functional stability of the community were monitored. According to the experiments, data were gathered on survival, morphological traits (vegetative height, lateral spread, shoot and root biomass), leaf traits (specific leaf area, leaf dry matter content) and reproductive traits (probability of flowering, capitulum production) of the genotypes (native diploid, native tetraploid and introduced tetraploid) of both model species. Through the use of statistical models and multivariate approaches, the effects of management and biotic factors on survival, growth and reproduction of native and introduced genotypes of these two worldwide invasive species were assessed. Invasion strategy of the two model species was investigated through a growth experiment in optimal conditions in a pot experiment. High investment in seed production could explain invasive success of S. inaequidens through high propagule pressure, whereas C. maculosa's strategy seemed to be oriented towards interactions with belowground communities, as shown by the shift in rhizosphere bacterial communities between genotypes. For both species, polyploidisation in the native range could be linked to a specialisation towards higher competitive ability, which could have allowed the first step of invasion. Introduction in the new range could be related to a loss of specialisation through selection of traits allowing coping with various or changing environments, improving successful spread. Survival of both model species was highly affected by community spatial pattern, management and neighbouring competition. Responses of growth and reproductive output of native and introduced genotypes to management, community spatial pattern and community diversity were species-specific. Growth and reproductive output of both genotypes of S. inaequidens were affected by experimental factors whereas introduced genotypes of C. maculosa were less affected than native ones. Comparison of response of native and introduced genotypes to experimental factors allowed defining two strategies of invasion based on phenotypic plasticity. Centaurea maculosa was able to maintain fitness in all kinds of environments, either favourable or stressful ("Jack-of-all-trades" invader). Senecio inaequidens was able to deal with all kinds of environments and was also able to increase its fitness in favourable conditions ("Jack-and-Master" invader). The combination of the effects of management, community spatial pattern and community diversity on species genotypes allowed defining, for both invasive species, two invasion phases which were impacted by different factors. The introduction phase corresponds to the survival of seedlings and their ability to deal with neighbouring competition. If seedlings manage to survive despite neighbouring competition, they grow and reproduced in order to spread, which corresponds to the establishment phase. In terms of management perspectives, regular mowing or use of highly covering species could limit invasive success of S. inaequidens and C. maculosa respectively. The synthesis of all the experiments conducted in this thesis with C. maculosa and S. inaequidens highlights (1) the importance of polyploidisation in the invasion process as well as (2) the species-specific invasion strategies and consequently (3) the species-specific response of invasive species to abiotic and biotic factors. It also emphasizes on (4) the temporal evolution of the interaction between invasiveness and invasibility since the community factors that affected invasive species fitness changed according to the invasion stage (introduction vs. establishment phase) of the invader. The provided insights into the importance of the interaction between species invasiveness and community invasibility will contribute to improve knowledge in ecology of invasions, in addition to supply some clues for management efficiency.

EPFL2009

Theoretical and Experimental Studies of Dendritic Metacommunities

Francesco Carrara

The present thesis deals with the understanding of the origins and the mechanisms of maintenance of biodiversity in natural landscapes, in particular by identifying key processes that define large-scale patterns of abundance and diversity. Biological communities often occur in spatially structured habitats where connectivity directly affects dispersal and metacommunity processes. Recent theoretical work suggests that dispersal constrained by the connectivity of specific habitat structures affects diversity patterns and species interactions. This is particularly relevant in dendritic networks epitomized by fluvial ecological corridors. This thesis addresses whether connectivity alone can explain observed features of biodiversity and selectively promote different components of community composition in river-like landscapes, such as local species richness or the among-community similarity. The relevance of this thesis lies in the major ecological challenges posed by the topic, and its fundamental importance to conservation biology. The studies pursued herein are also deemed relevant because of the influence of the spatial connectivity and dispersal on population dynamics and of the relevance of biodiversity to ecosystem functioning. Mechanisms of species coexistence were investigated with a blend of theoretical tools (broadly related to statistical mechanics and the theory of stochastic processes) and experimental work using laboratory microbial communities. The research tools ranged from aspects of modern coexistence theory in a local perspective to the recent concept of the metacommunity in spatial ecology, within a unified framework. The study of biodiversity in riverine ecosystems guided by observational data has been addressed by combining theoretical metacommunity models with laboratory experiments. The results are diverse. First, they show experimentally that connectivity per se shapes key components of biodiversity in metacommunities. Local dispersal in isotropic lattice landscapes homogenizes local species richness and leads to pronounced spatial persistence. By contrast, dispersal along dendritic landscapes leads to higher variability in local diversity and among-community composition. Although headwaters exhibit relatively lower species richness, they are crucial for the maintenance of regional biodiversity. By suitably arranging patch sizes within river-like networks the effect of local habitat capacity (i.e., the patch size) and dendritic connectivity on biodiversity can be experimentally disentangled in aquatic microcosm metacommunities. It is shown in this thesis that species coexistence and community assembly depend on intricate, non-trivial interactions of local community capacity and network positioning. Furthermore, an interaction of spatial arrangement of habitat capacity and dispersal along river-like networks also affects a key ecosystem descriptor, namely regional evenness. High regional evenness in community composition is found only in landscapes preserving geomorphological scaling properties of patch sizes. In riverine environments some of the rarer species sustained regionally more abundant populations and were better able to track their own niche requirements compared to landscapes with homogeneous patch size or landscapes with spatially uncorrelated patch size. All the experimental results were supported and extended by a theoretical analysis where the above mechanisms have been generalized. This thesis provides the first direct experimental evidence that spatially constrained dendritic connectivity is a key factor for community composition and population persistence in riverine landscapes. As such, this thesis assesses a longstanding issue in spatial community ecology. It offers unique insights into the ecological forces structuring natural communities in a key ecosytem, and demonstrates principles that can be further tested in theoretical metacommunity models possibly to be extended to real riverine ecosystems. Taken together, the analyses show how the structure of ecological networks interacts with the spatial environmental matrix in determining biodiversity patterns and the functioning of biological communities. The analyses also suggest that altering the natural linkage between dendritic connectivity and patch size strongly affects community properties at multiple scales. The first part of this thesis (chapters 2 and 3) addresses key aspects of biodiversity-ecosystem functioning research where the combination of theory-guided experiments and theoretical investigations shows how a stochastic implementation of population dynamics proves fundamental for key community properties such as species persistence and community stability. The diversity-productivity and diversity-stability relationships are explored. Both experimental findings and the results of a stochastic model fitted to the experimental interaction matrix, suggest the emergence of strong stabilizing forces when species from different functional groups interact in the same environment, increasing species coexistence and community biomass production. The last part (chapter 6) provides a synthesis of this thesis work, in that it aims at unifying aspects from niche-theory, usually adopted in spatially implicit models, with those characteristic of a spatially explicit context from a typical real-life mountainous regions. It is dedicated to the possible explanation for a macroecological pattern routinely observed from organisms in different domains of life, that is, the mid-elevational peak in local species richness. Guided by empirical observations on diversity of macroinvertebrates in Swiss river basins, a theoretical ansatz is provided which is deemed to capture the essential geomorphological drivers and controls relating species-fitness to altitude. A set of overarching conclusions and perspectives for future research are discussed in the concluding chapter.

EPFL2013

Ecohydrological Footprints

Samir Suweis

Ecohydrological footprints are defined as the response of ecosystem functions or services to changes in their hydrologic drivers. In this thesis, several diverse footprints are addressed: noise-driven effects on storage-discharge relations and catchment streamflow distributions, that are important drivers of biodiversity; soil salinization and its ecohydrological implications; topological effects of the ecological interaction networks on living communities (e.g. on their species persistence); and form and function of the global virtual water trade network. The coherence of the conceptual framework is provided by the study of drivers and controls of ecohydrological variability using methodological approaches based on statistical mechanics. In fact, this thesis work outlines a significant portion of environmental statistical mechanics, an overarching discipline that is emerging in recent years, which applies mathematical tools from statistical mechanics to model several ecohydrological processes. The proposed relevance of this thesis lies in the major effects of hydrologic drivers on ecological process. The view that emerges from current research in ecohydrology, that this thesis supports, is that there exists a definite need for an integrated understanding of ecological and hydrological processes. Because stochasticity is intrinsic to environmental and ecohydrological variability, noise plays an important and constructive role in ecohydrological processes. In this thesis, a stochastic approach is applied to analyze different ecohydrological processes, ranging from green and blue water flows in river basins (part I), ecosystem dynamics affected by the directional dispersal provided by river networks (part II) to water footprints of human society (part III).Methods range from novel exact solutions to stochastic differential equations to random graph theory applications, and imply the analysis of suitable field data. An analytical framework for quantitative analysis is laid out to tackle complex problems and to estimate the effects of environmental change on the interaction of the hydrologic processes with the biota. The main results of this thesis are: i) the achievement of exact solutions for the probability distribution of catchment streamflow, that takes in account stochastic fluctuations in the storage-discharge relation and for the condition of a noise induced phenomena to the streamflows regimes; ii) the stationary solutions of soil salinity under stochastic hydrologic forcing; iii) a novel solution of the Ito-Stratonovich problem in multiplicative Poisson processes; iv) the proper framework for species' persistence time distributions, as a function of topological constraints on the ecosystem, and its connection with other important macroecological laws. A related length-bias sampling problem is also solved. v) A statistical analysis of the global virtual trade network and a semi-analytical model that is able to describe most of the observed properties.

EPFL2011