Plant–frugivore networks are less specialized and more robust at forest–farmland edges than in the interior of a tropical forest

Forest fragmentation and local disturbance are prevailing threats to tropical forest ecosystems and affect frugivore communities and animal seed dispersal in different ways. However, very little is known about the effects of anthropogenic forest edges and of local disturbance on the structure and robustness of plant–frugivore networks. We carried out focal tree observations to record the frugivore species feeding on eight canopy tree species in the forest interior and at forest–farmland edges in a little and a highly disturbed part of a Kenyan rain forest. For each frugivore species, we recorded its body mass and its forest dependence. We examined how forest edge and local disturbance affected the abundance, the richness and the composition of the frugivore community and tested whether forest edge and local disturbance affected plant frugivore networks. Abundance and species richness of frugivores were higher at edges than in the forest interior. Forest visitors and small‐bodied frugivores increased, while forest specialists decreased in abundance at forest edges. The changes in frugivore community composition resulted in plant–frugivore networks that were more connected, more nested and more robust against species extinctions at forest–farmland edges than in the forest interior. Network specialization was lower at forest edges than in the forest interior because at the edges plant specialization on frugivores was very low in small‐fruited species. In contrast, small‐fruited plants were more specialized than large‐fruited plants in the forest interior. Our findings suggest that forest‐visiting birds may stabilize seed‐dispersal services for small‐fruited plant species at rain forest margins, while seed‐dispersal services for large‐fruited plant species may be disrupted at forest edges due to the decrease of large‐bodied frugviores. To assess the ultimate consequences of bird movements from farmland to forest edges for ecosystem functioning, future studies are required to investigate the seed‐dispersal qualities provided by forest‐visiting bird species in the tropics.     

Consequences of inferring diet from feeding guilds when estimating and interpreting consumer–resource stoichiometry

相似文献   

Mass–abundance scaling in avian communities is maintained after tropical selective logging

1. Selective logging dominates forested landscapes across the tropics. Despite the structural damage incurred, selectively logged forests typically retain more biodiversity than other forest disturbances. Most logging impact studies consider conventional metrics, like species richness, but these can conceal subtle biodiversity impacts. The mass–abundance relationship is an integral feature of ecological communities, describing the negative relationship between body mass and population abundance, where, in a system without anthropogenic influence, larger species are less abundant due to higher energy requirements. Changes in this relationship can indicate community structure and function changes.
2. We investigated the impacts of selective logging on the mass–abundance scaling of avian communities by conducting a meta‐analysis to examine its pantropical trend. We divide our analysis between studies using mist netting, sampling the understory avian community, and point counts, sampling the entire community.
3. Across 19 mist‐netting studies, we found no consistent effects of selective logging on mass–abundance scaling relative to primary forests, except for the omnivore guild where there were fewer larger‐bodied species after logging. In eleven point‐count studies, we found a more negative relationship in the whole community after logging, likely driven by the frugivore guild, showing a similar pattern.
4. Limited effects of logging on mass–abundance scaling may suggest high species turnover in logged communities, with like‐for‐like replacement of lost species with similar‐sized species. The increased negative mass–abundance relationship found in some logged communities could result from resource depletion, density compensation, or increased hunting; potentially indicating downstream impacts on ecosystem functions.
5. Synthesis and applications. Our results suggest that size distributions of avian communities in logged forests are relatively robust to disturbance, potentially maintaining ecosystem processes in these forests, thus underscoring the high conservation value of logged tropical forests, indicating an urgent need to focus on their protection from further degradation and deforestation.

     

Winners and losers: tropical forest tree seedling survival across a West African forest–savanna transition

Forest encroachment into savanna is occurring at an unprecedented rate across tropical Africa, leading to a loss of valuable savanna habitat. One of the first stages of forest encroachment is the establishment of tree seedlings at the forest–savanna transition. This study examines the demographic bottleneck in the seedlings of five species of tropical forest pioneer trees in a forest–savanna transition zone in West Africa. Five species of tropical pioneer forest tree seedlings were planted in savanna, mixed/transition, and forest vegetation types and grown for 12 months, during which time fire occurred in the area. We examined seedling survival rates, height, and stem diameter before and after fire; and seedling biomass and starch allocation patterns after fire. Seedling survival rates were significantly affected by fire, drought, and vegetation type. Seedlings that preferentially allocated more resources to increasing root and leaf starch (starch storage helps recovery from fire) survived better in savanna environments (frequently burnt), while seedlings that allocated more resources to growth and resource‐capture traits (height, the number of leaves, stem diameter, specific leaf area, specific root length, root‐to‐shoot ratio) survived better in mixed/transition and forest environments. Larger (taller with a greater stem diameter) seedlings survived burning better than smaller seedlings. However, larger seedlings survived better than smaller ones even in the absence of fire. Bombax buonopozense was the forest species that survived best in the savanna environment, likely as a result of increased access to light allowing greater investment in belowground starch storage capacity and therefore a greater ability to cope with fire. Synthesis: Forest pioneer tree species survived best through fire and drought in the savanna compared to the other two vegetation types. This was likely a result of the open‐canopied savanna providing greater access to light, thereby releasing seedlings from light limitation and enabling them to make and store more starch. Fire can be used as a management tool for controlling forest encroachment into savanna as it significantly affects seedling survival. However, if rainfall increases as a result of global change factors, encroachment may be more difficult to control as seedling survival ostensibly increases when the pressure of drought is lifted. We propose B. buonopozense as an indicator species for forest encroachment into savanna in West African forest–savanna transitions.     

Leaf shape and size track habitat transitions across forest–grassland boundaries in the grass family (Poaceae)

Grass leaf shape is a strong indicator of their habitat with linear leaves predominating in open areas and ovate leaves distinguishing forest‐associated grasses. This pattern among extant species suggests that ancestral shifts between forest and open habitats may have coincided with changes in leaf shape or size. We tested relationships between habitat, climate, photosynthetic pathway, and leaf shape and size in a phylogenetic framework to evaluate drivers of leaf shape and size variation over the evolutionary history of the family. We also estimated the ancestral habitat of Poaceae and tested whether forest margins served as transitional zones for shifts between forests and grasslands. We found that grass leaf shape is converging toward different shape optima in the forest understory, forest margins, and open habitats. Leaf size also varies with habitat. Grasses have smaller leaves in open and drier areas, and in areas with high solar irradiance. Direct transitions between linear and ovate leaves are rare as are direct shifts between forest and open habitats. The most likely ancestral habitat of the family was the forest understory and forest margins along with an intermediate leaf shape served as important transitional habitat and morphology, respectively, for subsequent shifts across forest–grassland biome boundaries.     

Differential seedling regeneration patterns across forest–grassland ecotones in two tropical treeline species (Polylepis spp.)

Successful forest expansion into grassland can be limited by seed dispersal and adverse conditions for tree seedlings in the grassland environment. In the high‐elevation Andes, human‐induced fragmentation has exacerbated the patchy distribution of Polylepis forests, threatening their unique biological communities and spurring restoration interest. Studies of Polylepis forest extent in Peru suggest that forest borders have remained stable over the past century despite decreasing anthropogenic disturbance, suggesting that tree seedling recruitment is being limited in the open grassland habitat. We studied natural seedling dispersion patterns of Polylepis sericea and Polylepis weberbaueri (Rosaceae) at forest–grassland edges across a range of environmental conditions to examine seedling recruitment and colonization of grasslands in Huascaran National Park (Peru). Using data from 2367 seedlings found in 48 forest–grassland edge plots (15 m × 15 m) at forest patches between 3900–4500 masl, we employed generalized mixed modelling to identify the significant associations of seedling densities with environmental covariates. In addition, we compared these associations to patterns of adult presence on the landscape. Seedling densities were associated with a combination of variables varying within (distance to forest edge) and among (elevation and dry season solar irradiation) plots across the landscape. For both species, seedling densities decreased with increasing distance away from the forest in a manner consistent with short‐distance seed dispersal by wind. Our results suggest that such short‐distance dispersal may slow forest expansion, but that there also appear to be substantial post‐dispersal limitations to seedling establishment in the grassland. Polylepis sericea densities decreased with elevation, while P. weberbaueri increased with elevation and decreased with solar irradiation. Associations of adult presence with elevation and solar irradiation mirrored those of seedling densities. Management of areas with forest patches dominated by these species should consider these differences in their environmental tolerances, particularly during species selection and zonation for reforestation.     

Nurse abundance determines plant facilitation networks of subtropical forest–grassland ecotone

The theory of complex networks has been recently used to explain ecological associations between nurses and seedlings in plant facilitation systems. The structure of these networks is potentially affected by morphological, ecological and evolutionary factors that can determine the facilitative interactions. In the present study, we evaluate the role of the projected crown area of the nurse plant, the dispersal syndrome, the abundance and the phylogeny relationships with seedlings as drivers of network structure in facilitation networks. To test these parameters, we used an ecotonal forest–grassland system in southern Brazil that experienced historical forest expansion over the grassland in the last few centuries. In two State Parks, Guartelá (GUA) and Vila Velha (VVL), we sampled tree species (nurses and seedlings) along four transects parallel to the forest edge, with five sampling stations along each transect. Tree seedlings (height >0.3 m) were sampled below the nurse's crown (i.e. isolated woody plants, height >1.5 m) and over the grassland area. A total of 160 nurses and 358 seedlings were sampled. Seedling abundance and richness were greater below a nurse's crown than on open grassland. Nurse abundance was the best predictor of the observed interaction matrix. Probabilistic matrices based on abundance, abundance and phylogeny, and dominance were able to predict the observed nestedness values, and they were also closest to the observed connectance, although all of the probabilistic matrices have underestimated this property. Specialization was predicted by abundance and phylogeny, abundance and projected crown area drivers for VVL, and the abundance based models were closest to the predicted specialization for GUA. The result of a similar pattern in predictability between sites indicates that the influences of morphology and evolutionary and ecological processes over the facilitation interactions are equivalent on a regional scale. Woody plant abundance was a key factor for the facilitation networks, driving forest expansion along the subtropical forest–grassland ecotone.     

Different responses of avian feeding guilds to spatial and environmental factors across an elevation gradient in the central Himalaya

Although elevational patterns of species richness have been well documented, how the drivers of richness gradients vary across ecological guilds has rarely been reported. Here, we examined the effects of spatial factors (area and mid‐domain effect; MDE) and environmental factors, including metrics of climate, productivity, and plant species richness on the richness of breeding birds across different ecological guilds defined by diet and foraging strategy. We surveyed 12 elevation bands at intervals of 300 m between 1,800 and 5,400 m a.s.l using line‐transect methods throughout the wet season in the central Himalaya, China. Multiple regression models and hierarchical partitioning were used to assess the relative importance of spatial and environmental factors on overall bird richness and guild richness (i.e., the richness of species within each guild). Our results showed that richness for all birds and most guilds displayed hump‐shaped elevational trends, which peaked at an elevation of 3,300–3,600 m, although richness of ground‐feeding birds peaked at a higher elevation band (4,200–4,500 m). The Normalized Difference Vegetation Index (NDVI)—an index of primary productivity—and habitat heterogeneity were important factors in explaining overall bird richness as well as that of insectivores and omnivores, with geometric constraints (i.e., the MDE) of secondary importance. Granivore richness was not related to primary production but rather to open habitats (granivores were negatively influenced by habitat heterogeneity), where seeds might be abundant. Our findings provide direct evidence that the richness–environment relationship is often guild‐specific. Taken together, our study highlights the importance of considering how the effects of environmental and spatial factors on patterns of species richness may differ across ecological guilds, potentially leading to a deeper understanding of elevational diversity gradients and their implications for biodiversity conservation.     

Carbon dynamics,net primary productivity and human‐appropriated net primary productivity across a forest–cocoa farm landscape in West Africa

Terrestrial net primary productivity (NPP) is an important metric of ecosystem functioning; however, there are little empirical data on the NPP of human‐modified ecosystems, particularly smallholder, perennial crops like cocoa (Theobroma cacao), which are extensive across the tropics. Human‐appropriated NPP (HANPP) is a measure of the proportion of a natural system's NPP that has either been reduced through land‐use change or harvested directly and, previously, has been calculated to estimate the scale of the human impact on the biosphere. Additionally, human modification can create shifts in NPP allocation and decomposition, with concomitant impacts on the carbon cycle. This study presents the results of 3 years of intensive monitoring of forest and smallholder cocoa farms across disturbance, management intensity, distance from forest and farm age gradients. We measured among the highest reported NPP values in tropical forest, 17.57 ± 2.1 and 17.7 ± 1.6 Mg C ha^?1 year^?1 for intact and logged forest, respectively; however, the average NPP of cocoa farms was still higher, 18.8 ± 2.5 Mg C ha^?1 year^?1, which we found was driven by cocoa pod production. We found a dramatic shift in litterfall residence times, where cocoa leaves decomposed more slowly than forest leaves and shade tree litterfall decomposed considerably faster, indicating significant changes in rates of nutrient cycling. The average HANPP value for all cocoa farms was 2.1 ± 1.1 Mg C ha^?1 year^?1; however, depending on the density of shade trees, it ranged from ?4.6 to 5.2 Mg C ha^?1 year^?1. Therefore, rather than being related to cocoa yield, HANPP was reduced by maintaining higher shade levels. Across our monitored farms, 18.9% of farm NPP was harvested (i.e., whole cocoa pods) and only 1.1% (i.e., cocoa beans) was removed from the system, suggesting that the scale of HANPP in smallholder cocoa agroforestry systems is relatively small.     

Land use types influenced avian assemblage structure in a forest–agriculture landscape in Ghana

The conservation of biodiversity within tropical forest regions does not lie only in the maintenance of natural forest areas, but on conservation strategies directed toward agricultural land types within which they are embedded. This study investigated variations in bird assemblages of different functional groups of forest‐dependent birds in three agricultural land types, relative to distance from the interior of 34 tropical forest patches of varying sizes. Point counts were used to sample birds at each study site visited. Data from counts were used to estimate species richness, species evenness, and Simpson's diversity of birds. Mean species richness, evenness, and diversity were modeled as responses and as a function of agricultural land type, distance from the forest interior and three site‐scale vegetation covariates (density of large trees, fruiting trees, and patch size) using generalized linear mixed‐effect models. Mean observed species richness of birds varied significantly within habitat types. Mean observed species richness was highest in forest interior sites while sites located in farm centers recorded the lowest mean species richness. Species richness of forest specialists was strongly influenced by the type of agricultural land use. Fallow lands, density of large trees, and patch size strongly positively influenced forest specialists. Insectivorous and frugivorous birds were more species‐rich in fallow lands while monoculture plantations favored nectarivorous birds. Our results suggest that poor agricultural practices can lead to population declines of forest‐dependent birds particularly specialist species. Conservation actions should include proper land use management that ensures heterogeneity through retention of native tree species on farms in tropical forest‐agriculture landscapes.     

Fire weather risk differs across rain forest—savanna boundaries in the humid tropics of north‐eastern Australia

Alternative stable state theory has been applied to understanding the control by landscape fire activity of pyrophobic tropical rain forest and pyrophytic eucalypt savanna boundaries, which are often separated by tall eucalypt forests. We evaluate the microclimate of three vegetation types across an elevational gradient and their relative fire risk as measured by McArthur's Forest Fire Danger Index (FFDI). Microclimatic data were collected from rain forest, tall eucalypt forest and savanna sites on eight vegetation boundaries throughout the humid tropics in north Queensland over a 3‐year period and were compared with data from a nearby meteorological station. There was a clear annual pattern in daily FFDI with highest values in the austral winter dry season and lowest values in the austral summer wet season. There was a strong association of the meteorological station FFDI values with those from the three vegetation types, albeit they were substantially lower. The rank order of FFDI values among the vegetation types decreased from savanna, tall eucalypt forest, then rain forest, a pattern that was consistent across each transect. Only very rarely would rain forest be flammable, despite being adjacent to highly flammable savannas. These results demonstrate the very strong effect of vegetation type on microclimate and fire risk, compared with the weak effect of elevation, consistent with a fire–vegetation feedback. This study is the first demonstration of how vegetation type influences microclimate and fire risk across a topographically complex tropical forest–savanna gradient.     

Seasonal and spatial variations of fish trophic guilds in a shallow,semi-enclosed tropical estuarine bay

A study on seasonal and spatial variations of feeding habits and trophic guilds of dominant fish species in Pattani Bay during March 2003 to February 2004 was aimed at classifying diet composition, identifying dominant food components of each species, categorizing trophic guilds of the community and evaluating effects of habitat characteristics and seasonality on guild organization. Most fishes showed high food intake, fed on a diverse range of food items but relied heavily on calanoid copepods and shrimps. All species, with the exception of Epinephelus coioides, were classified as specialist feeders. Four main dietary guilds were classified. Three of these were classified as the guilds dominated by at least two major food items. Significant variations in trophic guilds of 28 fish species based on habitat types and seasons were also identified. They could be divided into three seasonal groups and three site-groups and a single site. Trophic organization for each season and habitat ranged from two to four groups. This information identifies groups of fishes that seasonally and spatially utilize different food resources within a semi-enclosed estuarine bay ecosystem.     

Small mammal abundance and seed predation across boundaries in a restored‐grazed woodland interface

Passive restoration is an effective tool for the maintenance and conservation of biodiversity. Often areas in recovery are immersed in a matrix of land uses, in which the expansion and intensification of human activities exert new visible pressures at their boundaries. The degree of connectivity between these areas and their peripheral lands can be analyzed by mobile link species, organisms that actively move in the landscape by connecting areas to one another through their functional roles. We focus our design on the interface generated by the long‐term restoration area and surrounding grazing lands. We analyze the changes on boundary structure, small mammal abundance, and on the function of native seed dispersal by these vertebrate species. We captured small mammals and determined seed removal of Prosopis flexuosa at three distances inside and outside a fence that delineates passively restored and currently grazed areas. Our results indicate that small rodents find more suitable habitats at the site under restoration than in grazing lands. The restored‐grazing interface shows a decrease in small mammal abundance from the protected area to the grazed lands. From a functional perspective, an increase in small mammal abundance results in an increase in their seed removal activity with implications for seed fate, because the long‐term recovery of vegetation could enhance seed predation on a native tree species.     

Vertical stratification of feeding guilds and body size in beetle assemblages from an Australian tropical rainforest

Abstract The vertical stratification of insect species assemblages inhabiting tropical rainforests is well established but few have examined whether these patterns are reflected in vertical stratification of body size or feeding guilds. We used Malaise and Flight Interception Traps to sample beetle assemblages from five locations, at both canopy and ground zones of a tropical lowland rainforest site near Cape Tribulation, Australia. Beetles from 4 years of sampling were sorted to Family and morphospecies, and allocated to one of five feeding guilds. Within feeding guilds the number of species and individuals, from canopy‐ and ground‐caught traps were compared. The body lengths of species were measure and compared within feeding guilds and families. Herbivores was the dominant guild but was not the majority of all species or individuals. Most beetle species (69%) were less than 5 mm in length and the mean size of canopy‐caught species was greater than that for ground‐caught species. This was probably due to slightly more species of plant feeders (herbivores and xylophages) present in the canopy, which were significantly larger than saprophages, fungivores and predators. Among feeding guilds, there were few overall canopy–ground differences. These results contrast with species composition results presented elsewhere where strong differences between the canopy and the ground were evident. We suggest that our guild groupings may have been too coarse to detect fine‐scale differences and that resource partitioning may have also masked faunal stratification. We propose that fine‐scale differences in resources between the canopy and the ground, together with strong microclimate gradients, are likely to be important in structuring the vertical stratification of insect assemblages at the level of species, but not with respect to functional groups.     

Seasonal variations in the diversity and abundance of diazotrophic communities across soils

The nitrogen (N)-fixing community is a key functional community in soil, as it replenishes the pool of biologically available N that is lost to the atmosphere via anaerobic ammonium oxidation and denitrification. We characterized the structure and dynamic changes in diazotrophic communities, based on the nifH gene, across eight different representative Dutch soils during one complete growing season, to evaluate the amplitude of the natural variation in abundance and diversity, and identify possible relationships with abiotic factors. Overall, our results indicate that soil type is the main factor influencing the N-fixing communities, which were more abundant and diverse in the clay soils (n=4) than in the sandy soils (n=4). On average, the amplitude of variation in community size as well as the range-weighted richness were also found to be higher in the clay soils. These results indicate that N-fixing communities associated with sandy and clay soil show a distinct amplitude of variation under field conditions, and suggest that the diazotrophic communities associated with clay soil might be more sensitive to fluctuations associated with the season and agricultural practices. Moreover, soil characteristics such as ammonium content, pH and texture most strongly correlated with the variations observed in the diversity, size and structure of N-fixing communities, whose relative importance was determined across a temporal and spatial scale.     

Assessment of potential sources of protozoan contamination between two avian feeding guilds in a conservation area

Occasional screening of food and water quality available to organisms in protected areas could be beneficial to their successful conservation. This is important for areas receiving regular human visitors and exhibiting activities that may be detrimental to ecosystem health. This study determined the intestinal protozoan species harboured by insectivorous and granivorous birds within the Jos Wildlife Park, Nigeria and whether the two avian feeding guilds are more susceptible to protozoan infection through water or food (grass seeds and insects). Special boxes were used to collect faecal samples from trapped birds. Samples were later analysed in the laboratory for protozoans. Both food and water utilised by the birds in the area were microscopically screened. The composition and abundance of intestinal protozoans between the two feeding guilds did not show significant differences. However, considering parasite species individually, the degree of infection by protozoans such as Giardia lamblia was highly related to the range of infection sources that a feeding guild was exposed to. The composition of parasites observed in the two feeding guilds was strongly linked to both water bodies and avian diet obtainable in the Park. These observations showed that birds in the Park are orally susceptible to infection with protozoans, some of which have been reported to cause health implications in birds. Thus, any measure to curtail environmental contamination by the parasites may improve ecosystem health and survival of avian species in the Park.     

Group size of feeding stream case-bearing caddisfly grazers and resource abundance

Several herbivorous insects are known to benefit from feeding in groups; however, little is known about (1) the resource conditions under which herbivorous insects perform group feeding and (2) the optimum population size to get any benefits by group feeding, for example, in terms of growth performance. To test the hypotheses that the benefits from group feeding change with resource level and population size, we performed field investigations and an enclosure experiment using the grazer caddisfly larva Micrasema quadriloba. The field investigations revealed aggregated distributions of larvae (indicator of aggregation, I_δ=4.1±1.55, aggregated density: 12.7±5.3 individuals per 3.1×3.1 cm² (mean±1 SD)) when periphyton was abundant on stream cobbles and random distributions (I_δ=1.0±0.11) when periphyton was scarce. In the enclosure experiment, the relative growth rate (RGR) of the larvae at each population size showed different tendencies at high and low periphyton abundance levels; RGR with abundant periphyton had a convex curve with a peak at intermediate population size, whereas RGR with scarce periphyton decreased linearly with increasing population size. The benefits from group feeding thus changed with resource level; larvae obtained high growth performance by group feeding behavior only when the resource was sufficiently abundant. The present study revealed not only that the optimum group size of larvae increased their growth performance, but also that this optimum group size occurred frequently in the field. We also discuss the mechanisms and benefits of group feeding by case-bearing caddisfly grazers.     

Variation in tropical bird survival across longitude and guilds: a case study from the Amazon

Survival estimates of tropical birds have been used to examine life‐history variation across latitudes and dietary guilds. Here, we provide apparent survival estimates of 31 rainforest species from central Amazonian Brazil and compare our results with estimates from Ecuador, Peru (western Amazonia) and French Guiana (eastern Amazonia) to examine geographic variation in demography. Our averaged annual survival estimate for central Amazonian species (?= 0.59, SE = 0.10) was concordant with previously published estimates from the western Amazon (?= 0.58, SE = 0.02), and eastern Amazon (?= 0.63, SE = 0.06). Two predominate demographic patterns were detected across the study areas: within species, survival was highest or lowest in eastern or western Amazonia, but rarely in central Amazonia. The most striking demographic variation was exhibited by Pithys albifrons, for which apparent survival estimates were nearly twice as high in eastern Amazonia (?= 0.80, SE = 0.06) than in western Amazonia (?= 0.42, SE = 0.06) but intermediate in central Amazonia (?= 0.54, SE = 0.04). Although variation in survival may be associated with differences in life history characteristics, our analysis of flocking guild, body size, and nest architecture revealed only moderate differences in survival associated with nest architecture. These results suggest that geographic variation in population processes may be significant for widespread Amazonian species.     

Seasonal dynamics of avian guilds inside and outside core protected areas in an Andean Biosphere Reserve of southern Chile

Capsule: Many protected forests at high elevations show higher endemic biodiversity than unprotected habitats at low elevations. In seasonal ecosystems, however, harsh environmental conditions during winter may force individuals to move from protected forests to unprotected degraded forests in lowlands.

Aims: To examine how forest types and seasons affect avian diversity and habitat-use guilds in protected and unprotected forests.

Methods: Habitat and seasonal avian surveys were conducted to test forest type and seasonal differences on avian species richness, relative abundances and relative abundance of habitat-use guilds.

Results: Twenty-three avian species (～40% endemics) were recorded. Highland old growth forests showed the highest richness (4.1–4.9 species/point count). The lowest richness was recorded in mid-elevation secondary forests (3.1 species/point count). The highest relative abundances for habitat-specialist understorey users and large tree users were recorded in highland old growth forests. However, the abundance of large tree users was higher in secondary forests at low elevations during colder seasons.

Conclusions: Some species and guilds may move to lower elevations outside protected areas during winter. Increasing degradation of unprotected areas will likely affect the suitability of ‘winter habitat’ for specialist guilds, and thus the ecological processes and source/sink dynamics occurring across boundaries of protected and unprotected areas.