|
|
|
GIS & Data Integration
|
Implementation of A Geographic Information System (GIS) to Determine Wildlife Habitat Quality Using Habitat Suitability Index
YuChing Lai
Post Doctor, Division of Forest Management
Taiwan Forestry Research Institute
53 Nan-Hai Road, Taipei 100, TAIWAN
Walter L. Mills
Associate Professor of
Forestry,
Department of Forestry and Natural Resource
Purdue University, 1159 Forestry Building, Room 110
West Lafayette, IN 47907-1159
Chi-Chuan Cheng
Deputy Director, Taiwan Forestry Research Institute
Council of Agriculture, 53 Nanhai Rd., Taipei 100,TAIWAN
Key Words: Geographic Information System (GIS), Habitat Suitability Index (HSI), Wildlife Habitat, Spatial Pattern
Abstract: Habitat quality for many wildlife populations has a spatial component related to the arrangement of habitat elements across large geographic areas. There are a number of indices used to quantitatively describe the components of habitat assessment, yet, only a few of them incorporating spatial relationship in the model. Among these mathematical indices, habitat Suitability Index (HSI) models have been widely developed for use in habitat evaluation procedures (HEP). HEP, which is based on the assumption that habitat for a selected wildlife species can be described by a Habitat Suitability Index, can be used to document the quality and quantity of available habitat for a specific wildlife species. However, because of the difficult of quantifying spatial arrangement, HSI has ignored the spatial distribution of habitats. Geographic information system (GIS) provides wildlife managers and planners with techniques that can help overcome some of the problems inherent in developing, applying, and evaluating practical, spatially explicit habitat models. Given the desirability of deterministic results, seamless integration of data and analytical options available via GIS increases our ability to state, implement, test, and evaluate estimated or modeled habitat elements. This study examined possible ways to use the accumulated knowledge found in HSI to estimate the wildlife microhabitat quality across a landscape. In this study, GIS is used to generate parameters of HSI models, especially the spatial habitat parameters that are often of explicit importance for HSI models, incorporating spatial reasoning and constraints into HSI models, analyzing spatial patterns of habitat, and providing the mapping capability for converting stand-based information into maps of habitat. Species that occur in Missouri Ozark Forest Ecosystem with existing HSI models are selected. A moving window the size of the home ranges of each species is applied to calculate the average value of each life requisites. In conclusion, implementing GIS in wildlife habitat assessment improves the use of HSI models by automating tasks, identifying areas where site-specific analyses were needed, and reducing the spatial and temporal complexity involved with integrating different resource perspectives.
1. Introduction
Ecosystem management appears to be the wave of the future. General program goals such as "maintain biodiversity", "maintain viable populations of all native species", and "protect representative natural communities" are often emphasized in ecosystem management (USDA 1994, Grumbine 1994, and Irland 1994). Tools for maintaining biodiversity and viable species populations are likely to be focused on providing habitats in an appropriate spatial and temporal arrangement. Therefore, vegetation management is a major tool for maintaining and restoring biodiversity and to achieve delisting or to avoid listing of threatened and endangered species (USDA 1994). There have been numerous attempts to assess habitat quality for particular species. Probably the most widely practiced in the United States is the Habitat Suitability Index (HSI), which together with the Habitat Evaluation Procedures (HEP) developed by the U.S. Fish and Wildlife Service, has frequently been used to assess habitat. It is also a useful monitoring tool for biodiversity at community-ecosystem level (Noss and Cooperrider 1994). However, it ignores the effects of immigration and spatial distribution of habitats (Carroll and Meffe 1994).
In this study, the possibility of using HSI for ecosystem management assessment will be tested. Species with existing HSI models will be use. Since the same procedure can be applied regardless of what species is used, more species could be used when, and if, appropriate HSI are developed. Point sample data combined with spatial data of the Missouri Ozark Forest Ecosystem Project (MOFEP) will be used to assess ecological organization at the community-ecosystem level in this study.
2. Methodology
Study Area
The Missouri Ozark Forest was located in southeastern Missouri Ozarks consists of 9,200 acres of mature upland oak-hickory and oak-pine forest communities. Collectively, these counties are 84% forested with large contiguous blocks of forests separated only by roads and streams. Compartment One of Missouri Ozark Forest Ecosystem Project (MOFEP) was selected for this study because it has the most detailed inventory data. It is 380 hectares (989 acres) in size and contained 62 stands. Seventy-six permanent sample plots are distributed among these stands. Dominant overstory species included pine (Pinus echinata), hickory , and oak.
A cluster plot design was used to collect data in order to investigate the effects of forest management on the composition and spatial distribution of woody and herbaceous vegetation. In this design, a 0.2 ha (0.49 ac.) circular plot was used to sample trees 11.4 cm (4.5 in) dbh and larger and to tally the total number of den trees. Four line intercepted 17.2 m (56.4 feet) in length were located within each 0.2 ha plot to measure the coverage of down dead woody material. Four circular 0.02 ha (0.05 ac.) plots were located within the larger plot to sample woody plants between 3.5 cm (1.5 in.) and 11.2 cm (4.4 in.) dbh. One 0.004 ha (0.01 ac) plot was placed within each 0.02 ha plot, sharing the same plot center, to sample vegetation taller than 1.0 m (3.3 feet) and less than 3.5 cm (1.5 inches) dbh. Four one square m (10.8 square feet) plots were located within each 0.02 ha plot to sample vegetation less than 1.0 m (3.3 feet) in height.
2.2 HSI Models selection
While hundreds of habitat suitability index models were available, the selected HSI represent those that were valid for the Missouri Ozark Highland and for which data was available. Eleven species were selected for this study including seven bird species and four mammals. The bird species were selected when either their breeding range or year-round range include the Missouri Ozark highland. They were barred owl (Strix varia) (Allen 1987), brown thrasher (Toxostoma rufum) (Cade 1986), downy woodpecker (Picoides pubescens) (Schroeder 1982), hairy woodpecker (Picoides villosus) (Sousa 1987), pileated woodpecker (Dryocopus pileatus) (Schroeder 1982), eastern wild turkey (Meleagris gallopavo sylvestris) (Schroeder 1985), and northern bobwhite (Colinus virginianus) (Schroeder 1985). Four mammal HSI models were selected, i.e., Bobcat (Felis rufus) (Boyle and Fendley 1987, Schwartz and Schwartz 1981), eastern cottontail (Sylvilagus floridanus) (Allen 1984, Schwartz and Schwartz 1981), gray squirrel (Sciurus carolinensis) (Allen 1982, Schwartz and Schwartz 1981), and fox squirrel (Sciurus niger) (Allen 1982, Schwartz and Schwartz 1981).
In most of these wildlife HSI models, more than one life requisite (for example, food, cover, and reproduction) was usually used (Table 1). Each of the life requisite was described by a set of habitat variables (Table 1).
Table 1. Attributes of selected specie
Attributes
Species |
Home Range(ha) |
Life Requisities |
Variables |
|
BARRED OWL |
228.6 |
Reproduction |
# of trees >51 cm dbh, mean dbh, Canopy cover |
|
BOBCAT |
100 |
Food |
% by grass/forb-shrub |
|
BROWN THRASHER |
1 |
Food/cover/reproduction |
Density, Canopy cover, Litter cover |
|
DOWNY WOODPECKER |
4 |
Food/reproduction |
Basal area,# of snags >15cm dbh |
|
EASTERN COTTONTAIL |
4 |
Cover/ |
Herbaceous cover, Shrub cover, Canopy cover |
|
FOX SQUIRREL |
3.55 |
Food/cover |
Canopy closure of hard mast trees, Distance to grain, mean dbh
canopy cover, shrub cover |
|
GRAY SQUIRREL |
0.4 |
Food/cover |
Canopy closure of mast trees, Distance of hard mast trees, canopy cover, mean dbh, shrub cover |
|
HAIRY WOODPECKER |
4 |
Cover/reproduction |
Mean dbh, Canopy cover, Canopy closure of pine, # snags >25 cm dbh, mean dbh |
|
N BOBWHITE |
4.9 |
Food/cover/nesting |
Canopy cover of preferred herbaceous plants, Litter cover, # pine or oak, canopy cover, herbaceous cover, mean hight of herbaceous canopy, grass herbaceous canopy, soil moisture regime |
|
PILEATED WOODPECKER |
136 |
Food/cover/reproduction |
Canopy closure, # tree stumps, # tree >51 cm dbh, # snags >38 cm dbh, mean dbo of snags >38 cm dbh |
|
TURKEY |
29 |
Food/cover |
Herbaceous cover, Mean hight of herbaceous canopy, Mean dbh of hard mast producing trees, # hard mast-producing trees, canopy coverage of soft mast-produciing trees, shrub cover, shrub cover comprised of soft mast producing shrub, canopy closure |
|
|
|
|
|
|
|
