L'installation de récifs artificiels est une solution classiquement proposée pour soutenir la petite pêche artisanale. C'est dans ce but qu'a été déployé dans la baie de Marseille le plus grand ensemble de récifs artificiels de Méditerranée. Ce système offre une opportunité de comprendre le fonctionnement trophique de ces structures par (1) la caractérisation des sources de matière organique (MO) (2) la détermination de leur utilisation par les consommateurs et (3) leur devenir au sein des réseaux trophiques. Ces questions ont été appréhendées à l'aide des isotopes stables du C et du N, de la caractérisation biochimique des sources de MO et de l'analyse des contenus stomacaux des poissons.Les sources de MO présentent des différences isotopiques et biochimiques, reflet de leurs fonctionnements différents. La MO particulaire en suspension est un pool variable influencé par les apports allochtones et soumis aux forçages climatiques. Elle est la source principale de MO sur les récifs artificiels. La production primaire benthique est très hétérogène et de qualité nutritionnelle faible. Elle contribue aux réseaux trophiques des récifs par le biais des détritus qu'elle produit. Enfin, la MOS est un pool détritique de faible qualité nutritionnelle. Les récifs artificiels ne modifient pas la structure et le fonctionnement des communautés naturelles de poissons et leur offrent des ressources alimentaires variées et importantes. L'ensemble des résultats acquis, via les isotopes stables et les contenus stomacaux, montrent que les récifs artificiels sont des producteurs de biomasse à partir des productions primaires locales. / Artificial reefs are a classical tool used to sustain small scale fisheries. In Marseille's Bay, the largest Mediterranean artificial reef system was deployed for this purpose. This system is a valuable opportunity to better understand the trophic functioning of artificial reefs, by (1) characterizing the organic matter (OM) sources, (2) determining how they are used by low trophic level consumers and (3) following their fate in the trophic networks. These issues were assessed by three approaches:C and N stable isotope ratios, biochemical composition of OM sources and fish stomach content.OM sources display isotopic and biochemical differences reflecting their functioning. Suspended POM is variable, under the influence of allochtonous inputs and controlled by climatic forcing. POM represents the main OM source of the artificial reef food webs. The benthic primary production has heterogeneous C and N stable isotope ratios and exhibits mainly high insoluble carbohydrates contents. Benthic primary production is integrated into trophic networks mainly in the form of detritus. Eventually the sediment organic matter is a detrital pool, characterized by a low variability and a poor nutritional quality.Artificial reefs do modify the organization and the functioning of natural fish communities, and provide diversified and important food resources. Stable isotope and stomach content analyses confirm the consumption of artificial reef invertebrates by small carnivorous fishes, preyed themselves by piscivorous predators. All these results confirm that artificial reefs can efficiently increase fish biomass by local production based on phytoplanktonic and local benthic OM sources.

Most climate change predictions focus on the response of individual species to changing local conditions and ignore species interactions, largely due to the lack of a sound theoretical foundation for how interactions are expected to change with climate and how to incorporate them into climate change models. Much of the variability in species interaction strengths may be governed by fundamental constraints on physiological rates, possibly providing a framework for including species interactions into climate change models. Metabolic rates, ingestion rates and many other physiological rates are relatively predictable from body size and body temperature due to constraints imposed by the physical and chemical laws that govern fluid dynamics and the kinetics of biochemical reaction times. My dissertation assesses the usefulness of this framework by exploring the community-level consequences of physiological constraints. In Chapter 2, I incorporated temperature and body size scaling into the biological rate parameters of a series of realistically structured trophic network models. The relative magnitude of the temperature scaling parameters affecting consumer energetic costs (metabolic rates) and energetic gains (ingestion rates) determined how consumer energetic efficiency changed with temperature. I systematically changed consumer energetic efficiency and examined the sensitivity of network stability and species persistence to various temperatures. I found that a species' probability of extinction depended primarily on the effects of organismal physiology (body size and energetic efficiency with respect to temperature) and secondarily on the effects of local food web structure (trophic level and consumer generality). This suggests that physiology is highly influential on the structure and dynamics of ecological communities. If consumer energetic efficiency declined as temperature increased, that is, species did best at lower temperatures, then the simulated networks had greater stability at lower temperatures. The opposite scenario resulted in greater stability at higher temperatures. Thus, much of the community-level response depends on what species energetic efficiencies at the organismal-level really are, which formed the research question for Chapter 3: How does consumer energetic efficiency change with temperature? Existing evidence is scarce but suggestive of decreasing consumer energetic efficiency with increasing temperature. I tested this hypothesis on seven rocky intertidal invertebrate species by measuring the relative temperature scaling of their metabolic and ingestion rates as well as consumer interaction strength under lab conditions. Energetic efficiencies of these rocky intertidal invertebrates declined and species interaction strengths tended to increase with temperature. Thus, in the rocky intertidal, the mechanistic effect of temperature would be to lower community stability at higher temperatures. Chapter 4 tests if the mechanistic effects of temperature on ingestion rates and species interaction strengths seen in the lab are apparent under field conditions. Bruce Menge and I related bio-mimetic estimates of body temperatures to estimates of per capita mussel ingestion rates and species interaction strengths by the ochre sea star Pisaster ochraceus, a keystone predator of the rocky intertidal. We found a strong, positive effect of body temperature on both per capita ingestion rates and interaction strengths. However, the effects of season and the unique way in which P. ochraceus regulates body temperatures were also apparent, leaving room for adaptation and acclimation to partially compensate for the mechanistic constraint of body temperature. Community structure of the rocky intertidal is associated with environmental forcing due to upwelling, which delivers cold, nutrient rich water to the nearshore environment. As upwelling is driven by large-scale atmospheric pressure gradients, climate change has the potential to affect a wide range of significant ecological processes through changes in water temperature. In Chapter 5, my coauthors and I identified long-term trends in the phenology of upwelling events that are consistent with climate change predictions: upwelling events are becoming stronger and longer. As expected, longer upwelling events were related to lower average water temperatures in the rocky intertidal. Furthermore, recruitment rates of barnacles and mussels were associated with the phenology of upwelling events. Thus climate change is altering the mode and the tempo of environmental forcing in nearshore ecosystems, with ramifications for community structure and function. Ongoing, long-term changes in environmental forcing in rocky intertidal ecosystems provide an opportunity to understand how temperature shapes community structure and the ramifications of climate change. My dissertation research demonstrates that the effect of temperature on organismal performance is an important force structuring ecological communities and has potential as a tractable framework for predicting the community level effects of climate change. / Graduation date: 2013 / Access restricted to the OSU Community, at author's request, from Nov. 20, 2012 - Nov. 20, 2014