Seedling and Sapling Dynamics of Treefall Pits in Puerto Rico1

Seedling and sapling dynamics in a Puerto Rican rain forest were compared between forest understory and soil pits created by the uprooting of 27 trees during Hurricane Hugo. Soil N and P, organic matter, and soil moisture were lower and bulk densities were higher in the disturbed mineral soils of the pits than in undisturbed forest soils ten months after the hurricane. No differences in N and P levels were found in pit or forest soils under two trees with N–fixing symbionts (Inga laurina and Ormosia krugii) compared to soils under a tree species without N–fixing sym–bionts (Casearia arborea), but other soil variables (Al, Fe, K) did vary by tree species. Forest plots had greater species richness of seedlings (<10 cm tall) and saplings (10–100 cm tall) than plots in the soil pits (and greater sapling densities), but seedling densities were similar between plot types. Species richness and seedling densities did not vary among plots associated with the three tree species, but some saplings were more abundant under trees of the same species. Pit size did not affect species richness or seedling and sapling densities. Recruitment of young Cecropia schreberiana trees (>5 m tall) 45 months after the hurricane was entirely from the soil pits, with no tree recruitment from forest plots. Larger soil pits had more tree recruitment than smaller pits. Defoliation of the forest by the hurricane created a large but temporary increase in light availability. Recruitment of C. schreberiana to the canopy occurred in gaps created by the treefall pits that had lower soil nutrients but provided a longer–term increase in light availability. Treefall pits also significantly altered the recruitment and mortality of many understory species in the Puerto Rican rain forest but did not alter species richness.     

Leaf litter decomposition rates in degraded and fragmented tropical rain forests of Borneo

Previously extensive tracts of primary rain forest have been degraded by human activities, and we examined how the effects of forest disturbance arising from habitat fragmentation and commercial selective logging affected ecosystem functioning in these habitats by studying leaf litter decomposition rates in litter bags placed on the forest floor. The rain forests of Borneo are dominated by trees from the family Dipterocarpaceae, and we compared leaf litter decomposition rates of three dipterocarp species at eight forest fragment sites (area 3–3529 ha) that had different histories of disturbance pre‐fragmentation: four fragments had been selectively logged prior to fragmentation and four had been formed from previously undisturbed forest. We compared these logged and unlogged forest fragments with sites in continuous forest that had been selectively logged (two sites) and fully protected and undisturbed (two sites). After 120 d, undisturbed continuous forest sites had the fastest rates of decomposition (52% mass loss). Forest fragments formed from unlogged forest (32% mass loss) had faster decomposition rates than logged forest fragments (28% mass loss), but slower rates than continuous logged forest (39% mass loss). Leaves of a light‐demanding species (Parashorea malaanonan) decomposed faster than those of a shade‐tolerant species (Hopea nervosa), but decomposition of all three dipterocarp species that we studied responded similarly to logging and fragmentation effects. Reduced decomposition rates in logged and fragmented forest sites may affect nutrient cycling and thus have detrimental consequences for forest regeneration. Conservation management to improve forest quality should be a priority, particularly in logged forest fragments.     

Disturbance Level Determines the Regeneration of Commercial Tree Species in the Eastern Amazon

The effects of reduced‐impact logging (RIL) on the regeneration of commercial tree species were investigated, as long‐term timber yields depend partly on the availability of seedlings in a managed forest. On four occasions during a 20‐month period in the Tapajós National Forest (Eastern Amazon, Brazil), seven commercial tree species were assessed as follows: the long‐lived pioneers Bagassa guianensis and Jacaranda copaia; the partially shade‐tolerant Hymenaea courbaril, Dipteryx odorata, and Carapa guianensis; and the totally shade‐tolerant Symphonia globulifera and Manilkara huberi. In 2439 10 × 10 m plots, all individuals < 20 cm diameter at breast height (dbh) were assessed over three intervals, before, during, and after the forest being logged. Before logging, the density of seedlings and saplings of the seven species did not change. Logged trees were spatially aggregated, with 9.2 percent of the plots being heavily impacted by logging. After logging, the recruitment rate increased more than the mortality rate, so that post‐harvesting densities of seedlings and saplings increased. The increase in density was concentrated in logged plots with more disturbances. It is concluded that post‐harvesting heterogeneity of micro‐environments created by RIL may be an important component to be taken into account for sustainable forest management and conservation of commercial species.     

The confounding effects of logging on tree seedling growth and herbivory in Central Amazon

Light availability is an important modulator of seedling growth and plant–herbivore dynamics. Logging increases light levels in forests, potentially altering herbivore–plant interactions that drive seedling establishment. We conducted a transplant experiment to evaluate how logging and herbivory affect seedling growth and survival in three shade‐tolerant tree species, at paired canopy gap and understory sites in logged forest and an adjacent unlogged area in central Amazonia (Brazil). Seedlings were either left exposed to naturally occurring insect herbivores or protected from insects by a fine netting structure. We measured the herbivore damage and growth rate of seedlings after 18 mo. In general, logged areas received more light than unlogged sites. Growth and herbivory rates were positively influenced by light, and herbivory was also influenced positively by logging. In gaps, increased growth mitigated foliar damage. Logging resulted in a loss of foliar tissue due to increased herbivory. Herbivory rates were higher in the understory of logged sites than in that of unlogged understory sites, but growth was similar in these areas. Thus, the understory of logged areas provided the least favorable sites for shade‐tolerant tree regeneration, due to higher herbivory rates. The effect of logging on biotic interactions can extend beyond the gaps it creates into untouched understory sites. To our knowledge, this is the first time such a pattern has been observed, highlighting the importance of evaluating the impact of logging on biotic interactions.     

Response of bird assemblages to windstorm and salvage logging — Insights from analyses of functional guild and indicator species

Natural disturbances, such as fire, windstorms and insect outbreaks, are important drivers of biodiversity in forest ecosystems, but at the same time cause large economic losses. Among the natural disturbances in Europe, windstorms cause the highest economic loss. After such storms, damaged forest stands are commonly salvage logged to restore economic value. However, such interventions could affect species assemblages of various taxonomic groups, including breeding birds. Despite these potential effects, investigations of the impacts of post-storm logging are largely lacking. We thus investigated assemblages of breeding birds in 21 logged and 21 unlogged windstorm-disturbed forest plots and 18 undisturbed, control forest plots using fixed-radius point-stop counts three, five and seven years after a windstorm within the Bavarian Forest National Park as part of the European Long-Term Ecosystem Research Network. We recorded 2100 bird individuals of 55 bird species. Bird assemblages were predominantly altered by the consequences of the windstorm and affected only to a minor degree by subsequent logging of storm-felled trees. Nevertheless, bird species richness was significantly reduced by post-storm logging within the first season. In general, the windstorm led to a shift in bird assemblage composition from typical forest species towards open- and shrub-land species. Assemblages of logged and unlogged disturbed plots consisted mainly of long-distance migrants and ground-foraging bird species, whereas assemblages of undisturbed control plots consisted of resident species that forage within vegetation. Both unlogged and logged storm-felled coniferous forest stands were inhabited by endangered or declining bird species, such as Water Pipit (Anthus spinoletta) on logged plots and Eurasian Redstart (Phoenicurus phoenicurus) on unlogged plots. Indicator species analyses suggested that species of unlogged storm disturbed plots depended on storm-created legacies, such as pits and snags, for foraging and nesting. Hence, we recommend reducing post-storm logging of these legacies to support species restricted to unlogged disturbed forest. To increase the diversity of breeding birds on the landscape scale, diverse logged and unlogged post-disturbance stands should be provided, which could be gained by a partial benign neglect strategy of storm disturbed forest stands.     

Shrub Cover Influence on Seedling Growth and Survival Following Logging of a Tropical Forest

Logging in tropical forests may create large canopy openings. These gaps provide suitable conditions for some opportunistic shrubs and herbs to take advantage of the surge in resources and rapidly colonize disturbed sites. This dense plant cover may limit forest regeneration by interfering with tree seedling establishment, growth, and survival by altering the light and nutrients available to seedlings, modifying herbivore behavior, or a number of other factors. In Kibale National Park (Uganda), old logging sites are mainly covered by dense stands of Acanthus pubescens Engl., which appear to inhibit tree regeneration. We wanted to identify the ecological processes underlying this regeneration collapse. To do so, we designed a factorial experiment to evaluate the influences of herbivory and vegetation cover on the growth and survival of tree seedlings. We compared the survival and growth of transplanted tree seedlings in A. pubescens stands and logged forests, in the presence or absence of the understory vegetation layer (logged forest) or vegetation cover (A. pubescens), and with or without herbivory. We found no evidence to support the hypothesis that herbivory is significantly higher under dense A. pubescens cover. Seedling survival was not influenced by the environment. Seedling growth, however, was positively influenced by the removal of A. pubescens, suggesting that changes in resource availability associated with the presence of A. pubescens, may be important for regeneration. Our results suggest that sustained cutting of A. pubescens cover could foster the growth of established seedlings and could lead to tree regeneration and habitat restoration.     

El Niño droughts and their effects on tree species composition and diversity in tropical rain forests

In this study I investigated the effects of the extreme, 1997/98 El Niño related drought on tree mortality and understorey light conditions of logged and unlogged tropical rain forest in the Indonesian province of East Kalimantan (Borneo). My objectives were to test (1) whether drought had a significant effect on tree mortality and understorey light conditions, (2) whether this effect was greater in logged than in undisturbed forest, (3) if the expected change in tree mortality and light conditions had an effect on Macaranga pioneer seedling and sapling densities, and (4) which (a)biotic factors influenced tree mortality during the drought. The 1997/1998 drought led to an additional tree mortality of 11.2, 18.1, and 22.7% in undisturbed, old logged and recently logged forest, respectively. Mortality was highest in logged forests, due to extremely high mortality of pioneer Macaranga trees (65.4%). Canopy openness was significantly higher during the drought than during the non-drought year (6.0, 8.6 and 10.4 vs 3.7, 3.8 and 3.7 in undisturbed, old logged and recently logged forest, respectively) and was positively correlated with the number of dead standing trees. The increase in light in the understorey was accompanied by a 30 to 300-fold increase in pioneer Macaranga seedling densities. Factors affecting tree mortality during drought were (1) tree species successional status, (2) tree size, and (3) tree location with respect to soil moisture. Tree density and basal area per surface unit had no influence on tree mortality during drought. The results of this study show that extreme droughts, such as those associated with El Niño events, can affect the tree species composition and diversity of tropical forests in two ways: (1) by disproportionate mortality of certain tree species groups and tree size classes, and (2) by changing the light environment in the forest understorey, thereby affecting the recruitment and growth conditions of small and immature trees.     

Factors Limiting Post‐logging Seedling Regeneration by Big‐leaf Mahogany (Swietenia macrophylla) in Southeastern Amazonia,Brazil, and Implications for Sustainable Management1

Post‐logging seedling regeneration density by big‐leaf mahogany (Swietenia macrophylla), a nonpioneer light‐demanding timber species, is generally reported to be low to nonexistent. To investigate factors limiting seedling density following logging within the study region, we quantified seed production rates, germinability, dispersal patterns, and seed fates on the forest floor through germination and the first seedling growing season in southeastern Amazonia, Brazil. Fruit production rates were low by three logged and one unlogged populations compared to reports from other regions. Commercial‐sized trees (>60 cm diameter) were more fecund than noncommercial trees (30–60 cm diameter) at two sites, averaging 14.5 vs. 3.9 fruits/tree/year, respectively, at Marajoara, a logged site, over 8 yr. Fruit capsules contained an average of 60.3 seeds/fruit, 70 percent of which appeared viable by visual inspection. Sixty‐seven to 72 percent of apparently viable seeds germinated in nursery beds 2.5 mo after the dispersal period, when wet season rains began. Dry season winds blew most seeds west‐northwest of parent trees, with median dispersal distances of 28 and 9 m on west and east sides of parent trees, respectively. Nearly 100 percent of seeds fell within an area of 0.91 ha. On the forest floor beneath closed canopies, mammals, invertebrates, and fungal pathogens killed 40 percent of apparently viable seeds, while 36 percent germinated. Nine months after seedling establishment—midway through the first logging season following seed dispersal—14 percent of outplanted seeds survived as seedlings, representing 5.8 seeds/fruit. We conclude that seedlings are likely to survive in logging gaps at appreciable densities only in rare cases where previous year fruit production rates by logged trees were high (4–12.5% of commercial‐sized trees/year at Marajoara) and where tree crowns were felled in west or northwest directions.     

Natural regeneration of selected timber species in the Republic of Congo

Natural regeneration of timber species is critical to the sustainable management of tropical forests. To understand what determines regeneration success of timber species in the Congo Basin, we evaluated whether seedling recruitment rates differed between forest logged 30 years previously and unlogged forest and determined the environmental factors that influence seedling density, growth and survival. We monitored the fate of 2186 seedlings of seven timber species within 462, 25‐m² plots located along 21 transects. We characterized seedling plots by light availability, soil nutrient availability and pH, and abundance of mammalian herbivores and then used linear and generalized linear mixed models to evaluate the variables that influenced seedling density, growth and survival. Light availability and canopy openness were 18% and 81% higher in logged than unlogged forest, and concentration of soil nutrients varied between sites. Seedling density was 32% higher in unlogged than logged forest. Taking all species together, seedling survival was positively correlated with calcium and negatively with magnesium and available phosphorus. Rates of seedling growth increased with available light. Taken separately, seedlings of the selected timber species responded differently to abiotic and biotic factors, demonstrating species‐specific regeneration requirements.     

Effects of Selective Logging on Populations of Two Tropical Understory Herbs in an Amazonian Forest1

This study examined the effects of (1) rime elapsed after logging and (2) logging intensity on the total, juvenile, and adult densities, and on reproduction of Calathea altissima and Ischnosiphon amuma (Matantaceae), two monocot herb species in the understory of Central Amazonian forests. Logging was carried out experimentally at various intensities in eight plots (4 ha each) during 1987 and three plots in 1993. Five plots were left as experimental controls. In 1998, total and adult densities of the two species were greater in the plots logged 11 years before than in controls. For I. arouma, total and adult densities were also higher in the plots logged 5 years before; however, juvenile densities were less affected. The intensity of logging influenced adult density of both species but not total or juvenile densities. Densities of reproductive individuals of both species were higher in the logged areas and increased with logging intensity. The effects of time after logging and logging intensity on reproduction were indirect due to the greater number of adult plants in those areas. The observed effects were probably mediated by changes in canopy cover in logged areas. Greater light intensities in logging gaps, roads, and their margins may lead to increased reproduction, recruitment, growth, and survival, which in turn can lead to increased plant density. These results indicate that logging has long–term effects on understory plant populations.     

Artificial canopy gaps accelerate restoration within an exotic Pinus radiata plantation

We created small‐scale artificial canopy gaps to accelerate the growth of mature indigenous forest canopy species for restoration of an 18‐year‐old exotic Pinus radiata plantation forest, in the Marlborough Sounds, New Zealand. Small and large circular gaps were formed by felling. Seedlings of two indigenous forest canopy species, Podocarpus totara (Podocarpaceae) and Beilschmiedia tawa (Lauraceae), were planted within artificial gaps and undisturbed plantation canopy. Seedling height growth, mortality, and occurrence of animal browse were monitored at approximately 6‐month intervals over 17 months. Both P. totara and B. tawa differed significantly in height growth and in animal browse occurrence among artificial gap treatments. Growth of the light‐demanding P. totara was better under large canopy gaps, whereas growth of the shade‐tolerant B. tawa increased under gaps of any size but was most consistent under small gaps. For P. totara, any significant restoration benefit of gap formation on height growth was lost when browsed seedlings were taken into account. Animal browse significantly limited B. tawa height growth in large but not in small gaps. Small‐scale canopy gap creation is an effective method of modifying light transmission to the plantation understorey and accelerating seedling growth rates. Canopy gap size can be used to optimize understorey illumination according to species‐specific light requirements. The increased occurrence of animal browse in gaps requires consideration. Artificial canopy gaps within planted monocultures create structural heterogeneity that would otherwise take an extended period of time to develop. These results further support the role of plantations as indigenous forest restoration sites.     

Effects of fire and selective logging on the tree species composition of lowland dipterocarp forest in East Kalimantan, Indonesia

Tree species composition (diameter at breast height (dbh) 10 cm) was studied in primary, selectively logged and heavily burnt forests in East Kalimantan, Indonesia. The number of trees, tree species, and the Fishers's- diversity index were determined for the first 15 years (burnt forest) and 25 years (selectively logged forest) after disturbance. Additionally the population structure of six common and typical Macaranga pioneer tree species was compared through time between selectively logged, burnt and primary forest. Both selectively logged and burnt forest showed a significant reduction in number of trees and tree species per surface area directly after disturbance. Fire especially affected dominant tree species, while for selective logging the opposite was observed. In selectively logged forest the number of trees, tree species and the Fishers's- index reached pre-disturbance levels within c. 15 years. For burnt forest, only the number of trees recovered to pre-disturbance levels. The number of tree species stayed constant after disturbance, while the Fishers's- index decreased. The six studied Macaranga pioneer tree species seedlings were present in all forest types. Their density seems to be unrelated to light levels in the forest understorey but strongly related to the number of mature parent trees. Their sapling densities were strongly related to light levels in the forest understorey. The studied Macaranga species formed an important part of both under- and over-storey in burnt forest 15 years after disturbance, while they were almost absent in the understorey and only moderately common in the overstorey of selectively logged forest.     

Arrested succession in logging gaps: is tree seedling growth and survival limiting?

Thirty years after selective timber harvest in the Kibale National Park, Uganda, many abandoned logging gaps are dominated by Acanthus pubescens, and show little forest recovery. To examine if this arrested successional state was caused by limited tree seedling growth and survival, we planted seedlings of four forest tree species (Albizia grandibracteata, Mimusops bagshawei, Prunus africana and Uvariopsis congensis) in A. pubescens‐dominated logging gaps and in control areas of adjacent forest. To assess if clearing A. pubescens facilitates forest regeneration, we planted seedlings of two species (A. grandibracteata and U. congensis) in small clearings cut within the logging gaps. We examined mortality, growth, herbivory and site characteristics among the treatments. Finally, we described the physical attributes of the A. pubescens‐dominated gaps. Seedlings of all the four species survived and grew equally well in A. pubescens and forest treatments, and most site characteristics were also similar. Seedlings planted in clearings grew more than in either forest or A. pubescens sites. Very few established trees were found in A. pubescens sites, and most of these were near the forest edges. We also discussed the role of elephants (Loxodonta africana) and collapsing A. pubescens canopies in the maintenance of an arrested successional state in these logging gaps.     

Logging and Edaphic Factors in the Invasion of an Asian Woody Vine in a Mesic North American Forest

Factors contributing to an invasion of the Asian woody vine Celastrus orbiculatus in a mesic forested Massachusetts (USA) sanctuary were investigated. Planting germinated C. orbiculatus seeds in the field revealed that, although they tolerate a wide range of conditions, seedlings grow largest in moist circumneutral soil under high irradiance. A study of naturally occurring vines in a logged forest suggested that C. orbiculatus invaded two years after harvest, and invasion may have been triggered by logging disturbance. Vine stems were more abundant on former logging roads than in surrounding selectively logged areas. Former logging roads had a significantly higher soil pH than surrounding areas and, due to the direction of construction, a significantly more southerly aspect. Fourteen years after harvest, logging roads continued to provide this aggressive exotic with superior habitat for establishment and growth.     

Changes in liana community structure and functional traits along a chronosequence of selective logging in a moist semi-deciduous forest in Ghana

ABSTRACT

Background: Lianas are an important component of tropical forests that respond to logging disturbance. Determining liana response to selective logging chronosequence is important for understanding long-term logging effects on lianas and tropical forests.

Aims: Our objective was to quantify the response of liana communities to selective logging chronosequence in a moist semi-deciduous forest in Ghana.

Methods: Liana community characteristics were determined in ten 40 m × 40 m plots randomly and homogenously distributed in each of four selectively logged forest stands that had been logged 2, 14, 40 and 68 years before the surveys and in an old-growth forest stand (ca. >200 years).

Results: Liana species composition differed significantly among the forest stands, as a function of logging time span, while species richness fluctuated along the chronosequence. The abundance of liana communities and of reproductive and climbing guilds was lower in the logged forests than in the old-growth forest. The ratio of liana abundance and basal area to those of trees was similar in the logged forests, but significantly lower than those in the old-growth forest.

Conclusions: Logging impacts on liana community structure and functional traits were largely evident, though no clear chronosequence trends were recorded, except for species composition.     

Hyperspectral remote detection of niche partitioning among canopy trees driven by blowdown gap disturbances in the Central Amazon

Advanced recruitment and neutral processes play important roles in determining tree species composition in tropical forest canopy gaps, with few gaps experiencing clear secondary successional processes. However, most studies are limited to the relatively limited spatial scales provided by forest inventory plots, and investigations over the entire range of gap size are needed to better understand how ecological processes vary with tree mortality events. This study employed a landscape approach to test the hypothesis that tree species composition and forest structural attributes differ between large blowdown gaps and relatively undisturbed primary forest. Spectral mixture analysis on hyperspectral satellite imagery was employed to direct field sampling to widely distributed sites, and blowdown plots were compared with undisturbed primary forest plots. Tree species composition and forest structural attributes differed markedly between gap and non-gap sites, providing evidence of niche partitioning in response to disturbance across the region. Large gaps were dominated by classic Neotropical pioneer genera such as Cecropia and Vismia, and average tree size was significantly smaller. Mean wood density of trees recovering in large gaps (0.55 g cm⁻³) was significantly lower than in primary forest plots (0.71 g cm⁻³), a difference similar to that found when comparing less dynamic (i.e., tree recruitment, growth, and mortality) Central Amazon forests with more dynamic Western Amazon forests. Based on results, we hypothesize that the importance of neutral processes weaken, and niche processes strengthen, in determining community assembly along a gradient in gap size and tree mortality intensity. Over evolutionary time scales, pervasive dispersal among colonizers could result in the loss of tree diversity in the pioneer guild through competitive exclusion. Results also underscore the importance of considering disturbance processes across the landscape when addressing forest carbon balance.     

Different responses to shade of evergreen and deciduous oak seedlings and the effect of acorn size

An evergreen oak species, Cyclobalanopsis multinervis, and a deciduous oak species, Quercus aliena var. acuteserrata were grown from acorns under two light levels (full sunlight and shade at about 18 % of full sunlight, simulating the light intensities in forest clearings and gaps, respectively) for one growing season. Three hypotheses were tested: (i) the deciduous species grows faster than the evergreen species in forest gaps and clearings; (ii) the deciduous species responds more strongly in terms of growth and morphology to variation in light climate than the evergreen species; and (iii) seedling size is positively correlated to acorn size. The results showed: (i) at both light levels, the deciduous seedlings gained significantly more growth in biomass and height than the evergreen seedlings; (ii) both species produced significantly more biomass in full sunlight than in shade, without showing any significant difference in height between treatments. Increase in light intensity improved the growth of the deciduous seedlings more strongly; (iii) at a similar age, the deciduous seedlings showed a greater response in leaf morphology and biomass allocation to variation in light levels, but when compared at a similar size, biomass allocation patterns did not differ significantly between species; (iv) bigger acorns tended to produce larger seedlings, larger leaf sizes and more leaf area, between and within species. These differences demonstrate that the deciduous species is gap-dependent and has the advantage over the evergreen species in forest gaps and clearings.     

Tree seed germination and seedling establishment in treefall gaps and understorey in a subtropical forest of northeast India

Abstract Seed germination, and survival and growth of seedlings of four dominant tree species, Quercus dealbata, Quercus griffithii, Quercus glauca and Schima khasiana were studied in the treefall gaps and forest understorey of an undisturbed mature-phase humid subtropical broadleaved forest in northeast India. Three important microenvironmental factors namely photosynthetically active radiation (PAR), soil moisture and litter depth, were also measured in the forest understorey and gaps and correlated with seedling mortality. Seed germination of S. khasiana was significantly higher in the treefall gaps than in the understorey; among the tree species studied, it had the highest germination. Quercus seedlings were abundant in the understorey and small gaps, while S. khasiana seedlings were more numerous in the large gaps. The survivorship curves for the seedling populations revealed that the three Quercus species survived better in the understorey, while S. khasiana did so in the gaps. PAR and soil moisture were positively correlated with tree seedling mortality, which occurred mainly during the winter months. The Quercus seedlings grew better in the forest understorey and small gaps and S. khasiana seedlings in the large gaps. The differential performance of the tree seedlings to the conditions prevailing in the understorey and gaps of two sizes indicates that different species were adapted to different light environments depending upon their optimum requirements. This could be an effective mechanism for promoting species coexistence in the forest community.     

Disturbance, selective logging and bird diversity: a Neotropical forest study

The composition and structure of bird communities, and the damage to forest structure were surveyed in northern French Guiana (northeastern Amazonia) one year and ten years after selective logging and compared with the situation in a similar undisturbed primary forest. A point-count method was used in which 937 0.25ha sample plots were censused for 20 minutes each, and their vegetation structure was measured. On average, 38% of the forest undergrowth was destroyed, then invaded by dense regrowth, and up to 63% of the canopy was substantially opened as a result of selective logging. Hunting pressure also increased due to access roads opened for logging. Among the 256 species recorded, overall bird species richness and abundance were depressed by 27–34% in the logged areas compared to primary forest. The most vulnerable guilds, which decreased by 37–98% in abundance, were mature forest understorey species, especially terrestrial ones and mixed flocks of insectivores. Hummingbirds, small gaps, vine tangles and canopy species did not decrease, nor increase significantly after logging. Only species naturally associated with dense second growth, forest edges and large gaps actually increased. Habitat specialization was the major determinant of vulnerability to logging, and, to a lesser degree, size (large) and diet (insectivorous), but foraging behaviour and rarity had little effect. Bird sensitivity to changes in logged forest structure may involve physiological intolerance, reduced food categories, increased exposure to predators, too dense understorey for their specific foraging behaviour and/or avoidance of gaps. Suggested improvements of current forest management and logging techniques for the maintenance of a higher proportion of the original biodiversity include minimizing logging damages, long rotations (>50 years) between cuts, and keeping unlogged forest patches within logging concessions.