Food webs, defined as "who eats whom" in nature, have become a central topic within community ecology and thus they have been used to understand general ecological patterns such as biodiversity and species interactions as well as material and nutrient flows within ecosystems. In South America the knowledge of the taxonomy and distribution of freshwater invertebrates is incomplete and fragmented. Previous studies have focused on specific taxonomic groups and some countries such as Brazil and Argentina. In contrast, there have been many aquatic food webs published for UK freshwater systems with high levels of taxonomic resolution. This thesis aims to examine food web patterns in two geographically separated systems. The effects of systematic taxonomic aggregation on food web properties were examined and the relationship between consumer and prey body size revisited. A total of 24 food webs were examined in Chilean and Welsh streams, where 6128 invertebrate guts were examined to establish feeding interactions. These Chilean and Welsh food webs are amongst the largest, most complete and fully resolved. In both systems there was a high proportion of basal species, combined with low proportions of top and intermediate species. Significant differences were detected in most food web properties, in comparison to previous studies, where basal species were aggregated to coarser categories. No significant relationship between the body size of the consumers and their prey was found in either Chilean or Welsh streams. These results differ substantially from published data, and we attribute these differences to the greater taxonomic resolution particularly on the basal resources.

A major goal of ecology is to discover how the dynamics and structure of multi-trophic ecological communities are related. It is difficult to understand links between dynamics and structure because mathematical models of the dynamics of systems of realistic complexity have a large number of unmeasured parameters, and whole-community data are limited and typically comprise only a snapshot or time-averaged picture. The resulting 'plague of parameters' means most studies of multi-species population dynamics have been very theoretical. Dynamical models parameterised using physiological allometries suggest a solution to the plague of parameters. These models are a synthesis of allometric scaling and Lotka-Volterra style dynamical models (Yodzis & Innes, 1992): model parameters are computed from empirically-observed inter-specific power-law relationships between physiological rates and body masses. This approach avoids the need to derive species- or population-specific parameters, sacrificing some accuracy for generality and making it possible to investigate the dynamics of complex communities. These models have been used in a large number of theoretical studies that have drawn conclusions on a wide range of topics. Despite their increasing use, this class of dynamical models are rarely tested against empirical data. This PhD examined this modelling approach and some of its assumptions. Outcomes of this work are 1) publication of a new dataset of field metabolic rate data of individual birds and mammals together with an analysis of this data using linear mixed-effects models, leading to a better understanding of one of the model's principal assumptions, 2) an open-source R package for analysing and visualising empirical food-web data, 3) an open-source R package for simulating community dynamics using the model of interest and 4) validation of the model's ability to recreate static patterns seen in empirical community data.

Food web structure is of underlying importance to ecological functions and processes. Whilst it is understood that a range of biotic and abiotic factors affect structure, relatively little is known of the role of biodiversity per se in structuring food webs. In this thesis I utilise novel multi-dimensional estimates of food web structure based on stable isotope ratios of nitrogen (δ15N) and carbon (δ13C) to quantify structural responses to changing community diversity. I additionally investigate methodological aspects of sample preparation and stable isotope quantifications of food chains. Using an arthropod prey-predator system, in chapter 2 I demonstrate that tissue selection and lipid extraction are important methodological procedures for deriving accurate δ15N and δ13C signatures. In chapter 3 I test the utility of δ15N to quantify food chain length, and δ13C to trace primary energy sources through to end consumers. Bayesian resampling of variance in sample means for plant and arthropod food chains produces robust isotopic estimates that match known food chain length well despite some error variance, and estimates of δ13C-range that trace trophic transfers. Chapter 4 represents a change in system from lab to field as I determine δ15N and δ13C signatures for plant and invertebrate species within three grassland communities representing a gradient of biodiversity. Quantifications of community bivariate isotopic space using isotopic metrics revealed that greater taxonomic richness increased both diversity of resource space exploited and overlap in resource space. These results therefore suggest that loss of diversity affected structure through altering relative patterns of niche partitioning in resource exploitation amongst community members. In chapter 5, I additionally find evidence that grassland management mediated changes in food web compartmental structure that were associated with differences in generalist invertebrate predator feeding habits. Taken together, these findings develop and demonstrate the utility of isotopic approaches to quantifying food web structure, and provide evidence of important mechanisms by which biodiversity affects food web structure. I conclude that the preservation of natural food web structure and trophic dynamics are further reasons for halting loss of biodiversity.