Bird Diversity and Abundance in Forest Fragments
William Oduro, Institute of Renewable Natural Resources, Kwame Nkrumah University of Science and Technology, Kumasi, Ghana.
Bright Kankam, Forestry Research Institute of Ghana, Kumasi, Ghana.
Summary
Throughout the world, tropical forests are rapidly being converted to agriculture. In most regions, these forests will survive only in fragments for the next several decades. Research is needed on how well plants and animals can persist in such fragments. This project will provide the first data on diversity and abundance of one taxon (namely birds) in forest patches of Ghana, a tropical African nation where forest fragmentation is most severe. Such research will help determine what combinations of patch size and management regime provide a suitable refuge for various bird species. This project has support from the Forestry Research Institute of Ghana (FORIG), the University of Science and Technology (UST) in Kumasi, Ghana's Nature Conservation Research Centre (NCRC), and a Fulbright award from the African Regional Research Program.
Introduction
Occupying less than 7% of the world's land mass, tropical forests harbor over 50% of all plant and animal species (Myers 1992, Bowles et al. 1998). As a group, birds are closely associated with forests, and approximately 30% of the world's species of birds are so restricted to tropical forests (either for winter or year-round habitat) that they would disappear if all tropical forests were lost (Myers 1992). Excluding the impacts of selective logging, these forests are being lost at a rate of about 1.2% annually (FAO 1993). Given the needs of growing human populations, and the agricultural economies of many tropical countries, tropical forests will increasingly become fragments in an agricultural landscape. Until (hopefully) economies mature and some agricultural lands can revert to forest in the 21st century, these fragments will be refuges for tropical biodiversity. It is incumbent that we become good managers of forest fragments, because (at least for the next 50-100 years) we will be unable to protect biodiversity by setting aside large areas of forest.
The most fragmented tropical forests today occur in the Philippines, peninsular Malaysia, Ghana, and Costa Rica (Whitmore 1997); thus these nations represent the future of tropical forests and are the appropriate locations in which to study biodiversity in forest fragments. Conservation biologists recently produced the first 2 volumes on biodiversity in tropical forest fragments (Schelhas and Greenberg 1996; Laurance and Bierregaard 1997), focusing largely on those countries and regions where fragmentation is most advanced. However the 51 chapters in these 2 volumes contain only 2 reports from Africa, neither of which addressed vertebrate diversity: Lebbie and Freudenberger (1996) catalogued the sizes of sacred forest groves in Sierra Leone, and Smith (1997) analyzed historic and current vegetation maps to describe patterns of forest loss in western Madagascar. Studies in Central America, South America, and Asia suggest that small forest fragments can support a surprisingly large fraction of a region's native vertebrate species, that certain subgroups (e.g., fruit-eating birds, larger mammals) are more vulnerable to loss, and that the type of management (e.g., park, timber production reserve, community-based reserve) also influences which species persist (Bennet and Caldecott 1981, Duff et al. 1984, Brown and Brown 1992, Robinson and Redford 1991, Magsalay et al. 1995, Warkentin et al. 1995, Greenberg 1996, Redford and Mansour 1996, Bierregaard and Stouffer 1997, Thiollay 1997, Warburton 1997, Marsden 1998). The only similar study in Africa (Newmark 1991) showed similar patterns for 9 fragments in a dry woodland region. Additional work is needed in Africa to determine how diversity fares in forest fragments, and as a starting point for better management of those fragments. Although many taxa would be appropriate for such a study, birds are a diverse and important group that can readily be inventoried in a project of modest duration. Furthermore, forest birds as a group may be more sensitive than other taxa to fragmentation. For example, Corlett and Turner (1997) reported that over 50% of forest interior bird species (in contrast to only 29% of vascular plant species) became locally extinct due to massive forest fragmentation in Singapore.
This project will be the first to examine vertebrate diversity in African forest fragments. Our goals are to (1) to determine bird diversity and abundance in about 25 forest fragments in each of 4 major forest zones in Ghana, (2) to characterize such diversity as a function of various factors (including fragment size, human disturbance, type of management, and proximity to other forest patches), and (3) determine if other factors (such as nest location, body size, or trophic level of a bird species) are correlated with a species' vulnerability to fragmentation. Such data would have immediate application to management and would form a springboard for future research projects. The project will involve as collaborators Dr. William Oduro, the wildlife ecologist in the Institute for Renewable Natural Resources (IRNR) at the University of Science and Technology (UST) in Kumasi, and Mr. Bright Kankam, a recent IRNR graduate who in 1998 became the first wildlife ecologist on the staff of the Forest Research Institute of Ghana (FORIG).
Methods
We will study birds in 4 forest regions, following the classification system of Hall and Swaine (1981) and Wagner et al. (1991). In order of increasing aridity (and distance from the coast) the 4 regions are moist evergreen forest, moist semi-deciduous forest, dry semi-deciduous forest, and Guinea savannah woodland. Research will be conducted during July 1999-August 2000.
In each region we will select 20 to 30 forest patches, with about 5 patches in each of 5 size classes, namely 2-5 ha, 6-30 ha, 31-149 ha, 150-600 ha, and >600 ha. In each regions, we will include the two largest protected forest areas in the largest size class; supposedly these will contain a full complement of native bird species. Each fragment will be at least 250 m from other forest patches. We will attempt, insofar as is possible, to achieve a balanced design with respect to distance to nearest larger patch (e.g., classes of 250-1000m, 1-5 km, and >5 km), and with respect to other factors such as road density, proximity to human settlements, and degree of previous logging - and to categories created by combinations of these factors. Attention to the factor combinations will mitigate problems due to "natural confounding" that would occur, for example, if size and interpatch distance were 100% correlated in a poorly-chosen sample of sites. However, we realize that with only about 30 patches in each type, it will be impossible to achieve a fully balanced design for all combinations of factors.
Each forest region contains many fragments in each size class. Two large National Parks, namely Kakum (in the moist semi-deciduous forest zone) and Mole (in the Guinea savannah woodland) would provide the largest reference sites in these two zones. In addition, each zone contains 1-3 large (>20,000 ha) Forest Reserves and 12 to 30 smaller (45 to 10,000 ha) Reserves under the jurisdiction of the Ghana Forestry Department. Hawthorne and Abu-Juam (1995) and Wagner et al. (1991) list the size, forest condition, and other data for these forests. Over 95% of remaining intact forest exists within these Reserves and National Parks (Wagner et al. 1991), and most of the balance is in small sacred groves established near villages throughout the country. No new Forest Reserves (and probably no new sacred groves) have been established since the late 1970's. Agricultural clearing since their establishment has left most reserves and groves as isolated islands within agricultural lands that had previously been forested. The edges of forest patches are obvious to even the casual observer from an automobile or aircraft, or in satellite imagery.
Objective 1: Determine bird diversity and abundance in forest fragments each of 4 major forest zones in Ghana
In each forest zone, we will focus on 50-80 bird species that are restricted to, reach their greatest abundance in, or occur commonly in that forest type. We will train field observers to recognize these 50 species by sight and sound. Our previous experience suggests that over 75% of detections will be by sound alone, and therefore training in bird song recognition is most important.
We will conduct point counts on five circular plots (50-m radius) in each fragment (Ralph et al. 1995), with locations selected to distinguish between area-sensitive species (which would decline with patch size even in counts made at the patch center) from edge-sensitive species (which would decline with proximity to edge regardless of patch size). Plots will be at least 150m apart, except in the smallest patches where this would be physically impossible. In each patch, two plots will abut corners or peninsulas of the patch (locations with maximum edge); two plots will be about 100 m from, and near the midpoint of, the longest relatively straight edge; and the 5th plot will be at the approximate center of the patch. Within a forest type, we will conduct all point counts within as short a period as possible. Because our ability to travel quickly to the plots and our familiarity with the birds will increase over time, the point counts will probably occur near the end of the study, but avoiding the May-June rainy season. To avoid pseudo-replication (Hurlburt 1984), we will sum each species' abundance across the 5 plots within a patch to yield a single observation per patch.
Although the point counts provide a repeatable measure of diversity across patches of different sizes, they will certainly provide incomplete lists of species present. To obtain a more complete species list, observers will census each fragment 3 to 6 times, with the duration of each census proportional to patch area, e.g., 30 minutes for sites <2 ha, 1 hour for sites 2-5 ha, 1.5 hours for sites 5-20 ha, and 5 hours for the largest patches (Warburton 1997, Ralph and Scott 1981). For patches >150ha, we will census a 150-ha rectangular subarea, with length < 3 times its width, and oriented to encompass most of the diversity (topographic, vegetative, or edaphic) in the patch. In each census, observers will travel slowly through the plot, stopping frequently for 1-5 minutes, looking and listening for species not previously documented in that patch and noting evidence of breeding (such as courtship, or birds carrying nest material or food) for all species. Because breeding seasons for African birds are diffuse within species and non-synchronous among species, and because breeding birds are more detectable than non-breeders, at least 4 weeks will elapse between censuses. To the maximum extent possible, we will rotate observers between visits to a patch. For any patch in which >2 new species were detected during the 3rd visit, additional censuses will be carried out until <2 new species are detected during a census, up to a maximum of 6 visits. To control for the effects of different sample sizes, we will use rarefaction (Heck et al. 1975, James and 1981) in comparing species richness across plots.
For descriptive purposes, we will also undertake limited sampling of the agricultural lands surrounding the forest patches.
Objective 2: Characterize avian diversity as a function of fragment size, human disturbance, type of management, proximity to other forest patches, and other factors
During the first visit we will classify each patch with respect to 8 physiognomic parameters, namely canopy cover near the edge, tree density (using the rapid method of Marsden 1998), canopy cover in the interior, density of treefall gaps, amount of woody debris on the ground, density of lianas, abundance of non-native pioneer plant species, and density of herbaceous and shrub understories. We will also record several edaphic and landscape variables, including road density, proximity to human settlements, water availability (based on stream length of area or open water), topographic roughness (fraction of land in steep slopes), degree of disturbance of the understory (based on farm patches, palm tapping, trail density), and an estimate of past logging damage (based on stumps and logging tracks, or logging records when available). For each parameter that cannot be measured quickly, we will develop objective and repeatable criteria to quickly classify each patch into one of at most 5 ordinal classes.
Many bird species in tropical forests are remarkably rare, even in undisturbed primary forest. For instance Thiollay (1997) reported that 70% of the species were present on fewer than 2% of his 0.25-ha plots, and only 5% of the species were present on over 8% of the plots. If a similar pattern occurs in West Africa, the abundances of many species will be too low to allow statistical inferences about them. Therefore, following Thiollay (1997), we will group species into about 12 groups according to body size, diet, and broad habitat affinity. Examples of such groups are "small terrestrial insectivores associated with forest understory" and "large to medium frugivores associated with upper canopy of forests."
For each forest type, we will graph the probability of occurrence of species and diversity of species groups as a function of patch size class. We will similarly plot abundance of species and species groups as a function of patch size. Doubtless species will vary in their response to fragmentation, with some species present in all patch sizes, and others showing aversion to small patches with thresholds at various patch sizes. Such patterns are evident for birds in North America (Robbins et al. 1989) and Australia (Warburton 1997), and are likely to occur in Africa as well. This sort of analysis will identify west African species that are most sensitive to forest fragmentation, a group commonly referred to as "area-sensitive" birds (Robbins et al. 1989).
We will use partial correlations and stepwise multiple regression to correlate species richness (number of species in a patch) with patch size, distance to nearest patch, and the physiognomic and landscape parameters mentioned above. We will also use stepwise regression to assess the effects of the landscape parameters on the forest structural variables. These analyses will help identify which patches are most valuable from a conservation perspective, and what simple management actions (e.g., closing roads) might increase conservation value of a patch. To avoid pseudo-replication, we will conduct most regressions on a per-patch basis. Because such analyses may fail to detect bird responses to local conditions, we also may conduct additional regressions on a per-point basis. The latter analyses will require reasonable screening of overly-dependent points, and cautious interpretation.
Objective 3: Determine species traits (nest location, body size, trophic level) that are associated with vulnerability to fragmentation
In addition, we will use partial correlations and multivariate analysis to identify environmental and life history attributes that are correlated with sensitivity to fragmentation. These attributes will include nest location (ground, understory, low canopy, high canopy), fecundity, body size, geographic range size, natural abundance (in large patches of preferred habitat), trophic level (frugivore, insectivore, omnivore, raptor), and feeding zone (ground, understory, canopy). These will help us formulate hypotheses regarding the mechanisms that underlie a species' sensitivity to fragmentation. Testing these hypotheses - as has been done for area-sensitive birds in US forests by Robinson et al. (1995) and Donovan et al. (1996), and others - would be fertile and exciting ground for future research by Ghanaian researchers.
Finally, we will use nonmetric multidimensional scaling (NMDS; Minchin 1987) to describe major patterns of variation in the forest structural variables and bird communities, and we will use nested subsets analysis (Patterson 1987) to determine whether predictable subsets of species drop out as patch size decreases. This would suggest that a highly nonrandom component would be susceptible to extinction in small patches, and that management of larger patches should focus on such species while management of smaller patches focuses on the species that can reasonably be expected to reside there.
LITERATURE CITED
Bennet, E. L., and J. O. Caldecott. 1981. Unexpected abundance: the trees and wildlife of the Limas Belas Estate forest reserve, Perak. The Planter 57: 516-519.
Bierregaard, R. O., and P. C. Stouffer. 1997. Understory birds and habitat mosaics in Amazonian rainforests. Pages 138-155 In Laurance, W. F., and R. O. Bierregaard, Jr., eds., Tropical forest remnants: ecology, management, and conservation of fragmented communities. University of Chicago Press.
Bowles, I. A., R. E. Rice, R. A. Mittermeier, G. A. B. da Fonseca. 1998. Logging and tropical forest conservation. Science 280:1899-1900.
Brown, K. S., and G. G. Brown. 1992. Habitat alteration and species loss in Brazilian forests. Pages 119-142 in T.C. Whitmore and J. A. Sayer, eds., Tropical deforestation and species extinction. Chapman and Hall.
Corlett, R. T., and I. M. Turner. 1997. Long-term survival in tropical forest remnants in Singapore and Hong Kong. Pages 333-356 In Laurance, W. F., and R. O. Bierregaard, Jr. eds. Tropical forest remnants: ecology, management, and conservation of fragmented communities. University of Chicago Press.
Donovan, T. M., F. R. Thompson, J. Faaborg, and J. R. Probst. December 1995. Reproductive success of migratory birds in habitat sources and sinks. Conservation Biology 9:1380-1395.
Duff, A. B., R. A. Hall, and C. Marsh. 1984. A survey of wildlife in and around a commercial tree plantation in Sabah. Malaysian Forester 47:197-213.
FAO (Food and Agriculture Organization of the United Nations). 1993. Forest resources assessment 1990: tropical countries. FAO Forestry Paper 112, FAO, Rome.
Greenberg, R. 1996. Managed forest patches and the diversity of birds in southern Mexico. Pages 59-90 In Schelhas, J., and R. Greenberg, eds., Forest patches in tropical landscapes. Island Press.
Grimes, L. G. 1987. The birds of Ghana. British Ornithologists Union, London.
Hall, J. B., and M. D. Swaine. 1981. Distribution and ecology of vascular plants in a tropical rain forest: forest vegetation of Ghana. W. Junk Publishers, The Hague. 383pp.
Hawthorne, W. D., and M. Abu-Juam. 1995. Forest protection in Ghana. IUCN Publications, Gland, Switzerland and Cambridge, UK. 203pp.
Heck, K. K., G. Vanbelle, and D. S. Simberloff. 1975. Explicit calculation of the rarefaction diversity measurement and the determination of sufficient sample size. Ecology 56:1459-1461.
James, F. C., and S. Rathbun. 1981. Rarefaction, relative abundance, and diversity of avian communities. Auk 98:785-800.
Johns, A. G. 1997. Timber production and biodiversity conservation in tropical rain forests. New York, Cambridge University Press. 225pp.
Kramer, R., C. van Schaik, and R. J. Johnson. 1997. Last stand: protected areas and the defense of tropical biodiversity. New York, Oxford University Press. 242pp.
Laurance, W. F., and R. O. Bierregaard, Jr. 1997. Tropical forest remnants: ecology, management, and conservation of fragmented communities. Chicago, University of Chicago Press. 616pp.
Lebbie, A. R., and M. S. Freudenberger. 1996. Sacred groves in Africa: forest patches in transition. Pages 300-326 In Schelhas, J., and R. Greenberg, eds, Forest patches in tropical landscapes. Island Press.
Magsalay, P., T. Brooks, G. Dutson, and R. Timmins. 1995. Extinction and conservation on Cebu. Nature 373:294.
Marsden, S. J. 1998. Changes in bird abundance following selective logging on Seram, Indonesia. Conservation Biology 12:605-611.
Meffe, G. K. 1998. Into the millenium (editorial). Conservation Biology 12:1-3.
Minchin, P. R. 1987. An evaluation of the relative robustness of techniques for ecological ordination. Vegetatio 69:89-107.
Myers, N. 1992. The primary source: tropical forests and our future. Norton, New York. 416pp.
Newmark, W. D. 1991. Tropical forest fragmentation and the local extinction of understory birds in the eastern Usambara Mountains, Tanzania. Conservation Biology 5:67-78.
Patterson, B. D. 1987. The principle of nested subsets and its implication for biological conservation. Conservation Biology 1:323-334.
Ralph, C. J., and J. M. Scott. 1981. Estimating numbers of terrestrial birds. Studies in Avian Biology 6. Cooper Ornithological Society, Lawrence, Kansas. 630 pp.
Ralph, C. J., J. R. Sauer, and S. Droege. 1995. Monitoring bird populations by point counts. USDA Forest Service General Technical Report PSW-149. 187pp.
Redford, K. H., and J. A. Mansour. 1996. Traditional peoples and biodiversity conservation in large tropical landscapes. America Verde Publications. 278pp.
Robinson, J. G., and K. H. Redford. 1991. Neotropical wildlife use and conservation. Chicago, University of Chicago Press. 520pp.
Robbins, C. S., D. K. Dawson, and B. A. Dowell. 1989. Habitat area requirements of breeding forest birds of the Middle Atlantic States. Wildlife Monographs 103. 34pp.
Robinson, S. K., F. R. Thompson, T. M. Donovan, D. R. Whitehead, and J. Faaborg. 1995. Regional forest fragmentation and the nesting success of migratory birds. Science 267:1987-1990.
Schelhas, J., and R. Greenberg. 1996. Forest patches in tropical landscapes. Washington, D.C., Island Press. 426 pp.
Sponsel, L. E., T. N. Headland, and R. C. Bailey. 1996. Tropical deforestation: the human dimension. New York, Columbia University Press. 365pp. Written by anthropologists who care about biodiversity; scientific citations; readable.
Struhsaker, T. T. 1997. Ecology of an African rain forest : logging in Kibale and the conflict between conservation and exploitation. Univ. Florida Press. 432pp.
Thiollay, J. 1997. Disturbance, selective logging and bird diversity: a Neotropical forest study. Biodiversity and Conservation 6:1155-1173.
Wagner, M. R., S. K. N. Atuahene, and J. R. Cobbinah. 1991. Forest entomology in west tropical Africa: forest insects of Ghana. Kluwer Academic Publishers, Boston. 210pp.
Warkentin, I. G., R. Greenberg, and J. S. Ortiz. 1995. Songbird use of gallery woodlands in recently cleared and older settled landscapes of the Selva Lacandona, Chiapas, Mexico. Conservation Biology 9:1095-1106
Warburton, N. H. 1997. Structure and conservation of forest avifauna in isolated rainforest remnants in tropical Australia. Pages 190-206 In Laurance, W. F., and R. O. Bierregaard, Jr. 1997. Tropical forest remnants: ecology, management, and conservation of fragmented communities. Chicago, University of Chicago Press.
Whitmore, T. C. 1997. Tropical forest disturbance, disappearance, and species loss. Pages 3-12 In Laurance, W. F., and R. O. Bierregaard, Jr. 1997. Tropical forest remnants: ecology, management, and conservation of fragmented communities. University of Chicago Press.
Last Updated ( Thursday, 10 July 2008 )