Final Report Summary - ECOFUN (Analysis of biodiversity changes on structural and functional properties of marine ecosystems under cumulative human stressors)
This is the third and last scientific review report that provides an update on the scientific progress of the ECOFUN project during the last year of its development (January - December 2011). It also gives a summary of the resources used, and it includes an updated version of the fellow curriculum vitae with new publications and summaries of main results, updated list of contribution to congresses and other deliverables from the ECOFUN project.
a) Scientific overview of the project
The objective of my research developed during the Marie Curie International Outgoing Fellowship (IOF) was to quantify how the structure of marine food-webs changes with various human stressors and develop an understanding of how these changes impact ecosystem functioning. To accomplish this objective I used an interdisciplinary approach applying meta-analysis of historical and recent data on species loss and ecosystem changes, food-web modelling, and laboratory experiments.
During the third and last year of the ECOFUN project, I finished the synthesis of ecological data on biodiversity changes in marine ecosystems, and the modelling of the consequences of these biodiversity changes on food-web structure and functioning, and continued the analysis of aquatic microcosm experiments to measure ecosystem functioning changes due to biodiversity changes.
A detailed description of the progress made during the third year is provided below:
(1) Synthesis of ecological data on biodiversity changes in marine ecosystems
I compiled a database to synthesise available data on long-term changes in marine biodiversity under gradients of different human impacts for some ecosystems (mainly some areas of the Mediterranean Sea and the Atlantic Canada). This database also records observed / measured consequences of these changes from individual species and populations to ecosystem levels in terms of structure or topology of the food web (e.g. absence / presence, trophic level, connectance) and functioning (changes in biomass, production, and trophic flows allocation and energy transfer efficiency, such as cascading effects, proliferations of non-abundant species and increase in detritus). The goal of this task was to document historical changes in ecosystems due to changes in marine resources.
In this context, the IndiSeas project, on which I have been collaborating since 2007, produced a database (see http//:www.indiseas.org online) that now is available to the scientific community and the general public. This database summarises now the state of 34 exploited marine ecosystems worldwide regarding fishing impacts and results document and analyze ecosystem changes in exploited ecosystems due to fishing activities going back to the mid 1960s. An updated on scientific advances of INDISEAS project have been recently submitted to reviews in fish biology and fisheries (Shin et al., submitted). Moreover, during November 2011, I attended a new meeting of INDISEAS where the phase of the project continued. The new phase will last two more years and it aims to:
1) extend the application of INDISEAS original ecological indicators to other ecosystems;
2) complement the list of ecological indicators to include biodiversity and conservation-based indicators, environmental indicators and socio-economic indicators; and
3) investigate ways to provide reference levels for indicators, as well as create composite indicators.
Within this second phase of the project, or INDISEASII, I am co-chairing the task group of biodiversity and conservation-based indicators with Dr. Lynne J. Shannon, from the University of Cape Town, South Africa. The participation on INDISEASII allowed me to collect new data from different ecosystems and perform more complex analysis with different human stressors.
In addition, during 2011 I continued the collaboration with Dr Denise Breitburg and colleagues from the marine ecology lab at the Smithsonian Environmental Research Centre in Maryland (in United States of America) under the context of the project 'Linking nutrients, hypoxia, fisheries and fishes'. This project aims at better understanding the effects of nutrient enrichment, hypoxia, and the interaction between these stressors with fisheries removals on fish biomass and the composition in estuaries and semi-enclosed seas ('fish' being used loosely here to include crabs, shrimp, etc.). In this project, we are constructing a database with data from nutrients and hypoxia events, fisheries and ecosystem traits and a series of publications are being prepared.
Building from the data that I have been collecting, during 2011 I also collaborated in a study on ecological history of the Adriatic Sea (Lotze, Coll and Dunne, 2011) and on recovery examples in marine ecosystems using historical data and modelling results (Lotze et al., 2011, Coll and Lotze, in preparation). Moreover, continuing the work I did under the umbrella of the Census of Marine Life in 2009 - 2010, a new study was published (Coll et al., 2011).
In Lotze, Coll and Dunne (2011), we used a multidisciplinary approach combining our paleontological, archeological, historical, fisheries, and ecological data to reconstruct past changes in marine populations, habitats, and water quality in the Adriatic Sea. Then, we constructed binary food webs for different historical periods to analyze possible changes in food-web structure and functioning over time. Our results indicated that human activities have influenced marine resource abundance since at least Roman times and have accelerated in the nineteenth and twentieth centuries. Today, 98 % of traditional marine resources are depleted to less than 50 % of former abundance, with large (>1 m) predators and consumers being most affected. With 37 % of investigated species rare and 11 % extirpated, diversity had shifted towards smaller, lower trophic level species, further aggravated by more than 40 species invasions. Species providing habitat and filter functions had been reduced by 75 %, contributing to the degradation of water quality and increased eutrophication. Increased exploitation and functional extinctions had altered and simplified food-web structure over time, especially by changing the proportions of top predators, intermediate consumers, and basal species. Moreover, simulations of species losses indicated that today's ecosystems may be less robust to species extinctions than in the past. Our results illustrated the long-term and far-reaching consequences human activities can have on marine food webs and ecosystems.
In Lotze et al. (2011), we reviewed the growing research on marine recoveries to reveal how common recovery is, its magnitude, timescale and major drivers. Results showed that overall, 10 - 50 % of depleted populations and ecosystems show some recovery, but rarely to former levels of abundance. In addition, recovery can take many decades for long-lived species and complex ecosystems. Major drivers of recovery include the reduction of human impacts, especially exploitation, habitat loss and pollution, combined with favourable life-history and environmental conditions. Awareness, legal protection and enforcement of management plans are also crucial. We concluded that learning from historical recovery successes and failures is essential for implementing realistic conservation goals and promising management strategies.
In Coll et al. (2011), we aimed to identify the main areas where the interaction between marine biodiversity and threats is more pronounced and to assess their spatial overlap with current marine protected areas in the Mediterranean. We first identified areas of high biodiversity of marine mammals, marine turtles, seabirds, fishes and commercial or well-documented invertebrates. We mapped potential areas of high threat where multiple threats are occurring simultaneously. Finally we quantified the areas of conservation concern for biodiversity by looking at the spatial overlap between high biodiversity and high cumulative threats, and we assessed the overlap with protected areas. Our results showed that areas with high marine biodiversity in the Mediterranean Sea are mainly located along the central and north shores, with lower values in the south-eastern regions. Areas of potential high cumulative threats are widespread in both the western and eastern basins, with fewer areas located in the south-eastern region. The interaction between areas of high biodiversity and threats for invertebrates, fishes and large animals in general (including large fishes, marine mammals, marine turtles and seabirds) is concentrated in the coastal areas of Spain, Gulf of Lions, north-eastern Ligurian Sea, Adriatic Sea, Aegean Sea, south-eastern Turkey and regions surrounding the Nile Delta and north-west African coasts. Areas of concern are larger for marine mammal and seabird species. These areas may represent good candidates for further research, management and protection activities, since there is only a maximum 2 % overlap between existing marine protected areas (which cover 5 % of the Mediterranean Sea) and our predicted areas of conservation concern for biodiversity.
(2) Modelling the consequences of biodiversity changes on food-web structure and functioning
The available datasets collected on biodiversity changes were used in ecological modelling techniques to analyse the consequences of biodiversity changes on food-web structure and dynamics. To do so, I used different food-web modelling techniques: stochastic static and bioenergetic dynamic modelling, and mass-balance and trophodynamic modelling. i have also incorporated the use of geographic information systems (GIS, mainly learning ArcGIS from ESRI software).
The ecological modelling approach included the analysis of the consequences of diversity changes on food webs in various Mediterranean seas (Coll and Libralato, 2011; Navarro et al., 2011; Lotze, Coll and Dunne, 2011) and on seagrass ecosystems in the Northern Atlantic (Coll et al., 2011b; Schmidt et al., 2011). Moreover, I am working on a last manuscript presenting results on the recovery of marine ecosystems (Coll and Lotze, in preparation).
In Coll and Libralato (2011), we present a meta-analysis of Ecopath with Ecosim applications in the Mediterranean Sea. Ecological modelling tools are worldwide applied to support the ecosystem-based approach of marine resources. Numerous applications of this approach have been attempted in the last decades in the Mediterranean Sea, mainly using Ecopath with Ecosim tool. These models are used to analyse a variety of complex environmental problems. Many applications analyse the ecosystem impacts of fishing and have been used to assess management options, such as the reduction of fishing effort or the establishment of protected areas. Other studies deal with the accumulation of pollution through the food web, the impact of aquaculture, or the ecosystem implications of climate change. These applications contributed to the scientific aspects of an ecosystem-based approach in the region since they integrate human activities within the ecosystem context and evaluate their impact along the marine food web, quantifying direct and indirect impacts, and including the environmental factors. These studies also gathered an important bulk of information at an ecosystem level. Therefore, as a second part of this study, we used this information to quantify main structure and functional traits of Mediterranean Sea marine ecosystems at regional scales. Results highlighted several differential traits between ecosystem types and a few between basins, which illustrate the environmental heterogeneity and complexity of the Mediterranean Sea. The analysis of key species evidenced the importance of top predators and small pelagic fish in Mediterranean ecosystems in addition to the structural role of benthos and plankton organisms. The impact of fishing was found to be of a similar intensity in the Western, Central and Eastern regions, with some differences between ecosystem types.
In Navarro, Coll et al. (2011), we examine the trophic dynamics of representative species of the South Catalan marine ecosystem (North-western Mediterranean) using the Ecopath ecosystem modelling tool and the stable isotope approach. By using the two methodologies we depicted the trophic position (trophic level and d15N values) and trophic width (omnivory index and total isotopic area) of several species of fish, cephalopods, cetaceans, seabirds and one sea turtle. Our results showed a clear correlation between the trophic levels estimated by the Ecopath model and the d15N values, which validate the trophic position of several species in the study area and indicate that both methodologies are useful to determine the trophic position of marine species in Mediterranean marine food webs. In contrast, the two estimators of trophic width (the omnivore index and the total isotopic area) were not strongly related, since the relationship between modelling and stable isotope results was evident only for some species. Further comparisons of trophic width calculated in other marine food webs may provide a better understanding of our results and validate the accuracy of both methodological approaches to calculate trophic width.
In Coll et al. (2011b), we analysed the main structural features of food webs associated with Zostera marina communities in New Brunswick, Prince Edward Island and Nova Scotia, in Atlantic Canada, and across a gradient of eutrophication. The goals of this study were to quantify changes in food-web structure across regions and eutrophication levels, to assess whether changes in food-web structure translate into changes in functioning by evaluating the robustness of food webs against simulated species loss, and to compare food-web structure in seagrass beds with those of other aquatic ecosystems worldwide. Based on data from field surveys, we constructed individual food webs for 16 study sites and cumulative food webs for regions and eutrophication levels. Network models were used to calculate 16 structural properties for each food web. Our results indicated that food-web structure was similar among low eutrophication sites across regions in Atlantic Canada. Moreover, our seagrass webs were similar to a tropical seagrass web, and significantly differed from other aquatic webs. A trend of increasing degradation was observed across the eutrophication gradient in both NB and PEI. With increasing eutrophication food webs showed more variability and less predictable patterns as well as a general structural simplification indicated by fewer trophic groups, lower maximum trophic level of the highest top predator, fewer number of trophic links connecting top to basal species, a higher fraction of herbivore and intermediate consumers, and higher number of prey species per predator. Several food-web properties showed non-linear responses to increasing eutrophication and partially differed between regions indicating site-specific attributes that modulate the responses to enrichment. More eutrophied sites are also less robust to species loss. The effect of species deletion was stronger in PEI compared to NB and NS, and may indicate a macro-ecological gradient of degradation across regions. Overall, our study suggested that eutrophication alters food-web structure and functioning of seagrass habitats, and that the spatial scale at which food-web structure is studied is a critical factor determining the results of food-web analyses.
In Schmidt, Coll et al. (2011), we used models of predator-prey feeding interactions to evaluate food-web structure and their robustness to simulated species removals. Despite disparate three-dimensional canopy structure, both habitats significantly enhanced overall abundance and diversity of associated flora and fauna. Yet significant differences occurred in the species assemblages within and between habitats that are attributed to different settlement opportunities, food availability, predation risk, and maneuverability. While eelgrass was a better nitrogen filter, rockweed maintain eight-fold greater biomass and thus 14-fold greater nitrogen and eight -fold greater carbon storage. 53 % of the food-web properties differed between the two habitats. Eelgrass communities showed higher food-web complexity, and a greater robustness to simulate species removals. Our study demonstrated that marine vegetation provided important habitat, food, shelter and stores nitrogen and carbon, yet the extent of these services depended on the foundation species and its architecture. Changes in canopy structure will have profound effects on associated food webs and ecosystem services. Thus, increasing human pressures on coastal ecosystems threaten the continued supply of essential functions and services these ecosystems provide, and the protection of marine vegetated habitats should be a management priority.
During this year 2011, I also visited the Fisheries Centre (University of British Columbia, Vancouver, Canada) for three months (September - December). There I participated in the Ecopath research and development consortium initial meeting.
(3) Application of aquatic microcosm experiments to measure ecosystem functioning changes due to biodiversity changes
In parallel, in 2011 I continue analysing the results from the aquatic microcosm experiments than were conducted in 2008 and 2009 to observe patterns of human impacts on marine food webs under laboratory conditions and to quantify changes in structure and function. Different methodologies and techniques were used, including the analysis of stable isotopes and bacterial activity. During 2011, I finished the analysis of the results of experiments performed during 2008 (preliminary trial) and 2009, where a series of microcosm experiments were developed to reproduced different levels of diversity change under climate change scenarios in laboratory conditions. The first results are now published (Coll and Hargadon, in press) and another one is being prepared (Coll et al., in preparation).
In Coll and Hargadon (in press), we explored food web changes in trophic structure and ecosystem functioning following biomass removal of top predators in representative temperate and tropical rock pool communities that contained similar assemblages of zooplankton and benthic invertebrates. We observed changes in species abundances following predator removal in both temperate and tropical communities, in line with expected inverse effects of a trophic cascade, where predation release benefits the predator's preys and competitors and impacts the preys of the latter. We also observed several changes at the community and ecosystem levels including a decrease in total abundance and mean trophic level of the community, and changes in chlorophyll-a and total dissolved particles. Our results also showed an increase in variability of both community and ecosystem processes following the removal of predators. These results illustrate how predator removal can lead to inverse trophic cascades both in structural and functioning properties, and can increase variability of ecosystem processes. Although observed patterns were consistent between tropical and temperate communities following an inverse cascade pattern, changes were more pronounced in the temperate community. Therefore, aquatic food webs may have inherent traits that condition ecosystem responses to changes in top-down trophic control and render some aquatic ecosystems especially sensitive to the removals of top predators.
(4) Meta-analysis on ecosystem functioning changes due to marine biodiversity changes
The application of a meta-analytical approach combining the results from data synthesis, modelling and laboratory experiments were used to quantify general patterns of structural and functional food-web changes across ecosystems due to biodiversity changes. This task started at the end of 2010 and continued during 2011. The first analyses using modelling results are now published in Fish and Fisheries (Coll and Libralato, 2011) and as a book chapter (Heymans et al., in press), and are in preparation for publication (Heymans et al. in preparation).
a) Scientific overview of the project
The objective of my research developed during the Marie Curie International Outgoing Fellowship (IOF) was to quantify how the structure of marine food-webs changes with various human stressors and develop an understanding of how these changes impact ecosystem functioning. To accomplish this objective I used an interdisciplinary approach applying meta-analysis of historical and recent data on species loss and ecosystem changes, food-web modelling, and laboratory experiments.
During the third and last year of the ECOFUN project, I finished the synthesis of ecological data on biodiversity changes in marine ecosystems, and the modelling of the consequences of these biodiversity changes on food-web structure and functioning, and continued the analysis of aquatic microcosm experiments to measure ecosystem functioning changes due to biodiversity changes.
A detailed description of the progress made during the third year is provided below:
(1) Synthesis of ecological data on biodiversity changes in marine ecosystems
I compiled a database to synthesise available data on long-term changes in marine biodiversity under gradients of different human impacts for some ecosystems (mainly some areas of the Mediterranean Sea and the Atlantic Canada). This database also records observed / measured consequences of these changes from individual species and populations to ecosystem levels in terms of structure or topology of the food web (e.g. absence / presence, trophic level, connectance) and functioning (changes in biomass, production, and trophic flows allocation and energy transfer efficiency, such as cascading effects, proliferations of non-abundant species and increase in detritus). The goal of this task was to document historical changes in ecosystems due to changes in marine resources.
In this context, the IndiSeas project, on which I have been collaborating since 2007, produced a database (see http//:www.indiseas.org online) that now is available to the scientific community and the general public. This database summarises now the state of 34 exploited marine ecosystems worldwide regarding fishing impacts and results document and analyze ecosystem changes in exploited ecosystems due to fishing activities going back to the mid 1960s. An updated on scientific advances of INDISEAS project have been recently submitted to reviews in fish biology and fisheries (Shin et al., submitted). Moreover, during November 2011, I attended a new meeting of INDISEAS where the phase of the project continued. The new phase will last two more years and it aims to:
1) extend the application of INDISEAS original ecological indicators to other ecosystems;
2) complement the list of ecological indicators to include biodiversity and conservation-based indicators, environmental indicators and socio-economic indicators; and
3) investigate ways to provide reference levels for indicators, as well as create composite indicators.
Within this second phase of the project, or INDISEASII, I am co-chairing the task group of biodiversity and conservation-based indicators with Dr. Lynne J. Shannon, from the University of Cape Town, South Africa. The participation on INDISEASII allowed me to collect new data from different ecosystems and perform more complex analysis with different human stressors.
In addition, during 2011 I continued the collaboration with Dr Denise Breitburg and colleagues from the marine ecology lab at the Smithsonian Environmental Research Centre in Maryland (in United States of America) under the context of the project 'Linking nutrients, hypoxia, fisheries and fishes'. This project aims at better understanding the effects of nutrient enrichment, hypoxia, and the interaction between these stressors with fisheries removals on fish biomass and the composition in estuaries and semi-enclosed seas ('fish' being used loosely here to include crabs, shrimp, etc.). In this project, we are constructing a database with data from nutrients and hypoxia events, fisheries and ecosystem traits and a series of publications are being prepared.
Building from the data that I have been collecting, during 2011 I also collaborated in a study on ecological history of the Adriatic Sea (Lotze, Coll and Dunne, 2011) and on recovery examples in marine ecosystems using historical data and modelling results (Lotze et al., 2011, Coll and Lotze, in preparation). Moreover, continuing the work I did under the umbrella of the Census of Marine Life in 2009 - 2010, a new study was published (Coll et al., 2011).
In Lotze, Coll and Dunne (2011), we used a multidisciplinary approach combining our paleontological, archeological, historical, fisheries, and ecological data to reconstruct past changes in marine populations, habitats, and water quality in the Adriatic Sea. Then, we constructed binary food webs for different historical periods to analyze possible changes in food-web structure and functioning over time. Our results indicated that human activities have influenced marine resource abundance since at least Roman times and have accelerated in the nineteenth and twentieth centuries. Today, 98 % of traditional marine resources are depleted to less than 50 % of former abundance, with large (>1 m) predators and consumers being most affected. With 37 % of investigated species rare and 11 % extirpated, diversity had shifted towards smaller, lower trophic level species, further aggravated by more than 40 species invasions. Species providing habitat and filter functions had been reduced by 75 %, contributing to the degradation of water quality and increased eutrophication. Increased exploitation and functional extinctions had altered and simplified food-web structure over time, especially by changing the proportions of top predators, intermediate consumers, and basal species. Moreover, simulations of species losses indicated that today's ecosystems may be less robust to species extinctions than in the past. Our results illustrated the long-term and far-reaching consequences human activities can have on marine food webs and ecosystems.
In Lotze et al. (2011), we reviewed the growing research on marine recoveries to reveal how common recovery is, its magnitude, timescale and major drivers. Results showed that overall, 10 - 50 % of depleted populations and ecosystems show some recovery, but rarely to former levels of abundance. In addition, recovery can take many decades for long-lived species and complex ecosystems. Major drivers of recovery include the reduction of human impacts, especially exploitation, habitat loss and pollution, combined with favourable life-history and environmental conditions. Awareness, legal protection and enforcement of management plans are also crucial. We concluded that learning from historical recovery successes and failures is essential for implementing realistic conservation goals and promising management strategies.
In Coll et al. (2011), we aimed to identify the main areas where the interaction between marine biodiversity and threats is more pronounced and to assess their spatial overlap with current marine protected areas in the Mediterranean. We first identified areas of high biodiversity of marine mammals, marine turtles, seabirds, fishes and commercial or well-documented invertebrates. We mapped potential areas of high threat where multiple threats are occurring simultaneously. Finally we quantified the areas of conservation concern for biodiversity by looking at the spatial overlap between high biodiversity and high cumulative threats, and we assessed the overlap with protected areas. Our results showed that areas with high marine biodiversity in the Mediterranean Sea are mainly located along the central and north shores, with lower values in the south-eastern regions. Areas of potential high cumulative threats are widespread in both the western and eastern basins, with fewer areas located in the south-eastern region. The interaction between areas of high biodiversity and threats for invertebrates, fishes and large animals in general (including large fishes, marine mammals, marine turtles and seabirds) is concentrated in the coastal areas of Spain, Gulf of Lions, north-eastern Ligurian Sea, Adriatic Sea, Aegean Sea, south-eastern Turkey and regions surrounding the Nile Delta and north-west African coasts. Areas of concern are larger for marine mammal and seabird species. These areas may represent good candidates for further research, management and protection activities, since there is only a maximum 2 % overlap between existing marine protected areas (which cover 5 % of the Mediterranean Sea) and our predicted areas of conservation concern for biodiversity.
(2) Modelling the consequences of biodiversity changes on food-web structure and functioning
The available datasets collected on biodiversity changes were used in ecological modelling techniques to analyse the consequences of biodiversity changes on food-web structure and dynamics. To do so, I used different food-web modelling techniques: stochastic static and bioenergetic dynamic modelling, and mass-balance and trophodynamic modelling. i have also incorporated the use of geographic information systems (GIS, mainly learning ArcGIS from ESRI software).
The ecological modelling approach included the analysis of the consequences of diversity changes on food webs in various Mediterranean seas (Coll and Libralato, 2011; Navarro et al., 2011; Lotze, Coll and Dunne, 2011) and on seagrass ecosystems in the Northern Atlantic (Coll et al., 2011b; Schmidt et al., 2011). Moreover, I am working on a last manuscript presenting results on the recovery of marine ecosystems (Coll and Lotze, in preparation).
In Coll and Libralato (2011), we present a meta-analysis of Ecopath with Ecosim applications in the Mediterranean Sea. Ecological modelling tools are worldwide applied to support the ecosystem-based approach of marine resources. Numerous applications of this approach have been attempted in the last decades in the Mediterranean Sea, mainly using Ecopath with Ecosim tool. These models are used to analyse a variety of complex environmental problems. Many applications analyse the ecosystem impacts of fishing and have been used to assess management options, such as the reduction of fishing effort or the establishment of protected areas. Other studies deal with the accumulation of pollution through the food web, the impact of aquaculture, or the ecosystem implications of climate change. These applications contributed to the scientific aspects of an ecosystem-based approach in the region since they integrate human activities within the ecosystem context and evaluate their impact along the marine food web, quantifying direct and indirect impacts, and including the environmental factors. These studies also gathered an important bulk of information at an ecosystem level. Therefore, as a second part of this study, we used this information to quantify main structure and functional traits of Mediterranean Sea marine ecosystems at regional scales. Results highlighted several differential traits between ecosystem types and a few between basins, which illustrate the environmental heterogeneity and complexity of the Mediterranean Sea. The analysis of key species evidenced the importance of top predators and small pelagic fish in Mediterranean ecosystems in addition to the structural role of benthos and plankton organisms. The impact of fishing was found to be of a similar intensity in the Western, Central and Eastern regions, with some differences between ecosystem types.
In Navarro, Coll et al. (2011), we examine the trophic dynamics of representative species of the South Catalan marine ecosystem (North-western Mediterranean) using the Ecopath ecosystem modelling tool and the stable isotope approach. By using the two methodologies we depicted the trophic position (trophic level and d15N values) and trophic width (omnivory index and total isotopic area) of several species of fish, cephalopods, cetaceans, seabirds and one sea turtle. Our results showed a clear correlation between the trophic levels estimated by the Ecopath model and the d15N values, which validate the trophic position of several species in the study area and indicate that both methodologies are useful to determine the trophic position of marine species in Mediterranean marine food webs. In contrast, the two estimators of trophic width (the omnivore index and the total isotopic area) were not strongly related, since the relationship between modelling and stable isotope results was evident only for some species. Further comparisons of trophic width calculated in other marine food webs may provide a better understanding of our results and validate the accuracy of both methodological approaches to calculate trophic width.
In Coll et al. (2011b), we analysed the main structural features of food webs associated with Zostera marina communities in New Brunswick, Prince Edward Island and Nova Scotia, in Atlantic Canada, and across a gradient of eutrophication. The goals of this study were to quantify changes in food-web structure across regions and eutrophication levels, to assess whether changes in food-web structure translate into changes in functioning by evaluating the robustness of food webs against simulated species loss, and to compare food-web structure in seagrass beds with those of other aquatic ecosystems worldwide. Based on data from field surveys, we constructed individual food webs for 16 study sites and cumulative food webs for regions and eutrophication levels. Network models were used to calculate 16 structural properties for each food web. Our results indicated that food-web structure was similar among low eutrophication sites across regions in Atlantic Canada. Moreover, our seagrass webs were similar to a tropical seagrass web, and significantly differed from other aquatic webs. A trend of increasing degradation was observed across the eutrophication gradient in both NB and PEI. With increasing eutrophication food webs showed more variability and less predictable patterns as well as a general structural simplification indicated by fewer trophic groups, lower maximum trophic level of the highest top predator, fewer number of trophic links connecting top to basal species, a higher fraction of herbivore and intermediate consumers, and higher number of prey species per predator. Several food-web properties showed non-linear responses to increasing eutrophication and partially differed between regions indicating site-specific attributes that modulate the responses to enrichment. More eutrophied sites are also less robust to species loss. The effect of species deletion was stronger in PEI compared to NB and NS, and may indicate a macro-ecological gradient of degradation across regions. Overall, our study suggested that eutrophication alters food-web structure and functioning of seagrass habitats, and that the spatial scale at which food-web structure is studied is a critical factor determining the results of food-web analyses.
In Schmidt, Coll et al. (2011), we used models of predator-prey feeding interactions to evaluate food-web structure and their robustness to simulated species removals. Despite disparate three-dimensional canopy structure, both habitats significantly enhanced overall abundance and diversity of associated flora and fauna. Yet significant differences occurred in the species assemblages within and between habitats that are attributed to different settlement opportunities, food availability, predation risk, and maneuverability. While eelgrass was a better nitrogen filter, rockweed maintain eight-fold greater biomass and thus 14-fold greater nitrogen and eight -fold greater carbon storage. 53 % of the food-web properties differed between the two habitats. Eelgrass communities showed higher food-web complexity, and a greater robustness to simulate species removals. Our study demonstrated that marine vegetation provided important habitat, food, shelter and stores nitrogen and carbon, yet the extent of these services depended on the foundation species and its architecture. Changes in canopy structure will have profound effects on associated food webs and ecosystem services. Thus, increasing human pressures on coastal ecosystems threaten the continued supply of essential functions and services these ecosystems provide, and the protection of marine vegetated habitats should be a management priority.
During this year 2011, I also visited the Fisheries Centre (University of British Columbia, Vancouver, Canada) for three months (September - December). There I participated in the Ecopath research and development consortium initial meeting.
(3) Application of aquatic microcosm experiments to measure ecosystem functioning changes due to biodiversity changes
In parallel, in 2011 I continue analysing the results from the aquatic microcosm experiments than were conducted in 2008 and 2009 to observe patterns of human impacts on marine food webs under laboratory conditions and to quantify changes in structure and function. Different methodologies and techniques were used, including the analysis of stable isotopes and bacterial activity. During 2011, I finished the analysis of the results of experiments performed during 2008 (preliminary trial) and 2009, where a series of microcosm experiments were developed to reproduced different levels of diversity change under climate change scenarios in laboratory conditions. The first results are now published (Coll and Hargadon, in press) and another one is being prepared (Coll et al., in preparation).
In Coll and Hargadon (in press), we explored food web changes in trophic structure and ecosystem functioning following biomass removal of top predators in representative temperate and tropical rock pool communities that contained similar assemblages of zooplankton and benthic invertebrates. We observed changes in species abundances following predator removal in both temperate and tropical communities, in line with expected inverse effects of a trophic cascade, where predation release benefits the predator's preys and competitors and impacts the preys of the latter. We also observed several changes at the community and ecosystem levels including a decrease in total abundance and mean trophic level of the community, and changes in chlorophyll-a and total dissolved particles. Our results also showed an increase in variability of both community and ecosystem processes following the removal of predators. These results illustrate how predator removal can lead to inverse trophic cascades both in structural and functioning properties, and can increase variability of ecosystem processes. Although observed patterns were consistent between tropical and temperate communities following an inverse cascade pattern, changes were more pronounced in the temperate community. Therefore, aquatic food webs may have inherent traits that condition ecosystem responses to changes in top-down trophic control and render some aquatic ecosystems especially sensitive to the removals of top predators.
(4) Meta-analysis on ecosystem functioning changes due to marine biodiversity changes
The application of a meta-analytical approach combining the results from data synthesis, modelling and laboratory experiments were used to quantify general patterns of structural and functional food-web changes across ecosystems due to biodiversity changes. This task started at the end of 2010 and continued during 2011. The first analyses using modelling results are now published in Fish and Fisheries (Coll and Libralato, 2011) and as a book chapter (Heymans et al., in press), and are in preparation for publication (Heymans et al. in preparation).