FINAL REPORT

REMOTE SENSING ANALYSIS OF BAT SPECIES RICHNESS AND ABUNDANCE IN CACHE VALLEY AND THE BEAR RIVER RANGE

BRAD LENGAS and DAN ROBERTS

Merlin D. Tuttle

INTRODUCTION

Nine species of bats occur in the Bear River Range of northern Utah in a wide variety of habitats. State and federal agencies are becoming increasingly concerned that current management plans do not specifically address bat species habitat requirements.
Defining bat habitat is not an easy task. Bats are highly mobile, often moving several kilometers daily between foraging and roosting habitat. The area situated between these two types of habitat may be considered the "commuting" habitat. In addition to these extensive habitat requirements, bat habitat is three-dimensional.
Bats select roosts based on the shelter provided by the physical features: rock outcrops and overhangs, hollow trees, bark and rock crevices, caves, mines, and buildings. Foraging habitat is generally found along riparian corridors and edge habitats where high numbers of insects occur. Water is an essential component of the foraging habitat, as all bats come to still pools or eddies to drink.
Bat species richness patterns have been shown to be related on a microgeographic scale to insect densities, the presence of freestanding water, and the degree of urbanization (McCoy and Conner 1980, de Jong and Ahlen 1991, Kurta and Teramino 1992, Rydell 1992). Habitat preferences for bats are difficult to determine, due to their high degree of mobility as mentioned above. Nonetheless, certain habitat types, such as forested areas, may be preferred by bats over other areas.
The objective of this study was to search for patterns of bat species richness and abundance as they relate to the TM spectral data and the GAP vegetation classification, based on mist netting data for nine bat species captured at 14 sites in the Bear River Range and Cache Valley between 1992 and 1994.

BAT SPECIES RICHNESS AND ABUNDANCE DATA AT 14 STUDY SITES (1992-1994)

SITE SPECIES 1 2 3 4 5 6 7 8 9 10 11 12 13 14 TOTAL Little Brown 1 4 20 36 14 0 10 5 0 2 1 10 7 5 115 Myotis Silver-haired 16 4 26 9 2 7 1 3 0 0 0 2 4 3 77 Bat Long-legged 1 0 1 21 0 0 0 1 1 0 0 10 1 6 42 Myotis Big Brown 2 0 2 8 0 5 10 1 2 0 0 3 8 0 41 Bat Long-eared 0 0 0 18 1 0 0 2 2 0 0 1 2 10 36 Myotis Hoary Bat 1 3 4 0 4 0 12 3 1 0 0 0 0 1 29 Western Small- 0 0 0 0 1 0 0 2 0 0 0 1 1 0 5 footed Myotis Fringed 0 0 0 2 0 0 0 0 0 0 0 1 1 0 4 Myotis Townsend's 0 0 0 0 0 1 0 0 1 0 0 0 0 0 2 Big-eared Bat TOTAL 21 11 53 94 22 13 33 17 7 2 1 28 24 25 351

THE STUDY AREA

The study area encompassed the following nine quadrads in the northeast section of Cache County:
Richmond, Naomi Peak, Tony Grove, Smithfield, Mount Elmer, Temple Peak, Logan, Logan Peak, Boulder Mountain
Below is the GAP Vegetation image for the nine-quad study area.

BATS OF THE STUDY AREA

J. Scott Altenbach
Little Brown Myotis (Myotis lucifugus)
The Little Brown Myotis occurs in forested areas up to elevations of 3300 m. It is also associated with humans and can be found in towns and cities. It roosts in rock crevices, tree hollows, crevices under bark, and in caves, mines, and buildings (Armstrong et. al, 1994).

J. Scott Altenbach
Long-eared Myotis (Myotis evotis)
The Long-eared Myotis occurs mainly in coniferous forests at moderate to high elevations. It roosts most often under loose bark, and also in caves, mines, and buildings (Armstrong et. al, 1994).

J. Scott Altenbach
Long-legged Myotis (Myotis volans)
The Long-legged Myotis occurs mainly in forested mountain areas at moderate elevations. It roosts under loose bark, in rock crevices, and in caves, mines, and buildings (Armstrong et. al, 1994).

J. Scott Altenbach
Fringed Myotis (Myotis thysanodes)
The Fringed Myotis occurs mainly in coniferous forests at moderate to high elevations. It roosts in rock crevices, trees, caves, mines, and buildings (Armstrong et. al, 1994).

J. Scott Altenbach
Western Small-footed Myotis (Myotis ciliolabrum)
The Western Small-footed Myotis occurs in forested canyon areas, typically at lower elevations. It is most often found roosting in rock crevices, but this species also roosts under loose bark, and in caves, mines, and buildings (Armstrong et. al, 1994).

Merlin D. Tuttle
Big Brown Bat (Eptesicus fuscus)
The Big Brown Bat has a wide distribution and is a commensal with humans, often roosting in attics, barns and bridges. It also roosts in hollow trees, in rock crevices, caves, and mines (Armstrong et. al, 1994).

Merlin D. Tuttle
Silver-haired Bat (Lasionycteris noctivagans)
The Silver-haired Bat occurs mainly in forested areas at a wide range of elevations. It roosts under loose bark and in other tree crevices, in hollow trees, in open sheds and garages, and sometimes in soft-walled caves and mines (Armstrong et. al, 1994).

Randy Babb
Hoary Bat (Lasiurus cinereus)
The Hoary Bat is a solitary tree-roosting species that occurs in forested areas at a wide range of elevations. It roosts by hanging 3-5 m above the ground in the foliage of trees (Armstrong et. al, 1994).

Merlin D. Tuttle
Townsend's Big-eared Bat (Corynorhinus townsendii)
The Townsend's Big-eared Bat occurs in a wide variety of habitats up to 2860 m. This species is most commonly found roosting in caves and abandoned mines, although it sometimes roosts in buildings (Armstrong et. al, 1994).

METHODS
INITIAL DATA
Bat Species Richness and Abundance Data from 14 Sampling Sites
UTM Coordinates for the Bat Sampling Sites From GPS

IMAGES AND COVERAGES USED FOR THE STUDY AREA
Landsat TM Image of Cache County
30 Meter GAP Vegetation Map of the 1:100,000 Logan Quadrangle
Water Courses and Bodies of the 1:100,000 Logan Quad
7.5' 30 Meter DEM Data for Nine Quads:
Richmond, Naomi Peak, Tony Grove Smithfield, Mt. Elmer, Temple Peak Logan, Logan Peak, Boulder Mtn.

SOFTWARE
ARC/INFO 7.0
ERDAS IMAGINE 8.2
SAS
CRICKET GRAPH
QUATTRO PRO
CUSTOM C PROGRAMS

COVERAGES AND IMAGES CREATED FOR ANALYSIS
BAT SAMPLING SITES
BUFFERS
990 Meter (33 Pixel) Buffer
540 Meter (18 Pixel) Buffer
210 Meter ( 7 Pixel) Buffer

MASKS
from 990 Meter Buffer
from 540 Meter Buffer
from 210 Meter Buffer
TM Data for the 14 Buffered Bat Sample Sites
from 990 Meter Buffer
from 540 Meter Buffer
from 210 Meter Buffer

GAP Vegetation Data for the 14 Buffered Bat Sample Sites
from 990 Meter Buffer
from 540 Meter Buffer
from 210 Meter Buffer

AOIs for Each of the 14 Bat Sample Sites
from 990 M BUFFER
from 540 M BUFFER
from 210 M BUFFER

METHODS FOR BUILDING THE COVERAGES
All Raster Files (IMAGINE) - were Georeferenced to UTM, Spheroid Clarke 1866, Zone 12, Datum NAD27, with 30 Meter pixel sizes.
The Bat Sampling Site Coverage was created in ARC/INFO with the GENERATE command from a text file with the UTM coordinates. After BUILDing the coverage, POINTGRID was used to convert from a Point coverage to a Grid. The Grid was then Imported into IMAGINE as an image.
BUFFERs of the Bat Sites Image were created using the INTERPRETER, GIS ANALYSIS, and SEARCH Option in IMAGINE. Patterns of bat species richness and abundance (as they relate to the TM spectral data and the GAP vegetation classification) was assumed to be a function of the scale that most closely approximates the bat's usage of the habitat. Three different scales (990 m:large, 540 m:intermediate, and 210 m:small) were choosen to determine which scale was most appropriate for bats.
MASK files were created using the INTERPRETER, GIS ANALYSIS, and RECODE Option in Imagine. Areas that were to be retained were Coded as 1, and all other areas to be removed were Coded as 0. MASK files were created for each of the three Buffers around the Bat Sampling Sites.
The TM IMAGE and the GAP VEGETATION IMAGE were MASKed with the MASK files to create files that contained data only within the Buffered areas around the Bat Sampling Sites. The MASK operation was performed using the INTERPRETER, UTILITIES, and MASK Option.

STATISTICAL ANALYSIS
I. Since Bat Species and Richness differed among Sampling Sites, it was necessary to know if the reflectance values of the Sampling Sites also differed.
ANALYSIS OF VARIANCE (ANOVA) (Sokal and Rohlf 1981)
1. SAMPLING - Used an AOI for each Bat Sampling Site for the TM and GAP Images
2. The UTILITY, PIXEL TO TABLE Option - To create a TEXT File of the Pixel Values
3. C PROGRAM - To remove the UTM Data and Pixels with Zeros for all 6 TM Bands
4. FTP TEXT Files to VAX Account
5. SAS - PROC GLM Option (SAS Institute 1985)

II. Was the Variance of each TM Band at each Bat Sampling Site Correlated with Bat Species Richness?
1. Variances from the SAS PROC GLM and Bat Species Richness were entered into the VAX Account
2. SAS PROC CORR (SAS Institute 1985)(Sokal and Rohlf 1981)

III. Although the reflectance values of the Sampling Sites differed, we were interested in knowing if any of the sites were similar.
SEPARABILITY INDEX
1. The CLASSIFIER, SIGNATURE EDITOR Option - to select each Buffered Bat Sampling Site with varying Euclidean Distances to capture the entire Buffer.
2. The EVALUATE, SEPERABILITY, TRANSFORMED DIVERGENCE, REPORT Option - to create a TEXT file of the SEPERABILITY
CLUSTERING
1. Enter SEPERABILITY data into VAX Account
2. SAS PROC CLUS OPTION - CENTROID METHOD (SAS Institute 1985)

IV. In order to determine which Buffer size most closely approximates the bat's usage of the habitat, we needed to know which buffer gave the best association between TM reflectance values and Bat Species Richness.
CHI SQUARE (Sokal and Rohlf 1981)
1. TEXT files for TM Reflectance Values in Vax Account
2. Enter Data for Bat Species Richness into Vax Account
3. SAS PROC FREQ Option (SAS Institute 1985)

V. Are there relationships between the reflectance values of the TM DATA and Bat Species Richness or Abundance?
VI. Are there relationships between the GAP CLASS Richness and Bat Species Richness or Abundance?
LINEAR and POLYNOMIAL REGRESSION (Sokal and Rohlf 1981)
1. TEXT files for TM Reflectance Values in Vax Account
2. TEXT files for Bat Species Richness in Vax Account
3. Enter TEXT files for GAP CLASS Richness into VAX Account
4. SAS PROC REG OPTION - for p values (SAS Institute 1985)
5. CRICKET GRAPH - for graphic output and to compute and draw the Regression Equations

VII. Are there relationships between the GAP CLASS Diversity and Bat Species Diversity?
DIVERSITY INDICES (1/Simpson's D - Hill 1973)
1. Import data for GAP CLASS AND Bat Species Richness into QUATTRO PRO
2. Enter Formula for SIMPSON's Diversity Index - (Sum of all proportions squared)
3. Export Diversity Indices to TEXT files
4. FTP TEXT files for GAP CLASS Diversity and Bat Species Diversity into Vax Account
5. SAS PROC REG OPTION - for p values (SAS Institute 1985)
6. CRICKET GRAPH - for graphic output and to compute and draw the Regression Equations

VIII. Could the combinations of Bat Species found at the Sampling Sites be random?
NULL MODEL (Conner & Simberloff 1979)
1. Computation of Probabilities :

POSSIBILITIES 9 species 1 at a time 9 9 species 2 at a time 36 9 species 3 at a time 84 9 species 4 at a time 126 9 species 5 at a time 126 9 species 6 at a time 84 9 species 7 at a time 36 9 species 8 at a time 9 9 species 9 at a time 1 TOTAL 511

RESULTS

TABLE 1. RESULTS OF ANOVA AND CORRELATION FOR THE THREE BUFFER SIZES (210 m, 540 m, 990 m)
_____________________________________________________________________________________________________________________ BUFFER SIZE ANOVA OF SPECTRAL REFLECTANCES CORRELATION BETWEEN VARIANCES OF SPECTRAL REFLECTANCES FOR FOURTEEN SITES FOR EACH SITE AND BAT SPECIES RICHNESS _____________________________________________________________________________________________________________________ 210 m p < 0.0001 NONE 540 m p < 0.0001 NONE 990 m p < 0.0001 NONE _____________________________________________________________________________________________________________________

TABLE 2. RESULTS OF SITE CLUSTERING AND CHI-SQUARE TESTS OF GOODNESS OF FIT FOR SITE CLUSTERS AND BAT SPECIES RICHNESS

TABLE 2a. BEST CLUSTER FOR 210 METER BUFFER _______________________________________________________________________________________________ SITES CLUSTER BAT RICHNESS DF CHI-SQUARE p _______________________________________________________________________________________________ 1 1 5 2,3 2 3,5 4 3 6 5,6 4 5,3 7 5 4 54 69.0 0.08 8,9,12 6 7,5,7 10 7 1 11 8 1 13 9 7 14 10 5 _______________________________________________________________________________________________

TABLE 2b. BEST CLUSTER FOR 540 METER BUFFER _______________________________________________________________________________________________ SITES CLUSTER BAT RICHNESS DF CHI-SQUARE P _______________________________________________________________________________________________ 1,2,3,5,6,7 1 5,3,5,5,3,4 4,14 2 6,5 8,9 3 7,5 24 36.06 0.05 12,13 4 7,7 10,11 5 1,1 _______________________________________________________________________________________________

TABLE 2c. BEST CLUSTER FOR 990 METER BUFFER _______________________________________________________________________________________________ SITES CLUSTER BAT RICHNESS DF CHI-SQUARE P _______________________________________________________________________________________________ 1 1 5 2,3 2 3,5 4,14 3 6,5 5,6,7 4 5,3,4 36 44.07 0.167 8,9 5 7,5 10,11 6 1,1 12,13 7 7,7 _______________________________________________________________________________________________

RESULTS OF REGRESSION ANALYSES

FIGURE 1.

FIGURE 2.

FIGURE 3.

FIGURE 4.

FIGURE 5.

FIGURE 6.

FIGURE 7.

FIGURE 8.

NULL MODEL
______________________________________________________________________________________________________________________ SITES SPECIES COMBINATIONS PROBABIILITY SB BB HB LM BM EM FM SM TB OBSERVED EXPECTED _______________________________________________________________________________________________________________________ 1,3 X X X X X 0.143 0.002 _______________________________________________________________________________________________________________________ 10,11 X 0.143 0.002 _______________________________________________________________________________________________________________________ 12,13 X X X X X X X 0.143 0.002 _______________________________________________________________________________________________________________________ 2 X X X 0.071 0.002 _______________________________________________________________________________________________________________________ 4 X X X X X X 0.071 0.002 _______________________________________________________________________________________________________________________ 5 X X X X X 0.071 0.002 _______________________________________________________________________________________________________________________ 6 X X X 0.071 0.002 _______________________________________________________________________________________________________________________ 7 X X X X 0.071 0.002 _______________________________________________________________________________________________________________________ 8 X X X X X X X 0.071 0.002 _______________________________________________________________________________________________________________________ 9 X X X X X 0.071 0.002 _______________________________________________________________________________________________________________________ 14 X X X X X 0.071 0.002 _______________________________________________________________________________________________________________________

REFERENCES
Armstrong, D.M., R.A. Adams, and J. Freeman. 1994. Distribution and ecology of bats of Colorado. Natural History Inventory of Colorado No. 15: 1-84.
Conner, E.F., and D. Simberloff. 1979. The assembly of species communities: chance or competition? Ecology. 60:1132-1140.
Hill, M.O. 1973. Diversity and evenness: a unifying notation and its consequences. Ecology 54: 427-432.
Jong, J., de, and I. Ahlen. 1991. Factors affecting the distribution patterns of bats in Uppland, central Sweden. Holarctic Ecology 14: 92-96.
Kurta, A., and J. A. Teramino. 1992. Bat community structure in an urban park. Ecography 15: 257-261.
McCoy, E.D., and E.F. Conner. 1980. Latitudinal gradients in the species diversity of North American mammals. Evolution 34: 193-203.
Rydell, J. 1992. Occurrence of bats in northernmost Sweden (65*N) and their feeding ecology in summer. Journal of Zoology, London 227: 517-529.
SAS Institute, Inc. Cary, North Carolina, 1985.
Sokal, R.R. and F.J. Rohlf. 1981. Biometry, Second Edition. W.H. Freeman and Company, San Francisco, California, U.S.A.