A Qualitative Analysis of Marine Mammal Distribution Using GIS



A Qualitative Analysis of Marine Mammal Distribution Using GIS

Mike Loomis

Operational Oceanography

8 Sep 2008

Introduction

In recent years, the U.S. Navy has come under ever-increasing scrutiny with respect to the effects on marine mammals attributed to its use of active sonar (NRDC, 2008). In response, the Navy has become more sensitive to the potential impacts its activities may have on the environment, and specifically, on marine mammals. Before Navy vessels and aircraft can avoid marine mammals, they must know where the animals are located. Therefore, an understanding of the environmental factors favorable to marine mammal habitat can be of great assistance to determining where marine mammals may be located.

Geographic Information Systems (GIS) are an excellent means of storing and presenting environmental data. Various environmental parameters may be stored in a geodatabase, and the data can be queried and displayed in a variety of ways. The multiple tools within GIS software can be used by researchers to analyze the environmental data.

The following study will use the flexibility of GIS to qualitatively examine environmental and marine mammal data collected during the Summer 2008 Operational Oceanography research cruise. By presenting the data in different ways, using the tools available within the GIS software, the author will seek to determine if the marine mammal sightings collected during the research cruise correspond to any particular environmental parameters. If certain environmental parameters can be identified as areas of high marine mammal density, this information can be used as an additional tool for Navy planners to utilize as part of their marine mammal mitigation procedures.

Methods

All data for the study were collected during the 1-8 August, 2008 Operational Oceanography research cruise. Oceanographic data were collected from the Underway Data Acquisition System (UDAS) aboard the Research Vessel Point Sur. Marine mammal sighting data were collected by a trained observer aboard the ship and stored in an Excel spreadsheet. Only data collected from 4-7 August, corresponding to data collected in the San Nicholas Basin area, were used in this study. No marine mammal sightings were recorded on 8 August; therefore this day was not included in this study.

Data collected from the UDAS were stored in text format. The data was transferred from text files into Excel spreadsheets. After compilation into spreadsheets, the data was imported into an ArcGIS database using the ArcGIS software from ESRI. Because the data collected by the UDAS are stored as discrete ‘points’ of data at precise moments in time, the data were stored as points in the GIS.

The ArcGIS spatial analysis tool ‘cokriging’ was used to develop surface maps of the oceanographic variables analyzed during the study. Put very simply, cokriging is a geostatistical method to create an overlay of a parameter interpolated from discrete data points. The software compares known points to each other and calculates autocorrelation and cross-correlation between them. It then estimates the values of the unknown points and ‘fills them in’ on the overlay. In this manner, the oceanographic parameters of interest in the entire study area were interpolated from the point data collected by the UDAS.

Marine mammal data was stored by the shipboard observer in an Excel spreadsheet. The data contained latitude, longitude, species, and number of animals observed. For the purposes of this study, marine mammal species were collected into eight broad categories. Of the eight categories, only three categories of marine mammals were observed during the study period:

• Beaked whales

• Dolphins

• Whales other than beaked

Beaked whales were categorized distinct from other whale species due to the increased interest provided them by the Navy.

With all of the relevant cruise data imported into the GIS database, various overlays were generated using the tools available in ArcGIS to analyze the data qualitatively. The marine mammal sightings were overlaid on these surface maps. The maps were then visually evaluated to determine what, if any, oceanographic or bathymetric parameters coincided preferentially with marine mammal sightings.

Results

There were 37 recorded marine mammal sightings during the study period. The sightings comprised the following species:

Beaked whales

• Ziphius cavirostris (Cuvier’s Beaked Whale) 1 sighting

Dolphins

• Tursiops truncatus (Common Bottlenose Dolphin) 4 sightings

• Grampus griseus (Risso’s Dolphin) 2 sightings

• Delphinus delphis (Short-beaked Common Dolphin) 6 sightings

• unidentified species 4 sightings

Whales other than beaked

• Balaenoptera musculus (Blue Whale) 4 sightings

• Balaenoptera physalus (Fin Whale) 10 sightings

• Megaptera novaeangliae (Humpback Whale) 1 sighting

• unidentified species 5 sightings

Because the GIS database stores all of the information collected by the UDAS at each collection point, any parameter collected by the UDAS can be used in a GIS overlay. However, for the purposes of this study, three parameters were deemed most relevant as habitat discriminators: bathymetry, flouremetry, and sea surface temperature (SST).

The first overlay shows the locations of marine mammal sightings superimposed on a bathymetry map (figure 1). The overlay suggests the following:

• Whales seem to prefer sloping bathymetry. 16 of 20 sightings occurred on bathymetric slopes, roughly along the 1000’ depth contour.

• Dolphins preferred shallower water than whales. 8 of 16 sightings occurred in water shallower than that in which any whale sightings occurred. If the single whale sighting in the shallow water off the northwest coast of San Clemente Island is discounted, then 12 of 16 dolphin sightings occurred either shallower or coincidentally with whale sightings.

• The single beaked whale sighting occurred along a slope of over 3000’ depth.

The second overlay shows the locations of marine mammal sightings superimposed on a flouremetry map (figure 2). Inspection of the overlay does not suggest any particular preference by any of the marine mammals for any particular flouremetry quantity.

The third overlay shows the locations of marine mammal sightings superimposed on an SST map (figure 3). Inspection of the overlay suggests the following:

• A cool SST band bisects the study area from southwest to northeast across San Clemente Island.

• Marine mammals show some preference for this area with 17 of 37 sightings occurring in this narrow band.

• Marine mammal sightings occurred most frequently in the 17-19°C range, with 25 of 29 sightings within the mapped SST area occurring in this range. Four sightings occurred in water warmer than 19°C.

• Dolphins preferred the coldest water moreso than whales, with four dolphin sightings in water colder than 17°C and no whale sightings in water this cold.

Discussion

There are several obstacles present that make definitive conclusions about marine mammal distribution difficult. First, all of the data presented in this study were collected over a period of four days. The dynamic nature of oceanographic conditions means that any map of conditions over any period of time other than a particular instant will have inaccuracies. Second, the marine mammal sighting locations can necessarily only represent those marine mammals present within visual range of the ship’s track during the study period. The locations cannot be representative of the locations of all marine mammals during the study period. Finally, because the marine mammals could only be sighted visually during daylight hours, the sighting data is not representative of all marine mammals that were present within visual range of the ship during the study period. With these limitations in mind, though, it is still possible to explore how well the data presented fits with our ‘expectations’ of where marine mammals may be present.

Marine mammal sightings, bathymetry and SST

The biology and ecology of both the blue whale and fin whale are poorly understood (Reeves et al., 2002). However a study by NOAA (Moore et al., 2002) provides some useful information that supports inferences made from the data collected during the present study. Moore et al. (2002) found that areas with convergence of boundary currents combined with complex bathymetry can result in eddy formation, which can, “entrain and concentrate zooplankton and thereby attract the blue whales.” There are also studies that indicate that regions of upwelling along the California coast with steep topography can serve to collect and maintain large concentrations of krill, on which the whales feed (Croll et al., 1998, Fiedler et al., 1998).

If we define an SST front on the marine mammal and SST overlay, the data begins to suggest that the whale sightings occur along the front in an area of relatively complex, sloping bathymetry and deep water (figure 4). The anomalous band of cold water running southwest to northeast through the study area could represent an upwelling area, or it could be an artifact of the spatial analysis techniques applied over temporally and spatially varying SST collection points. The southeast side San Nicholas Basin provides upward sloping bathymetry against which deep currents could upwell (figure 5). Certainly, the data are compelling and seem to support findings in other published studies.

Similarly, the sightings of dolphins within the study area compare favorably to published studies of dolphin distribution. Selzer and Payne (1988) found that the common bottlenose dolphin and short-beaked common dolphin prefer deep waters with sloping bathymetry when found offshore. Both species prefer canyons, escarpments and slopes (Bearzi, 2005). Several large pods of both types of dolphins were sighted within the study area and were found in the vicinity of steeply sloping bathymetry. Several studies show that the species of dolphins noted in the current study prefer the sloping bathymetry and largely disregard the depth (Selzer and Payne, 1988; Gaskin, 1992). That seems to be supported by the data in the current study because the dolphins within the study area do not seem to have a particular preference for depth, but do seem to associate more with steeply sloping topography.

The association of dolphins with SST seems to be similar to that of the whales as well, suggesting that they are enjoying better feeding along the upwelling area (Bearzi, 2005; Klinowska et al., 1991). For example, the preferred prey for common bottlenose dolphins is anchovies which are known to concentrate in submarine basins and areas of upwelling (Mais, 1974, Hui, 1979).

Marine mammal sightings and flouremetry

The data do not seem to support any preference by the marine mammals for a particular flouremetry concentration. Flouremetry measures chlorophyll concentration, and therefore, should be a measure of primary productivity. Ideally, then, areas of high flouremetry could show areas of high prey density. However, other factors could be better indicators of prey density locations, such as areas of upwelling (Moore et al., 2005). In their long term study of blue whale locations and chlorophyll-a density, Moore et al. (2005) didn’t find any significant correlation between the two. Ultimately little is gained in the current study from examining flouremetry; however, the data resides in the database in the event that further exploration of this parameter is warranted in the future.

Conclusion

It is difficult to draw significant conclusions regarding marine mammal distribution in relation to environmental parameters based upon four days of data. Yet, even this small study verifies findings put forth in various previously cited papers. Beyond simply verifying data, though, by compiling the data within a GIS architecture, the database remains eminently available to be built upon, adding data from other sensors or research cruises into it. Rather than envisioning the current study as a means to an end, it is a data point upon which future research can be built, leveraging the utility of storing data in an easily built-upon and accessible format.

If such a database were compiled using historical and current marine mammal sightings, as well as all measured oceanographic variables like those used in this study, much more robust, statistically valid conclusions could be drawn. This is, of course, beyond the scope of what is possible in a short project for a single class. But, this study hopefully demonstrates what is possible to show with even a small dataset. Extrapolate this study into years of data and its benefits to organizations concerned about where marine mammals tend to concentrate, like the U.S. Navy, becomes self-evident.

Areas of future research

There are several additional areas of future research that suggest themselves after this study:

• Compile additional marine mammal sightings from previous and future research cruises into the existing GIS, along with UDAS data. Use this new database to draw more robust, statistically significant findings.

• Compile UDAS data and data from other sensors, like satellite data, into a GIS and compare them to each other to determine the accuracy of remotely sensed data.

• Use statistical analysis and spatial analysis tools available within ArcGIS software on the current or a future GIS database to present the data in a different manner (e.g. using graphs, area comparisons, correlation equations, etc.).

Appendix: Figures

[pic]

Figure 1, marine mammal sightings on bathymetry map

[pic]

Figure 2, marine mammal sightings with flouremetry map

[pic]

Figure 3, marine mammal sightings with SST

[pic]

Figure 4, SST front

[pic]

Figure 5, upwelling area

References

Bearzi, M. 2005. Habitat Partitioning by Three Species of Dolphins and Santa Monica Bay, California. Bull. Southern California Acad. Sci., 104(3):113-124.

Croll, D.A., B.R. Tershy, R. Hewitt, D. Demer, S. Hayes, P. Fiedler, J. Popp and V.L. Lopez. 1998. An Integrated Approach to the Foraging Ecology of Marine Birds and Mammals. Deep-Sea Res. II, 45:1353-1371.

Fiedler, P.C., S.B. Reilly, R.P. Hewitt, D. Demer, V.A. Philbrick, S. Smith, W. Armstrong, D.A. Croll, B.R. Tershy and B.R. Mate. 1998. Blue Whale Habitat and Prey in the California Channel Islands. Deep-Sea Res. II., 45:1781-1801.

Gaskin, D. E. 1992. Status of the Common Dolphin, Delphinus delphis, in Canada. Can. Field-Nat., 106:55-63.

Hui C.A. 1979. Undersea Topography and Distribution of Dolphins of the Genus Delphinus in the Southern California Bight. J. Mammal., 60(3):521-527.

Klinowska, M., and J. Cooke. 1991. Dolphins, Porpoises and Whales of the World: The IUCN Red Data Book. IUCN: Gland, Switzerland.

Mais, F. 1974. Pelagic fish surveys in the California Current. California Department of Fish and Game, Fish Bull., 162:1-79.

Moore, S.E., W.A. Watkins, M.A. Daher, J.R. Davies and M.E. Dahlheim. 2002. Blue Whale Habitat Associations in the Pacific: Analysis of Remotely-Sensed Data Using a Geographic Information System. Oceanography, 15(3):19-25.

Natural Resources Defense Council. 2008. Protecting Whales from Dangerous Sonar. Available: wildlife/marine/sonar.asp

Reeves, R.R., B.S. Steward, P.J. Clapham and J.A. Powell. 2002. Marine Mammals of the World. New York. Chanticleer Press, Inc.

Selzer, L. A. and P. M. Payne. 1988. The Distribution of White-sided (Lagenorhychus acutus) and Common Dolphins (Delphinus delphis) vs. Environmental Features of the Continental Shelf of the Northeastern United States. Mar. Mamm. Sci., 49(2): 141-153.

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download