| GISdevelopment.net ---> Application ---> Environment |
Using GIS for carnivores distribution mapping in fragmented landscape
Yu-Ching Lai, Kurtis Jai-Chyi Pei, and Kai Yuen Suen
Yu-Ching Lai, Kurtis Jai-Chyi Pei, and Kai Yuen Suen
Introduction
For wildlife management, the ability to model spatial distribution and changes in distribution of wildlife is of considerable importance. Once spatial distribution can be adequately modeled, the distribution and abundance can be monitored efficiently over time, and future changes can be predicted. These spatial characteristics and relationship are often difficult to identify and hard to display with traditional ground survey or statistical model. Therefore, using Geographic Information System (GIS) has become an evitable trend in ecology studies and developing wildlife habitat model (Rickers et al. 1995). The characteristics of spatial analysis and spatial display of GIS not only provides efficient way of data handling, storage, and analyzing, more importantly, it also enables mapping of wildlife distribution, identification of patterns, and generation of habitat spatial characteristics, hence, a useful tool in decision making for conservation and wildlife management (Scott et al.1992, Aspinall and Veitch 1993, Buckland and Elston 1993, Long et al. 1993, Ji-Wei and Clinton 2000, Lenton et al. 2000).
Auto-trigger camera has been used successfully in mammal survey worldwide to avoid possible inconsistency induced by weather, different investigators, limiting detective abilities of researchers at night, and etc. (Gysel and Davis 1956, Dodge and Snyder 1960, Carter and Slater 1991, Jones and Raphael 1993, Kucera and Barrette 1993, Pei et al. 1997). It has also been used in wildlife study to generate their activity patterns (Pearson 1960, Carley et al. 1970, Pei 1995, 1998), spatial distribution patterns (Pei et al. 1997, Pei 1999), and to estimate population density (Karanth 1995, Karanth and Nichols 1998, McCullough et al. 2000, Pei 2001b). It has the advantages of recording secretive animals, especially those do not leave prominent signs, and collecting data day and night in a more standardized and consistent way than most traditional methods (e.g., trapping, sign-searching, transect line, etc.)
In this study, the applicability of using the existed GIS technique coupled with field survey data from auto-triggered cameras were assessed for carnivore distribution mapping in fragmented landscape of Hong Kong Special Administration Region (the Hong Kong SAR).
Material and Methods
Auto-trigger cameras were used to conduct carnivores survey from October 2001 to March 2002. A total of 100 auto-trigger cameras were distributed randomly in 16 country parks and 1 special area of Hong Kong SAR according to the size of the country parks and their habitat diversity. Each auto-trigger camera was installed 1.5 to 2.5 meters above ground and their locations were recorded using GPS receiver. Depending on the abundance of the carnivores, film and battery was collected and replaced every 2 to 4 weeks.
Relative abundance, distribution pattern, and activity pattern for each species were generated from the field survey. The relative abundance for each species was represented by the Occurrence Index (OI = the number of pictures taken per 1,000 camera working hour). Serial pictures belong to the same individual taken during a short period (usually within 30 minutes) will be considered only 1 picture in the calculation to prevent the over-representative of a lingering individual, hence, to reduce the possibility of over-estimation of the abundance for the species. Camera working hours for each roll of film was the time span between the starting time of a new roll of film and the time recorded on the last picture in the case of the film was finished before the next checking by the researchers, or the time when the researchers arrived for collecting the film, in the case of the film was not finished. Comparisons were made both among species and within species among the different country parks and special area.The spatial distribution pattern for each species in Hong Kong SAR was generated using GIS. The map of country parks and special area were generated from the 1/20,000 Hong Kong base map using ArcView 8.1 software package (ESRI 2000). Range maps of each species were generated using the 17 study areas boundaries for distribution mapping. To understand the variation of the carnivore richness among the 17 study areas, a richness distribution map was produced based on the total number of carnivores species recorded in each study areas.
The level of OI-value for each species in each study areas was adapted to provided the information of the abundance for each species in study areas and to show the general distribution pattern for each species. To avoid arbitrary classification and unequal probability of occurrence due to different activity pattern of different species, the OI-values were grouped into 5 classes based on their distances to the mean, i.e. mean ± 0.5 std., <mean – 0.5 std., mean + 0.5 std. to mean+ 1.5 std., mean + 1.5 std. to mean +2.5 std., and >mean + 2.5 std. as levels of abundance to be “medium”, “low”, “abundant”, “high”, and “very high” respectively.
Results and Discussions
A total of 208,069.96 camera working hours (New Territories= 130,405.58 hours; Hong Kong Island= 36,195.26 hours; Lantau Island= 41,469.12 hours) and 2,137 pictures was carried out by the 100 auto-trigger cameras and a total of 714 pictures of carnivores were obtained. The number of carnivore picture taken in each study area was positively correlated with the number of camera working hours for the same area (r2= 0.80), which indicated the normal function of the auto-trigger cameras throughout the study period. The Occurrence Index (OI) for all carnivores for each study area, however, was not correlated with its camera working hours (r2= 0.01). This result showed that the abundance information (i.e., the OI-value) derived from data collected by the auto-trigger camera was independent from the survey effort, therefore a reliable index for relative abundance.
Among the 20 mammalian species recorded in this study, 9 were carnivores including Chinese ferret badger (Melogale moschata), yellow-bellied weasel (Mustela kathiah), masked palm civet (Paguma larvata), small Indian civet (Viverricula indica), small Asian mongoose (Herpestes javanicus), crab-eating mongoose (Herpestes urva), Chinese leopard cat (Felis bengalensis), feral cat (Felis catus), and feral dog (Canis familiaris). Chinese ferret badger (OI= 0.81), small Indian civet (OI= 0.99), and feral dog (OI= 0.95) were most abundant species (Table 1). Yellow-bellied weasel, small Asian mongoose and crab-eating mongoose were the rare species (Table 1).
Table 1 Survey data, species richness, and OI value of Carnivores in different Country Parks and the Tai Po Kau Special Areas
Habitat requirements should be the most important factors influence the abundance of species distribution in study areas. Except for the small Asian mongoose, all rare species also had very restricted distribution among the study area (Fig. 1). The small Asian mongoose can be considered as wider-distributed rare species because it had been recorded from the north (Pat Sin Leng Country Park) to the south (the Shing Mun Country Park) of the New Territories (Fig. 1e). Small Indian civet (Fig. 1d), masked palm civet (Fig. 1c), and leopard cat (Fig. 1g) were recorded only in the New Territories region and the Hong Kong Island. The reason for the locally extinction of Lantau Island’s population for these species requires further investigation, however, isolation of the Island, competition exclusion by species already existed, and the dispersal behavior of these species might be more important factors than the habitat suitability in this respect. Lastly, the interesting distribution pattern of the feral cat, i.e., common only in Lantau Island and the western part of the Hong Kong Island (Fig. 1h), seemed to be negatively correlated with the distribution of the leopard cat (Fig. 1g). The necessary of additional investigating on the possible competition status between these two species is indicated.
(a)
(b)
(c)
(d)
(e)
(f)
(g)
(h)
(i)
(a) ferret badger (b) yellow-bellied weasel (c) masked palm civet (d) small Indian civet (e) small Asian mongoose (f) crab-eating mongoose (g) leopard cat (h) feral cat (i) feral dog
Figure 1. Carnivores distribution for country parks and the Tai Po Kou special area, Hong Kong SAR. November 2001 - March 2002
In general, the Lantau Island had lower carnivores abundances, while Hong Kong Island and the New Territories had higher mammalian abundance (Table1). It is also important to note that a number of study areas were especially low in their mammal abundance including the Tai Lam in The New Territories, the Tai Tam Country Park in Hong Kong Island, and the Lantau South Country Park in Lantau Island. The fact that the New Territories and Hong Kong Island had similar maximum OI-values (i.e., range from 5 to 6), but the number of areas with maximum OI-values varies suggested an OI-value for all carnivores around 5.00 probably is the optimum capacity of the best habitat presently existing in the Hong Kong SAR (Table 1). Further improvement or proper management of protected areas may increase this maximum capacity. The New Territories was presumably having higher carnivores abundance for its larger in area hence possibly more good habitats. Other than larger in area, the distance to inland may also have contributed to its higher abundance. Further study is necessary for the possible cause.
Figure 2. Carnivore richness for country parks and the Tai Po Kou special area, Hong Kong SAR. November 2001 - March 2002
The carnivore richness for the study areas ranges from 3 to 8 (Table 1). The highest richness located in the New Territories region and both of the two country parks on Lantau Island were poor in species richness (Fig. 2). Other than poor habitat diversity, the higher isolation level may also negatively affect the number of carnivores. The same reason may also explain the low species richness in the Lam Tsuen Country Park, Kam Shan Country Park, and Lion Rock Country Park (Fig. 2). For corridor-liked areas, e.g. Kam Shan-Lion Rock Country Parks Connection in the New Territories region, the pattern of “line corridor” may have been more efficient in promoting the movement of animals between two ends, but provide few habitats for wildlife (Forman 1995). Fewer species inhabit in line corridors and they are mainly edge species (Forman and Godron 1981, 1986), but they usually packed together to form a high-density group (Niering and Goodwin 1974, Pollard et al. 1974, Chasko and Gates 1982, Sanders and Hobbs 1991, and Malanson 1993). This theory was further proofed by the high density value (OI) of the Kam Shan Country Park and the Lion Rock Country Park (Table 1). For country parks with both low species richness and low density such as Plover Cove Country Park, Lam Tsuen Country Park, Tai Po Kou special area, and Shek O Country Park, poor habitat quality may be responsible for its exceptional low species richness (Table 1).
Overall, distribution mapping using GIS and field survey data gives a clear understanding for carnivore distribution pattern that is otherwise difficult to see. It was proofed to be an useful tool for wildlife management. Further studies such as DNA Cytochrome sequencing checking, habitat requirements, and habitat survey and mapping are needed to examine possible causes for distribution patterns of carnivores.
References
For wildlife management, the ability to model spatial distribution and changes in distribution of wildlife is of considerable importance. Once spatial distribution can be adequately modeled, the distribution and abundance can be monitored efficiently over time, and future changes can be predicted. These spatial characteristics and relationship are often difficult to identify and hard to display with traditional ground survey or statistical model. Therefore, using Geographic Information System (GIS) has become an evitable trend in ecology studies and developing wildlife habitat model (Rickers et al. 1995). The characteristics of spatial analysis and spatial display of GIS not only provides efficient way of data handling, storage, and analyzing, more importantly, it also enables mapping of wildlife distribution, identification of patterns, and generation of habitat spatial characteristics, hence, a useful tool in decision making for conservation and wildlife management (Scott et al.1992, Aspinall and Veitch 1993, Buckland and Elston 1993, Long et al. 1993, Ji-Wei and Clinton 2000, Lenton et al. 2000).
Auto-trigger camera has been used successfully in mammal survey worldwide to avoid possible inconsistency induced by weather, different investigators, limiting detective abilities of researchers at night, and etc. (Gysel and Davis 1956, Dodge and Snyder 1960, Carter and Slater 1991, Jones and Raphael 1993, Kucera and Barrette 1993, Pei et al. 1997). It has also been used in wildlife study to generate their activity patterns (Pearson 1960, Carley et al. 1970, Pei 1995, 1998), spatial distribution patterns (Pei et al. 1997, Pei 1999), and to estimate population density (Karanth 1995, Karanth and Nichols 1998, McCullough et al. 2000, Pei 2001b). It has the advantages of recording secretive animals, especially those do not leave prominent signs, and collecting data day and night in a more standardized and consistent way than most traditional methods (e.g., trapping, sign-searching, transect line, etc.)
In this study, the applicability of using the existed GIS technique coupled with field survey data from auto-triggered cameras were assessed for carnivore distribution mapping in fragmented landscape of Hong Kong Special Administration Region (the Hong Kong SAR).
Material and Methods
Auto-trigger cameras were used to conduct carnivores survey from October 2001 to March 2002. A total of 100 auto-trigger cameras were distributed randomly in 16 country parks and 1 special area of Hong Kong SAR according to the size of the country parks and their habitat diversity. Each auto-trigger camera was installed 1.5 to 2.5 meters above ground and their locations were recorded using GPS receiver. Depending on the abundance of the carnivores, film and battery was collected and replaced every 2 to 4 weeks.
Relative abundance, distribution pattern, and activity pattern for each species were generated from the field survey. The relative abundance for each species was represented by the Occurrence Index (OI = the number of pictures taken per 1,000 camera working hour). Serial pictures belong to the same individual taken during a short period (usually within 30 minutes) will be considered only 1 picture in the calculation to prevent the over-representative of a lingering individual, hence, to reduce the possibility of over-estimation of the abundance for the species. Camera working hours for each roll of film was the time span between the starting time of a new roll of film and the time recorded on the last picture in the case of the film was finished before the next checking by the researchers, or the time when the researchers arrived for collecting the film, in the case of the film was not finished. Comparisons were made both among species and within species among the different country parks and special area.The spatial distribution pattern for each species in Hong Kong SAR was generated using GIS. The map of country parks and special area were generated from the 1/20,000 Hong Kong base map using ArcView 8.1 software package (ESRI 2000). Range maps of each species were generated using the 17 study areas boundaries for distribution mapping. To understand the variation of the carnivore richness among the 17 study areas, a richness distribution map was produced based on the total number of carnivores species recorded in each study areas.
The level of OI-value for each species in each study areas was adapted to provided the information of the abundance for each species in study areas and to show the general distribution pattern for each species. To avoid arbitrary classification and unequal probability of occurrence due to different activity pattern of different species, the OI-values were grouped into 5 classes based on their distances to the mean, i.e. mean ± 0.5 std., <mean – 0.5 std., mean + 0.5 std. to mean+ 1.5 std., mean + 1.5 std. to mean +2.5 std., and >mean + 2.5 std. as levels of abundance to be “medium”, “low”, “abundant”, “high”, and “very high” respectively.
Results and Discussions
A total of 208,069.96 camera working hours (New Territories= 130,405.58 hours; Hong Kong Island= 36,195.26 hours; Lantau Island= 41,469.12 hours) and 2,137 pictures was carried out by the 100 auto-trigger cameras and a total of 714 pictures of carnivores were obtained. The number of carnivore picture taken in each study area was positively correlated with the number of camera working hours for the same area (r2= 0.80), which indicated the normal function of the auto-trigger cameras throughout the study period. The Occurrence Index (OI) for all carnivores for each study area, however, was not correlated with its camera working hours (r2= 0.01). This result showed that the abundance information (i.e., the OI-value) derived from data collected by the auto-trigger camera was independent from the survey effort, therefore a reliable index for relative abundance.
Among the 20 mammalian species recorded in this study, 9 were carnivores including Chinese ferret badger (Melogale moschata), yellow-bellied weasel (Mustela kathiah), masked palm civet (Paguma larvata), small Indian civet (Viverricula indica), small Asian mongoose (Herpestes javanicus), crab-eating mongoose (Herpestes urva), Chinese leopard cat (Felis bengalensis), feral cat (Felis catus), and feral dog (Canis familiaris). Chinese ferret badger (OI= 0.81), small Indian civet (OI= 0.99), and feral dog (OI= 0.95) were most abundant species (Table 1). Yellow-bellied weasel, small Asian mongoose and crab-eating mongoose were the rare species (Table 1).
Table 1 Survey data, species richness, and OI value of Carnivores in different Country Parks and the Tai Po Kau Special Areas
| New Territories | Hong Kong | Lantau Island | No. of picture | OI | |||||||||||||||
| Plover Cove | Pat Sin Leng | Lam Tsuen | Tai Po Kou | Tai Mo Shan | Tai Lam | Shing Mun | Kam Shan | Lion Rock | Ma On Shan | Sai Kung | Pok Fu Lam | Aberd een | Tai Tam | Shek O | Lantau North | Lantau South | |||
| Ferret Badger | 0 | 8 | 0 | 0 | 4 | 6 | 13 | 1 | 6 | 44 | 28 | 13 | 5 | 0 | 13 | 19 | 4 | 168 | 0.81 |
| Yellow-bellied Weasel | 0 | 3 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 2 | 1 | 0 | 0 | 6 | 2 | 3 | 0.01 |
| Masked Palm Civet | 0 | 5 | 0 | 0 | 3 | 2 | 3 | 0 | 0 | 19 | 5 | 13 | 8 | 4 | 5 | 0 | 0 | 67 | 0.32 |
| Small Indian Civet | 5 | 49 | 1 | 5 | 8 | 34 | 13 | 2 | 1 | 31 | 18 | 4 | 12 | 6 | 16 | 0 | 0 | 205 | 0.99 |
| Small Asian Mongoose | 0 | 2 | 0 | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 3 | 0.01 |
| Crab-eating Mongoose | 1 | 6 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 7 | 0.03 |
| Leopard Cat | 0 | 1 | 4 | 0 | 3 | 2 | 5 | 4 | 1 | 22 | 10 | 2 | 2 | 0 | 2 | 0 | 0 | 58 | 0.28 |
| Feral Cat | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 1 | 2 | 1 | 0 | 0 | 6 | 2 | 26 | 0.12 |
| Feral Dog | 5 | 12 | 7 | 8 | 9 | 1 | 24 | 12 | 13 | 42 | 26 | 13 | 7 | 2 | 10 | 2 | 1 | 198 | 0.95 |
| Richness | 3 | 8 | 3 | 2 | 5 | 5 | 6 | 4 | 4 | 6 | 6 | 6 | 6 | 3 | 5 | 3 | 3 | ||
| No of picture | 11 | 86 | 12 | 13 | 27 | 45 | 59 | 19 | 21 | 159 | 88 | 47 | 35 | 12 | 46 | 27 | 7 | 714 | |
| OI | 2.56 | 4.66 | 4.31 | 2.55 | 3.71 | 1.81 | 3.24 | 4.05 | 5.56 | 6.02 | 6.04 | 5.84 | 2.84 | 1.55 | 5.70 | 2.02 | 0.25 | ||
Habitat requirements should be the most important factors influence the abundance of species distribution in study areas. Except for the small Asian mongoose, all rare species also had very restricted distribution among the study area (Fig. 1). The small Asian mongoose can be considered as wider-distributed rare species because it had been recorded from the north (Pat Sin Leng Country Park) to the south (the Shing Mun Country Park) of the New Territories (Fig. 1e). Small Indian civet (Fig. 1d), masked palm civet (Fig. 1c), and leopard cat (Fig. 1g) were recorded only in the New Territories region and the Hong Kong Island. The reason for the locally extinction of Lantau Island’s population for these species requires further investigation, however, isolation of the Island, competition exclusion by species already existed, and the dispersal behavior of these species might be more important factors than the habitat suitability in this respect. Lastly, the interesting distribution pattern of the feral cat, i.e., common only in Lantau Island and the western part of the Hong Kong Island (Fig. 1h), seemed to be negatively correlated with the distribution of the leopard cat (Fig. 1g). The necessary of additional investigating on the possible competition status between these two species is indicated.
(a)
(b)
(c)
(d)
(e)
(f)
(g)
(h)
(i)
(a) ferret badger (b) yellow-bellied weasel (c) masked palm civet (d) small Indian civet (e) small Asian mongoose (f) crab-eating mongoose (g) leopard cat (h) feral cat (i) feral dog
Figure 1. Carnivores distribution for country parks and the Tai Po Kou special area, Hong Kong SAR. November 2001 - March 2002
In general, the Lantau Island had lower carnivores abundances, while Hong Kong Island and the New Territories had higher mammalian abundance (Table1). It is also important to note that a number of study areas were especially low in their mammal abundance including the Tai Lam in The New Territories, the Tai Tam Country Park in Hong Kong Island, and the Lantau South Country Park in Lantau Island. The fact that the New Territories and Hong Kong Island had similar maximum OI-values (i.e., range from 5 to 6), but the number of areas with maximum OI-values varies suggested an OI-value for all carnivores around 5.00 probably is the optimum capacity of the best habitat presently existing in the Hong Kong SAR (Table 1). Further improvement or proper management of protected areas may increase this maximum capacity. The New Territories was presumably having higher carnivores abundance for its larger in area hence possibly more good habitats. Other than larger in area, the distance to inland may also have contributed to its higher abundance. Further study is necessary for the possible cause.
Figure 2. Carnivore richness for country parks and the Tai Po Kou special area, Hong Kong SAR. November 2001 - March 2002
The carnivore richness for the study areas ranges from 3 to 8 (Table 1). The highest richness located in the New Territories region and both of the two country parks on Lantau Island were poor in species richness (Fig. 2). Other than poor habitat diversity, the higher isolation level may also negatively affect the number of carnivores. The same reason may also explain the low species richness in the Lam Tsuen Country Park, Kam Shan Country Park, and Lion Rock Country Park (Fig. 2). For corridor-liked areas, e.g. Kam Shan-Lion Rock Country Parks Connection in the New Territories region, the pattern of “line corridor” may have been more efficient in promoting the movement of animals between two ends, but provide few habitats for wildlife (Forman 1995). Fewer species inhabit in line corridors and they are mainly edge species (Forman and Godron 1981, 1986), but they usually packed together to form a high-density group (Niering and Goodwin 1974, Pollard et al. 1974, Chasko and Gates 1982, Sanders and Hobbs 1991, and Malanson 1993). This theory was further proofed by the high density value (OI) of the Kam Shan Country Park and the Lion Rock Country Park (Table 1). For country parks with both low species richness and low density such as Plover Cove Country Park, Lam Tsuen Country Park, Tai Po Kou special area, and Shek O Country Park, poor habitat quality may be responsible for its exceptional low species richness (Table 1).
Overall, distribution mapping using GIS and field survey data gives a clear understanding for carnivore distribution pattern that is otherwise difficult to see. It was proofed to be an useful tool for wildlife management. Further studies such as DNA Cytochrome sequencing checking, habitat requirements, and habitat survey and mapping are needed to examine possible causes for distribution patterns of carnivores.
References
- Aspinall, R. and N. Veitch. 1993. Habitat mapping from satellite imagery and wildlife survey data using a Bayesian modeling procedure in a GIS. Photogrammetric Engineering and Remote Sensing 59: 537-543.
- Buckland, S. T. and D. A. Elston. 1993. Empirical models for the spatial distribution of wildlife. J. Applied Ecology (1994) 30: 478-495.
- Carley, C. J., E. D. Fleharty, and M. A. Mares. 1970. Occurrence and activity of Reithrodontomys megalotis, Microtus ochrogaster and Peromyscus maniculatus as recorded by a photographic device. Southwestern Nat. 15: 209-216.
- Carter, S. M. and E. Slater. 1991. Monitoring animal activity with automated photography. J. Wildl. Manage. 55: 689-692.
- Chasko, G. G. and J. E. Gates. 1982. Avian habitat suitability along a transmission-line corridor in an oak-hickory forest region. Wildl. Monographs 82: 1-41.
- Dodge, W.E. and D. P. Synder. 1960. An automatic camera device for recording wildlife activity. J. Wildl. Manage. 24: 340-342.
- Forman, R. T. T. 1995. Land mosaics- the ecology of landscapes and regions. Cambridge Univ. Press, Cambridge.
- Forman, R. T. T. and M. Godron. 1981. Patches and structural components for a landscape ecology. BioScience 31: 733-740.
- Forman, R. T. T. and M. Godron. 1986. Landscape ecology. John Wiley, New York.
- Gysel, L. W. and E. M. Davis Jr. 1956. A simple automatic photographic unit for wildlife research. J. Wildl. Manage. 20: 451-453.
- Ji-Wei and J. Clinton. 2000. Spatial modeling of the geographic distribution of wildlife populations: A case study in the lower Mississippi River region. Ecological Modelling 132 (1-2): 95-104.
- Jones, L. L. C. and M. G. Raphel. 1993. Inexpensive camera systems for detecting martens, fisher, and other animals: guideline for use and standardization. U.S. Forest Service General Technical Report PNW-GTR-306.
- Karanth, K. U. 1995. Estimating tiger populations from camera-trap data using capture-recapture models. Biological Conservation 71: 333-338.
- Karanth, K. U. and J. D. Nichols. 1998. Estimation of tiger densities in India using photographic captures and recaptures. Ecology 79(8): 2852-2862.
- Kucera, T. E. and R. H. Barrette. 1993. The Tarilmaster camera system for detecting wildlife. Wildl. Soc. Bull. 21: 505-508.
- Lenton, S.M., J.E. Fa, and J.P. Del Val. 2000. A simple non-parametric GIS model for predicting species distribution: endemic birds in Bioko Island, West Africa. Biodiversity and Conservation 9(7): 869-885.
- Long, A.J., M.H. Crosby, A.J. Stattersfield, and D.C. Wege. 1993. Towards a global map of biodiversity: patterns in the distribution of restricted-range birds. Global Ecology and Biogeography letters 5(4-5): 281-304.
- Malanson, G. P. 1993. Riparian landscapes. Cambridge Univ. Press, Cambridge.
- McCullough, D. R., K. C. J. Pei, and Y. Wang. 2000. Home range, activity patterns, and habitat relations of Reeves’ muntjacs in Taiwan. Journal Wildlife Management 64(2):430-441.
- Niering, W. A. and R. H. Goodwin. 1974. Creation of relatively stable shrublands with herbicides: arresting “succession” on rights-of-way and pasture land. Ecology 55: 784-795.
- Pearson, O. P. 1960. Habits of Microtus californicus revealed by automatic photographic records. Ecol. Monogr. 30: 231-249.
- Pei, K. 1995. Activity Rhythm of the Spinous Country Rat (Niviventer coxingi) in Taiwan. Zoological Studies 34(1): 55-58.
- Pei, K. 1997. Avian and mammal fauna of the Amentotaxus formosana Nature Reserve. Q. Jour. Chin. For. 30(2): 131-150.
- Pei, K. 1998. An evaluation of using auto-trigger cameras to record activity patterns of wild animals. Taiwan J. For. Sci. 13(4): 317-324.
- Pei, K. 1999. Spatial Distribution Patterns of the Red-bellied Tree Squirrel and Owston’s Long-nosed Tree Squirrel in Natural Forest in Southern Taiwan. Mammalian Sci. 39(1): 193-196.
- Pei, K. J. C. 2001b. The present status of the re-introduced Formosan sika deer (Cervus nippon taiouanus) in Kenting National Park. Q. Jour. Chin. For. 34 (4): 427-440.
- Pei, K., Chen, C. T., Wu, S. T. and Teng, M. C. 1997. Use of auto-trigger camera and Geographic Information System to study spatial distribution of forest wildlife. Q. Jour. Chin. For. 30(3): 279-289.
- Pollard, E., M. D. Hooper and N. W. Moore. 1974. Hedges. W. Collins, London.
- Sanders, D. A. and R. J. Hobbs, eds. 1991. Nature Conservation 2: the role of corridors. Surry Beatty, Chipping Norton, Australia.
- Scott, J. M., F. Davis, B. Csuti, R. Noss, B Butterfield, C. Graves, H. Anderson, S. Caicco, F. D’erchia, T. C. Edwards, Jr., J. Ulliman, and R. G. Wright. 1992. Gap analysis: a geographic approach to protection of biological diversity. Wildl. Monogr. 123:1-41.