Mirrored: 22nd of January 2012, 18:16 Original: www.mendeley.com Views: 33 Settings: Loading the mirror...
Mendeley reference manager logo Sign up for free Sign in E-mail address _____________________ Password (forgot?) _____________________ [ ] Remember me [ Sign in ] ...or sign in with Facebook Papers Search papersSearch groupsSearch people _____________________ Search * Main Navigation * Get Mendeley * What is Mendeley? * Papers * Groups Modeling the spatial autocorrelation of pelagic fish abundance by Km Kleisner, Jf Walter, Sl Diamond, Dj Die Biological Sciences > Miscellaneous Papers In your library Save reference to library Share _____________________ [ ] Short URL * Share on Facebook * Share on Twitter * Email this link * Overview * Related research Marine Ecology Progress Series (2010) Volume: 411, Pages: 203-213 * ISSN: 01718630 * DOI: 10.3354/meps08667 Available from www.int-res.com or Find this paper at: * openurl.ac.uk * WorldCat(R) * Google Scholar * Edit library access links Abstract ABSTRACT: The relationship between pelagic fish and ocean temperature is cited in many studies, the majority of which investigate correlations of pelagic species and sea surface temperatures (SST). While appropriate for surface-associated species, this may not be accurate for deep-diving fishes. A different way to examine this relationship is to model spatial autocorrelation of fish species and temperatures at an appropriate range of depths. Spatial autocorrelation, the distance at which data are interdependent, is a potential descriptor of the patch size of an organism. Here we modeled spatial autocorrelation for 5 pelagic species that inhabit different depths in the Gulf of Mexico: dolphinfish Coryphaena hippurus, wahoo Acanthocybium solandri, yellowfin tuna Thunnus albacares, swordfish Xiphias gladius, and bigeye tuna Thunnus obesus. Additionally, we modeled spatial autocorrelation for ocean temperatures at the surface, at 200, and 400 m. We hypothesized that autocorrelation distances will be greater for deeper water temperatures and for species that live at deeper depths due to greater homogeneity of deep waters over greater spatial ranges. Results show average distances of autocorrelation on the order of 55 to 60 km for wahoo and dolphinfish, 90 km for yellowfin tuna, and 135 to 145 km for swordfish and bigeye tuna; the same data for temperature were 75, 135, and 300 km for SST, and at 200 and 400 m depth, respectively. Autocorrelation distances of dolphinfish, wahoo, and yellowfin were correlated with that of of SST, while the autocorrelation distance of swordfish and bigeye were correlated with that of temperatures at 200 m. Results suggest that autocorrelation distances may be useful as a proxy for habitat delineation. Author-supplied keywords bigeye tunadolphinfishgeostatisticspelagic habitatpermitted without written consentpublisherresale republication notspatial analysisswordfishvariogramwahooyellowfin tuna Related research 1. Spatial autocorrelation and assessment of habitat-abundance relationships in littoral zone fish S G Hinch, K M Somers, N C Collins in Canadian Journal of Fisheries and Aquatic Sciences (1994) Save reference to library . Related research 7 readers 2. Variation in hydroacoustic abundance of pelagic fish S Hansson in Fisheries Research (1993) Save reference to library . Related research 3 readers 3. Prey harvests of seabirds reflect pelagic fish and squid abundance on multiple spatial and temporal scales WA Montevecchi, R A Myers in Marine ecology progress series Oldendorf (1995) Save reference to library . Related research 2 readers 4. Spatial autocorrelation as a tool for identifying the geographical patterns of aphid annual abundance Nadege Cocu, Richard Harrington, Maurice Hulle, Mark D A Rounsevell in Agricultural and Forest Entomology (2005) Save reference to library . Related research 11 readers 5. Failure to find the relationship between dispersal and spatial autocorrelation in species abundance Volker Bahn in Journal of Negative Results (2008) Save reference to library . Related research 8 readers More related papers Cite this document (BETA) * APA * BibTeX * Cell * Chicago * Harvard * MLA * Nature * Science APA BibTeX Cell Chicago Harvard MLA Nature Science [IMG] Click to zoom in 2 pages available to preview Available from www.int-res.com Close Page 1 hidden Modeling the spatial autocorrelation of pelagic fish abundance MARINE ECOLOGY PROGRESS SERIES Mar Ecol Prog Ser Vol. 411: 203-213, 2010 doi: 10.3354/meps08667 Published July 29 INTRODUCTION The biosphere is composed of many `patchy' eco- systems related to spatially-structured biological and environmental phenomena. Our ability to quantify spatial structure lies in the degree to which spatial autocorrelation, the increased similarity of variables at shorter distances (Legendre 1993), is present in these systems. Although most ecological phenomena display spatial autocorrelation, ecological studies have tradi- tionally taken a non-spatial perspective (Legendre & Fortin 1989). Recently, ecologists have begun to em- phasize the degree to which organisms are influenced by spatial patterns in the environment (Liebhold & Gurevitch 2002, Giannoulaki et al. 2003). There are 2 main approaches to dealing with spatial autocorrelation. The first is to remove spatial autocor- relation from consideration through random allocation (c) Inter-Research 2010 . www.int-res.com*Email: k.kleisner@fisheries.ubc.ca Modeling the spatial autocorrelation of pelagic fish abundance Kristin M. Kleisner1, 4,*, John F. Walter III2, Sandra L. Diamond3, David J. Die1 1Rosenstiel School of Marine and Atmospheric Science, University of Miami, 4600 Rickenbacker Causeway, Miami, Florida 33149, USA 2Southeast Fishery Science Center, NOAA, 75 Virginia Beach Drive, Miami, Florida 33149, USA 3School of Natural Sciences, University of Western Sydney, Locked Bag 1797, Penrith South DC, New South Wales 1797, Australia 4Present address: Fisheries Centre, University of British Columbia, 2202 Main Mall, Vancouver, British Columbia V6T 1Z4, Canada ABSTRACT: The relationship between pelagic fish and ocean temperature is cited in many studies, the majority of which investigate correlations of pelagic species and sea surface temperatures (SST). While appropriate for surface-associated species, this may not be accurate for deep-diving fishes. A different way to examine this relationship is to model spatial autocorrelation of fish species and temperatures at an appropriate range of depths. Spatial autocorrelation, the distance at which data are interdependent, is a potential descriptor of the patch size of an organism. Here we modeled spa- tial autocorrelation for 5 pelagic species that inhabit different depths in the Gulf of Mexico: dolphin- fish Coryphaena hippurus, wahoo Acanthocybium solandri, yellowfin tuna Thunnus albacares, swordfish Xiphias gladius, and bigeye tuna Thunnus obesus. Additionally, we modeled spatial auto- correlation for ocean temperatures at the surface, at 200, and 400 m. We hypothesized that autocor- relation distances will be greater for deeper water temperatures and for species that live at deeper depths due to greater homogeneity of deep waters over greater spatial ranges. Results show average distances of autocorrelation on the order of 55 to 60 km for wahoo and dolphinfish, 90 km for yel- lowfin tuna, and 135 to 145 km for swordfish and bigeye tuna; the same data for temperature were 75, 135, and 300 km for SST, and at 200 and 400 m depth, respectively. Autocorrelation distances of dolphinfish, wahoo, and yellowfin were correlated with that of of SST, while the autocorrelation dis- tance of swordfish and bigeye were correlated with that of temperatures at 200 m. Results suggest that autocorrelation distances may be useful as a proxy for habitat delineation. KEY WORDS: Spatial analysis . Geostatistics . Variogram . Pelagic habitat . Dolphinfish . Bigeye tuna . Yellowfin tuna . Wahoo . Swordfish Resale or republication not permitted without written consent of the publisher Page 2 hidden Mar Ecol Prog Ser 411: 203-213, 2010 of treatments to plots (Fisher 1926), i.e. random sample selection (Cochran 1977, Thompson 2002). This ap- proach is suitable if one has the luxury of a random sampling design. When this can not be achieved, the alternative approach is to use a model to describe the spatial autocorrelation of the process of interest, and then use the model to make predictions at unsampled locations (Matheron 1971, 1973). The model-based approach is useful for situations where the spatial autocorrelation of the process itself has ecological importance (Bez 2002) as certain parameters of the spatial model can reveal spatial patterns (i.e. patch size of an organism or spatial extent of suitable habitat), which reflect the biology of the species (Freire et al. 1992, Rossi et al. 1992). This study focuses on large pelagic fishes such as tuna and billfish whose role as apex predators is criti- cal to the structure and function of the pelagic eco- system. The range of movement of these species in both the vertical and horizontal spatial dimensions can vary greatly. Many of the large pelagic species are cir- cumglobal in distribution and highly migratory, and some, like bluefin tuna Thunnus thynnus and blue marlin Makaira nigricans, exhibit transoceanic move- ments (Block et al. 2001, Orbesen et al. 2008). Some pelagic fish species such as sailfish Istiophorus platy- pterus choose a specific depth range (0 to 60 m) as pre- ferred habitat (Hoolihan & Luo 2007), while others such as yellowfin tuna Thunnus albacares exhibit a diel pattern in depth distribution, diving to deeper waters during the day and remaining above the thermocline at night (Weng et al. 2009). Most fish populations, including pelagic fishes, ex- hibit positive spatial autocorrelation, not only because fish often form schools but also because the environ- mental features that serve to aggregate or structure fish distributions are also spatially autocorrelated (Schneider 1989, Petitgas et al. 2001, Nishida & Chen 2004). The idea that regional oceanographic features may define pelagic habitat is particularly important because there are no permanent structures in the pela- gic realm, only transitory water conditions that may define habitat patches. Temperature is highly corre- lated with pelagic fish distributions, with fish aggre- gating along temperature fronts or within a particular temperature zone (Fiedler & Bernard 1987, Block et al. 2001, Chen et al. 2005, Prince & Goodyear 2006, Loefer et al. 2007). Since water characteristics at depth are generally more homogeneous than at the surface (Cummins et al. 1990, Weaver & Sarachik 1990), tem- perature isobaths will likely occur over a greater hori- zontal distance at depth than at the surface. One impli- cation is that a species at the surface may be subject to greater partitioning of suitable habitat than a species that spends more time in deeper waters. Therefore, one might expect smaller patches of organisms at the surface than at depth. In this paper, we explore the spatial autocorrelation of marine pelagic species and ocean temperatures rel- ative to depth. The objectives of this work are (1) to determine differences in the spatial autocorrelation of a suite of pelagic fish species that occupy different ver- tical habitats as well as differences in spatial autocor- relation of water temperature at increasing depths, and (2) to determine what relationships exist in the patterns of spatial autocorrelation within species and between species and temperature at depth. We hypothesize that (1) due to greater homogeneity of ocean waters at depth, the distance at which species are spatially autocorrelated will be greater for deeper- dwelling species; (2) similarly, the distance of autocor- relation of water temperature measurements will in- crease with increasing depth. MATERIALS AND METHODS Study area. The Gulf of Mexico (GOM) is a semi- enclosed sea that is connected in the east to the Atlantic Ocean through the Straits of Florida, and in the south to the Caribbean Sea through the Yucatan Channel (Fig. 1). The dominant oceanographic feature in this region is the Loop Current, which enters the GOM from the Yucatan Channel and forms the begin- ning of the Gulf Stream. The Loop Current and the large anticyclonic rings that it sheds strongly influence the spatial structure of the biota in the GOM (Olson 2002). Description of data. Logbook catch data (1987 to 2005) from the GOM reported by commercial US long- line vessel captains for each set throughout the entire year (Beerkircher et al. 2003) were used in this study. Most of the longline fleet sets their gear in the north- eastern Gulf along Loop Current fronts or associated rings. Most surface longlines in the GOM are up to 30 km long and typically fish between 50 and 300 m below the surface with an average of 1000 hooks per set (Kerstetter 2005). The coordinates of the start of the sets are expressed as degrees latitude and longitude, and for this analysis they were projected to an Albers equidistant coordinate system using ArcGIS version 9. We used data from the 5 species most commonly encountered by commercial longlines in the GOM for the analysis. Dolphinfish Coryphaena hippurus and wahoo Acanthocybium solanderi are strictly epipelagic species (Ward & Myers 2005, Hammond 2006). Yel- lowfin tuna moves between epipelagic and deep waters (Holland et al. 1990, Brill et al. 1999). Swordfish Xiphias gladius and bigeye tuna Thunnus obesus are deep-dwelling (Musyl et al. 2003, Fritsches et al. 2005). 204 Sign up today - FREE Mendeley saves you time finding and organizing research. Learn more * All your research in one place * Add and import papers easily * Access it anywhere, anytime Start using Mendeley in seconds! _____________________ _____________________ _____________________ [ Sign up & Download ] Already have an account? Sign in Facebook Readership Statistics 9 Readers on Mendeley by Discipline 56% Biological Sciences 22% Mathematics 22% Environmental Sciences by Academic Status 44% Ph.D. Student 22% Researcher (at a non-Academic Institution) 22% Post Doc by Country 22% Canada 11% Iceland 11% United Kingdom * What is Mendeley? Features Overview Reference Manager How We Help Our Users Compare About Us Upcoming Events & Webinars Become an Advisor Awards & Reviews Our Team Jobs Contact Us Support Videos & Tutorials Citation Styles FAQ Feedback Release Notes Support Useful Links Blog Install Web Importer University Endorsements Developers Mendeley API Download and Upgrade Download Mendeley Free iPhone & iPad App Premium Packages Copyright (c) 2012 Mendeley Ltd. All rights reserved.Terms of UsePrivacy PolicyCopyrightFind us onFacebookTwitterYouTube