Research

Objective

The objective of this project is to evaluate the feasibility and effectiveness of spatially-explicit protective measures for highly-mobile marine vertebrates in the Alborán Sea, Western Mediterranean.

This study addresses three groups of marine vertebrate susceptible to a broad range of human impacts: birds, turtles, and cetaceans (Dayton et al. 1995, Roberts & Hawkins 1999). To evaluate diverse habitat-wildlife associations and potential MPA designs we will focus on a number of organisms with a wide range of life-histories.

Approach

We will follow a three step approach. First, we will quantify oceanographic habitats of marine birds, mammals, and turtles found in the Alborán Sea during summer using vessel-based surveys and concurrent oceanographic data from satellites. Next, we will use gap analysis to identify "priority" areas for conservation, by integrating these species-specific results into a comprehensive perspective of the areas and habitats where marine vertebrates overlap and concentrate. Finally, to assess the effectiveness of proposed MPAs and potential alternative designs, we will quantify the spatial and temporal predictability of those oceanographic features identified using predictive models by developing long-term climatologies of the available satellite imagery.

Explanatory Variables: Our analyses will consider both abiotic (e.g., bathymetry and hydrography) and biotic (e.g., ocean productivity) habitat features, and will include a variety of variables: static (distance to shore, median depth, depth gradients), dynamic (SST, SST gradients, Chl a, Chl a gradients, SSHA), survey effects (area, sea state), and a year co-variate to assess inter-annual variability. Sea surface temperature (SST) and chlorophyll a concentration (Chl a) are proxies of water mass distributions, fronts, and ocean productivity in the Mediterranean, and can be readily measured from satellites (Louzao et al. 2006). We will use daily and 8-day MODIS composites of SST and CHL, with a resolution of 4-km, publicly available since July 2002 (modis.gsfc.nasa.gov). To quantify sea surface height anomalies (SSHA) indicative of convergences / divergences and eddies, we will use the AVISO altimetry. We will sub-sample these coarse data, with a resolution of 0.25 o and a 10-day repeat-cycle, to match the MODIS imagery (Hyrenbach et al., 2006).

Model Building: To evaluate habitat-wildlife associations and potential MPA designs, we will focus on the most numerous species in the study area during summer (June - Sept.). We anticipate developing habitat models for 16 species with sufficient samples sizes. The cetacean / turtle models will build upon previous analyses and will rely on recent survey data (2002-06) (Cañadas et al. 2005). Because seabirds have not been surveyed quantitatively in the past, their models will rely on summer surveys during this study (2007-08). We will use habitat models to characterize the distributions of selected species in the study area: generalized linear models and generalized additive models (Redfern et al. 2006). Following our previous research, we will use a hierarchical approach to identify the habitat of the focal species by: (1) delineating their "range" using presence / absence data; and (2) identifying "concentration areas" where they aggregate. For each species, we will assemble the best-fit model using a step-wise procedure (Hyrenbach et al. 2006).

Model Validation: A critical aspect of habitat modeling entails validating the models with additional field data (Forney 2000, Redfern et al. 2006). We will use survey data from a cruise in the second year of the study (2007) to validate the models developed for vertebrate groups in the first year.

Gap Analysis: We will use "decision-support" GIS tools to integrate the species-specific survey data by pinpointing those areas where marine vertebrates aggregate, and to assess how MPA designs - with different areas and configurations - encompass these "hotspots". In particular, we will use the MARZONE software, a new version of the MARXAN software capable of incorporating multiple marine zone types and costs, currently under development to complete this task. This tool uses simulated annealing to develop objective functions - including multiple costs, penalties, and boundary designs - to facilitate the quantitative assessment of conservation goals under various management scenarios (Ball & Possingham 2000, Stewart & Possingham 2005).

Anticipated Results

Products - We will create the following tangible research and outreach products with collaborators. Most importantly, we will contribute to global portals of public information: the Network of Conservation and Education Practitioners (NCEP) program and the Ocean Biogeographic Information System (OBIS) portal.

Publications: We anticipate publishing four scientific papers, during this project:
- #1: Habitats of marine vertebrates in a dynamic seascape: the Alborán Sea.
- #2: Identifying Upper-trophic Predator Hotspots: the Alborán Sea.
- #3: Assessing the predictability of dynamic habitats: the Alborán Sea.
- #4: Putting it all together: Important Pelagic Habitats in the Alborán Sea.

Outreach: We will disseminate the results of our research to a broad audience, through four outreach venues: (i) posting cruise and project results on the Alnitak web-site, (ii) presenting yearly research results to WCPA and ACCOBAMS , (iii) making presentations at local coastal communities in Andalucia (Spain) after the 2007 and 2008 cruises, and (iv) presenting the results of our work at international scientific conferences in 2008 and 2009.

Contributions - We will make three major contributions to marine conservation:
Locally, the results of this project will reinforce the conservation importance of the Alborán Sea in three ways: (i) by advancing the habitat models previously developed for cetaceans (see methods section), (ii) by developing complementary habitat models for seabirds and loggerhead turtles in this area, and (iii) by identifying ecologically important areas at a broader ecological context, by both using gap analysis to combine the modeling results for individual taxa and by developing new ecosystem-level metrics. More specifically, our modeling results will provide a stronger scientific rationale for establishing a SPAMI in the Alborán Sea. Ultimately, it is our hope that these protections will mitigate certain human impacts (e.g. bycatch, oil spills, ship strikes) on these species.

Regionally, the results of this project will underscore the importance of establishing a network of representative Mediterranean MPAs not only for Spain, but on a broader basin-wide scale. Protecting an important migration and feeding area of many highly-migratory species will highlight the need for establishing other monitoring and protective measures for these species outside of these restricted MPAs. Moreover, the ability to identify predictable habitat-wildlife associations in the notoriously dynamic Alborán Sea will inspire similar studies in other areas, like Tunisia and Algeria (WWF 2006). In addition to motivating research and conservation actions, this project will educate the public at large about pelagic conservation and MPA networks.

On a global scale, this project will advance marine conservation science by providing the conceptual and quantitative tools for evaluating pelagic MPA sites and designs. More specifically, the results of the habitat modeling and gap analyses will demonstrate how sound science can provide the quantitative basis for setting and evaluating MPA designs, even when dealing with highly-mobile species and pelagic systems. This study will provide robust scientific approaches for the designation of MPAs, and for the objective measurement of their success through monitoring of habitat use patterns inside and outside the selected areas.