Elsevier

Applied Geography

Volume 32, Issue 2, March 2012, Pages 652-659
Applied Geography

Spatial patterning of Manta birostris in United States east coast offshore habitat

https://doi.org/10.1016/j.apgeog.2011.07.015Get rights and content

Abstract

This study aims to identify patterns in spatial distribution of Manta birostris along American East Coast by analyzing presence data using a suite of geostatistical techniques. Analysis of data from 1979 to 2008 revealed that the spatial patterns exhibited by the M. birostris include a strong clustering zone off the North Carolina from (however, not all years are represented in that range), Virginia and Maryland coast. Their presence is high near the coastal shelf edge and along the edge of the Gulf Stream. The mix of warm water and high chlorophyll concentration throughout the year in the region seem to create an ideal environment for this clustering pattern. However, in seasons with warmer water temperature, manta rays spread further north and inshore, possibly to avoid competition for prey. The results of geographic weighted regression indicate the role of sea surface temperatures on Manta presence or absence off the main current of the Gulf Stream near North Carolina and Virginia. Knowledge of the spatial distribution of M. birostris can assist in further protecting the species especially from dangers of by-catch and habitat degradation.

Highlights

M. birostris off the American East Coast display seasonal clustering patterns. ► Clusters are found where the shelf edge and Gulf Stream meet. ► Cold seasons display the most clustering likely due to strong attraction warm waters. ► Geographic weighted regression found that temperature played a role in patterning.

Introduction

Manta birostris, also called the manta ray (Donndorff, 1798), is a large pelagic ray found worldwide in tropical and subtropical waters (Bigelow and Schroeder, 1953, Compagno, 1999). Manta rays are in the Mobulid ray family and are the largest batoid ray (Bigelow & Schroeder, 1953). Mantas are of near threatened status according to the International Union for the Conservation of Nature or IUCN (Marshall et al., 2006); threats to manta rays include by-catch issues, shark netting, diving tourism impacts and uncontrolled direct fisheries (Dewar et al., 2008, Marshall et al., 2006, Paylado, 2004, Tibiriçá Yara et al., 2009, Zeeberg et al., 2006). These rays do utilize home ranges and preferred zone sites, tending to return to preferred cleaning stations and reefs sequentially. Because of their home range nature, direct local impacts on this species, although small, may have immense impacts on the global population. A study done by Dewar et al. (2008), with different styles of tags in the Komodo Marine Park in Indonesia, found that 81% of manta visits were to the same site. Site fidelity is common to multiple coastal and near-shore habitats across the globe, sometimes up to 15 years of fidelity to one location (Dewar et al., 2008, Homma et al., 1999; Notarbartolo-di-Sciara & Hillyer, 1989). However, M. birostris is a seasonal visitor to many areas and visitation to a specific feeding or cleaning station can occur over long periods. Individual mantas have been observed to return to the same cleaning stations for years with large temporal gaps between visitations. Even with this behavior being common in some areas, it is also common for individuals to be seen only once (Dewar et al., 2008). Habitat destruction and harvesting of mantas can be acute in small local areas, but because species are fond of these locations, these aforementioned impacts may parlay into huge population issues. A recent paper even found a M. birostris individual with melanistic traits previously seen only in the Pacific Ocean, in Atlantic waters off Brazil (Luiz, 2009). This would suggest individuals could travel transoceanically or even perhaps a transient population. Mantas have been recorded traveling up to 350 km and it is postulated that longer migrations are possible and likely (Dewar et al., 2008, Homma et al., 1999).

Coastal movements and abundance is better recorded than pelagic movements and these movements have been found to be heavily influenced by seasons, tides, lunar cycles, seasons, currents, and temperature (Dewar et al., 2008). Mantas are zooplanktovores, much like the whale sharks, Rhincodon typus, and are shown to employ similar feeding techniques and seasonality distributions in shared habitats (Luis Jr., 2009; Wilson, Pauly & Meekan, 2002). Both mantas and whale sharks have been recorded selectively feeding on surface populations of Pseudeuphausia latifrons, a zooplanktonic euphausiid (Hamner and Hamner, 2000, Wilson et al., 2002). Whale sharks are currently marked as Vulnerable by the IUCN. Currents, topography interactions, and upwelling create huge plankton blooms, which attract both mantas and whale sharks (Chin and Kyne, 2007, Monteiro et al., 2008). Movements of whale sharks have been linked to plankton blooms and currents associated with El Niño (Chin & Kyne, 2007). Off Ningaloo reef in Australia, mantas and whale sharks frequently inhabit locations with high surface convergence, strong vertical motion, and intense mixing that create dense planktonic aggregations and productivity (Wilson, Pauly & Meekan, 2002). In conjunction with that, whale shark distributions have been weakly correlated with sea surface temperatures (Chin & Kyne, 2007). A study done on zooplankton showed that abundances and biomass of this critical manta food source had strong patterns of correlation with temperature and chlorophyll concentration (Wilson, Meekan, Carleton, Stewart, & Knott, 2003) and highest observances of mantas in Ningaloo reef are at the peak of summer chlorophyll concentrations (Wilson, Pauly & Meekan, 2002). Due to similar food sources, hunting behaviors and habitat overlap, studies on whale sharks are a good model for transoceanic plankton feeders like mantas.

Due to the transoceanic pelagic nature of this species, it is critical to understand the spatial and temporal behaviors of this ray in order to ensure proper management. Mantas frequent high productivity areas, which also attract commercially important species (Zeeburg et al., 2006). By-catch vulnerability, population susceptibility and changes in environmental factors due to global climate change (Chin & Kyne, 2007), leave the M. birostris a highly susceptible species. However, there are very little published data on the species making research on distribution of manta ray’s, which is highly important. This study aims to assess environmental conditions mantas frequent in order understand possible abiotic and biotic relationships between the organism and its environment for its protection.

Geospatial techniques have been successfully used to examine the spatial patterns of animal species on both land and water. For instance, Jelinski, Krueger, and Duffus (2002) examined the spatio-temporal dynamics of the encounters between killer whales and different types of whale watching vessels in and around their reserves using geostatistical techniques, including movement of motorized and non-motorized whale watching vessels. The results of the study showed significant role of motorized whale watching vessels on the movement of whales. In another study by Schick et al., (2009), geostatistical techniques were used to examine the spatial patterns of whales around oil exploration activities in the Alaskan Beaufort Sea. The results of the study indicated a siginifcant temporary loss of available habitat near a drilling rig. GIS techniques have also been used to study the spatial distribution of terrestrial species in different parts of the world such as identification of eastern spadefoot toad habitat in eastern Connecticut (Moran & Button, 2011), Various spatial characteristics were taken into consideration in order to generate a GIS model, which included drainage class, texture class and soil deposit type along with digital elevation models used to calculate the relative elevation. Furthermore, geospatial techniques in the form of remote sensing have also been used to study larger areas, such as in case of eastern Kalimantan, Indonesia (Fuller, Meijaard, Christy, & Jessup, 2010). Composite maps of land use were evaluated statistically related to conservation threats, which revealed distinct spatial patterns in terms vulnerability of forests. Therefore, in this study a suite of GIS techniques have been used to analyze the spatial pattern of M. birostris in coastal areas.

Section snippets

Data and methods

Data on manta observances were obtained from the Ocean Biogeographic Information System (OBIS) run by Rutgers University. Originally part of the Census of Marine Life, the OBIS project allows for a collaborated effort to record the presence of important marine species across the globe. Environmental data included scientific quality Sea Surface Temperature (SST) and Cholorphyll from the Oceanwatch database. SST data were obtained from NASA’s Pathfinder Version 5.0m, while Chlorophyll-A (CHL-A)

Results

Four populations of M. birostris were identified from the OBIS data: the South American Pacific, South African Eastern Cape, California Baja, and American East Coast (AEC). Kernel density analysis revealed these locations as hot spots, however, they do not include every individual in the dataset. This technique was an effective tool in studying clustering effects because it does not only account for point density in a cell but also the spatial relationships between points (Fig. 1a). An

Discussion and conclusion

M. birostris is a species found typically in the tropics, so finding individuals in colder AEC waters may seem unusual. However, temperature seems to be a guiding factor in the manta’s distribution for this area. Highest values of both the C1 and R2 values are found in habitats that border the Gulf Stream; thus indicating significant influence on pelagic habitat selection in that region. The temperature preference for AEC M. birostris was from 19 to 22° °C, while in the Komodo Marine Park the

Acknowledgments

We would like to thank Jacob Crows for his assistance with image preparation for this paper. Also we are grateful for help and advice of Maria Estevanez throughout the publishing process.

References (24)

  • A.S. Fotheringham et al.

    Geographically weighted regression: the analysis of spatially varying relationships

    (2002)
  • W.M. Hamner et al.

    Behaviour of Antarctic krill (Euphausia superba): schooling, foraging, and antipredatory behaviour

    Canadian Journal of Fisheries and Aquatic Science

    (2000)
  • Cited by (18)

    • Spatiotemporal persistence of bat roadkill hotspots in response to dynamics of habitat suitability and activity patterns

      2021, Journal of Environmental Management
      Citation Excerpt :

      Thus, spatial and temporal changes in hotspots can give insights regarding distribution and predictability of important resources on the landscape (Fensome and Mathews, 2016). The notion of shifting hotspots relates to the concept of home range fidelity (Börger et al., 2006; Freedman and Roy, 2012; Switzer, 1997). The consistent classification of a road section as hotspot over time, termed hotspot fidelity, may reflect the existence of permanent important landscape features for bat ecological needs, such as foraging habitats or roosts in roadsides.

    • Predicting mobulid ray distribution in coastal areas of Lesser Sunda Seascape: Implication for spatial and fisheries management

      2020, Ocean and Coastal Management
      Citation Excerpt :

      Mobulid rays are highly mobile species, of which the movements are driven by biological and physiological requirements such as feeding, mating, cleaning, and thermoregulation (Ari and Correia, 2008; Jaine et al., 2012; Jaine et al., 2014; Rohner et al., 2013; Armstrong et al., 2016; Rohner et al., 2017a). Because mobulid rays are mainly zooplanktivorous, their occurrence is usually correlated with local oceanographic conditions and in particular, zones of high surface productivity as shown in many parts of their ranges (Freedman and Roy, 2012; Stewart et al., 2017), including in the northern Savu Sea region (Putra and Mustika, 2020b). This study confirms that the habitat preference of mobulid rays were distributed in the areas close to the 200-m isobath and those accounting for higher chlorophyll-a concentration as a proxy for their prey density.

    • Extending geographically and temporally weighted regression to account for both spatiotemporal heterogeneity and seasonal variations in coastal seas

      2018, Ecological Informatics
      Citation Excerpt :

      The hypothesis of space-time stationarity in global models is usually impractical since parameters tend to change across the research regions and over time. Therefore, some local models have been put forward to capture spatial variability in marine environmental processes (Freedman and Sen Roy, 2012; Keith et al., 2013). For instance, geographically weighted regression (GWR) model was formulated and fitted by Keith et al. (2013) to investigate the variations in the relations between macroalgae richness and environmental conditions over geographical regions.

    • Comparison of spatial and non-spatial logistic regression models for modeling the occurrence of cloud cover in north-eastern Puerto Rico

      2013, Applied Geography
      Citation Excerpt :

      The application of GWR has grown in forestry, ecology, and social sciences over the last decades. For example, it was used to model tree growth in northeast of US (Zhang & Shi, 2004), estimate net primary production for Chinese forest ecosystems (Wang, Ni, & Tenhunen, 2005) and aboveground biomass in tropical rainforests in Indonesia (Propastin, 2012), explore the relationships between water quality and physical properties or land use/land cover of the watersheds (Tu, 2011; Wu & Driscoll, 2009), analyze wealth and land cover in Massachusetts (Ogneva-Himmelberger, Pearsall, & Rakshit, 2009), explore multiple factors that caused afforestation in Hoa Binh Province of Northern Vietnam from 1993 to 2000 (Clement, Orange, Williams, Mulley, & Epprecht, 2009), or deforestation in the State of Mexico from 1993 to 2000 (Jaimes, Sendra, Delgado, & Plata, 2010), examine the influence of climate and land cover pattern on bird species richness in New York (Ma, Zuckerberg, Porter, & Zhang, 2012), identify spatial patterns of the distribution of Manta birostris along East Coast in the US (Freedman & Roy, 2012), and distributing of stray cats in Auckland, New Zealand (Aguilar & Farnworth, 2012). Very few studies have used response variables with binomial distribution (Atkinson, German, Sear, & Clark, 2003).

    • Quantifying spatial-temporal patterns in wildlife ranges using STAMP: A grizzly bear example

      2012, Applied Geography
      Citation Excerpt :

      This issue becomes particularly problematic for low-density wide-ranging populations in highly heterogeneous landscapes (Apps, McLellan, Woods, & Proctor, 2004; Wielgus & Bunnell, 1995) or when there is high inter-population variability in behaviour (e.g., Wiens, Crawford, & Gosz, 1985). The notion of shifting home ranges relates to the concept of home range fidelity, which has been described for a number vertebrate species (Borger et al., 2006; Freedman & Roy, 2012; Switzer, 1997). Home range fidelity occurs when a particular animal leaves a given site, say upon migration or hibernation, but returns to the previously occupied and familiar site (Switzer, 1993).

    View all citing articles on Scopus
    View full text