CE 547, Spring 2004

Class Project

Rick Winslow

 

Modeling locations of black bear depredation activity in New Mexico

(back to home)

 

 

 

Background:

 

During the Spring Legislative session of 2003 the lawmakers of the State of New

Mexico (State) amended Chapter 17 of the New Mexico Annotated Code (NMAC).  This section of the State’s laws pertains specifically to the authority and laws of the fish and wildlife of the State and the New Mexico Department of Game and Fish (Department).  The amendment was enacted on 17-1-14 NMAC which gives the State Game Commission the authority to manage fish and wildlife of the State for its citizens.  The amendment specifically added to section B) the 16th clause, “…shall have the authority to designate areas, public and private, which shall be required to have bear-proof garbage containers.”

            The passage of this legislation has created a need to determine which areas require bear-proofing.  The most effective way of identifying these areas is based upon a number of factors, including:  Conservation Officer’s records and perceptions of “problem” areas, depredation location records, nuisance bear captures, and bear habitat.  It turns out that the most effective way of relating these very different data sources is through the use of spatial analysis.  The records and perceptions of the Conservation Officers will be used to filter the products of the GIS analysis. 

            Black bears (Ursus americanus) are common large carnivores throughout much of the United States.  They are opportunistic omnivores which means they readily feed on most sources of nutrients.  They occur at relatively low densities through most of their range but are important and typically visible members of the natural community in most areas where they occur.  Because of their opportunistic nature black bears become habituated to human-related food sources easily and can pose a danger to the human community and to themselves in these situations.  (because of the combination of human fear and intrigue with these animals, ) Bear management is often controversial and is always an important item on the Public’s agenda.

Objective:

            This project is intended to create a predictive model of communities or areas where the bear-proof garbage container law should be implemented.  It is/will be based upon past depredation records, bear captures, bear habitat, game management units (GMUs) and communities of the State.

Methods

Data Sources:

New Mexico Department of Game and Fish, GMU shapefile and bear habitat model (Appendix 1).

RGIS, roads and communities information (Appendix 2).

ArcMap templates, State of New Mexico and surrounding states (Appendix 3).

EDAC, assistance with projecting UTM Zone 12 data points in Arcview (Appendix 4).

Projection used:         

NAD 83 UTM Zone 13.  This projection was picked because the locational data used is in UTMs in NAD 83 as is the raster image from the bear habitat model.

 

 

Software Environment:

I used Arcgis software to perform all the work that I have accomplished so far.  I did rely upon the assistance of EDAC as mentioned above to project the UTM Zone 12 bear capture locations in the Zone 13 raster of the bear model.  Actual UTM locations were recorded in Access or Excel and converted to dBASE 4 then projected in Arcmap as XY data. 

Analysis:

The primary analysis techniques used so far have been with the select by location, select by attribute, and the geoprocessing wizard within Arcmap.  It has been a visual and spatial examination of how bear habitat relates to human communities (habitat), road systems, and how and where bear nuisance activities occur within the human dominated portion of the landscape.

Results:

            I have used the model to identify communities that occur near or within bear habitat (Appendix 5).  This identification process requires different levels of scaling, i.e. communities within 5 or 10 kilometers of potential bear habitat.  It is also very effective at visually organizing the different data sets that are parts of the model.  I have created different layers incorporating bear capture locations, communities at various scales, roads, bear habitat by importance (3 levels of use) and bear habitat spatially, I have also use the buffer wizard to try to buffer the towns and habitat in different ways. 

Conclusion

            The model is successful at identifying the towns/areas where the bear-proof garbage container law may need implementation at some later date.  Those identified have been verified by discussion with the local Conservation Officers and through records of where bear depredation and nuisance bear captures have occurred in the past.

The problem at this point is the actual implementation of the law and the acquisition/purchase of bear-proof garbage containers by the towns, which will incur a significant financial hardship, or by the consumer/trash producer citizens who will eventually “bear” the burden of increased bills for refuse removal.

Future work:

            The model still requires significant work to make it as useful and as descriptive as I would like.  I need to add all the actual depredation records since 2000 (a somewhat arbitrary date but it closely corresponds to the time “good” records of depredation have been kept!); this is approximately 800-1000 UTM locations.  Through this I will have a very good idea of where potential problems occur.  I also want to be able to determine, by GMU, how much bear habitat there is versus habitat within 5 and 10 kilometers of communities (realizing that a bear can smell a pile of apples from at least 5 kilometers and then travel that distance in no time).  The eventual goal is to intersect buffers of the communities with bear habitat, identify target communities and determine the amount of bear habitat that lies within a given distance of any community.

 

 

 

 

 

 

Appendices:

Procedure: 

Step one:  Gather information, bear habitat model (from Costello et al. 2001), GMU shapefile, states shapefile, roads and communities shapefiles.

Step two:  Turn locational bear capture data into usable format.  The information was in UTMs NAD 83 (maybe, the odds are that is wasn’t, but the differences between NAD 83 and NAD 27 are not that significant for a large animal), also it was in the approximate format 12-34.5-447.2, which roughly corresponds to NAD 83 Zone 12 345000e 4472000n.  All of these records from 2000 to present (423 total) had to be converted to a proper UTM coordinate.  Some were unusable, some were duplicated and therefore redundant, some were in Alaska…etc.

Step three:  Import bear model raster and start adding shapefiles to get the intended results (Appendix 6).  Go see EDAC and get them to project the NAD 83 Zone 12 data points into a Zone 12 raster (is this cheating?) using Arcview (Appendix 7).  Use select by attribute and/or select by location to pare down the cities shapefile to cities with greater than 300 inhabitants (Appendix 8).  Create buffers around selected communities and find this to be less than useful (Appendix 9).  Finally use select by location to select communities within 5 and 10 kilometers of bear habitat to identify communities where the action may be required (Appendix 5 shows cities located within 10 kilometers of bear habitat).  Then, finally, use select by location to pick out cities within 10 kilometers of bear habitat, make a new map with those cities, add the other cities of population greater than 300 in and buffer the cities within 10 kilometers of bear habitat with 5 and 10 kilometer ring buffers (Appendix 10).  Pheww!