Background: Edge effects cause changes in bird community richness, abundance, and/or distribution within a land- scape, but the avian guilds most influenced can vary among regions. Although Southeast Asia has the highest rates of deforestation and projected species loss, and is currently undergoing an explosive growth in road infrastructure, there have been few studies of the effects of forest edges on avian communities in this region. Methods: We examined avian community structure in a dry evergreen forest in northeastern Thailand adjacent to a five-lane highway. We evaluated the richness and abundance of birds in 11 guilds at 24 survey points on three parallel transects perpendicular to the edge. At each point, 10-min surveys were conducted during February‒August 2014 and March‒August 2015. Vegetation measurements were conducted at 16 of the bird survey points and ambient noise was measured at all 24 survey points. Results: We found a strongly negative response to the forest edge for bark-gleaning, sallying, terrestrial, and under- story insectivores and a weakly negative response for arboreal frugivore-insectivores, foliage gleaning insectivores, and raptors. Densities of trees and the percentage canopy cover were higher in the interior, and the ambient noise was lower. In contrast, arboreal nectarivore-insectivores responded positively to the forest edge, where there was a higher vegetation cover in the ground layer, a lower tree density, and a higher level of ambient noise. Conclusion: Planners should avoid road development in forests of high conservation value to reduce impacts on biodiversity. Where avoidance is impossible, a number of potential mitigation methods are available, but more detailed assessments of these are needed before they are applied in this region. Keywords: Road edge, Avian guilds, Dry evergreen forest, Thailand Background conservation value of intact forest landscapes (Potapov Forest habitat conversion and degradation, which was et al. 2017) and bringing an increasing proportion of the already widespread globally by the mid-18th century, remaining forest closer to edges (Laurance et al. 2015). has continued to increase, especially in the tropics over Edges can act as barriers for birds and can affect the past half century (Haddad et al. 2015). The result - genetic diversity, species distributions, species abun- ing fragmentation means that much of the remaining dances, and nest survival, which can lead to local spe- forest is now potentially subject to edge effects (Barber cies extinctions (Newmark and Stanley 2011; Mammides et al. 2014; Haddad et al. 2015). Infrastructure develop- et al. 2015). Edges alter the physical environment of for- ment, especially roads and reservoirs, is projected to est habitats via increased sunlight, temperature extremes, continue to increase in the coming decades, reducing the and wind exposure, and reduced humidity, directly influencing vegetation structure and food availability, which may, in turn, cause changes in the avian commu- nity. Such changes may render edge habitats unsuitable *Correspondence: firstname.lastname@example.org Conservation Ecology Program, School of Bioresources and Technology, for some bird species (Murcia 1995). Roadside edges King Mongkut’s University of Technology Thonburi, Bangkok, Thailand not only change the physical environment, but also lead Full list of author information is available at the end of the article © The Author(s) 2018. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creat iveco mmons .org/licen ses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creat iveco mmons .org/ publi cdoma in/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Khamcha et al. Avian Res (2018) 9:20 Page 2 of 13 to increased traffic noise and potential road mortality Collar 2002; Lee et al. 2005; Moradi et al. 2009; Day- that may directly impact populations, distributions, and ananda et al. 2016); most ecological and conservation behaviours of some bird species (Halfwerk et al. 2011; studies in the tropics are conducted in continuous pri- Jack et al. 2015). mary forest and forest interiors. In view of the high rate Edge preference or avoidance by birds probably of forest loss and rapid increase in infrastructure devel- depends on multiple traits, including biogeographic opment in Southeast Asia, however, a better under- origin, trophic guild, body size, degree of habitat spe- standing of how avian communities respond to forest cialization, reproductive rates (Ewers and Didham 2006; edges and which mechanisms influence this response Barbaro and van Halder 2009; Vetter et al. 2011). Avian is needed to improve forest management and mitigate guilds respond to edge effects differently. Species that potentially adverse impacts from road construction. require forest interior for foraging or breeding, as well as The aim of this study, therefore, was to investigate the those with specific or specialized diets or foraging behav - effects of a roadside forest edge on avian community iours, may avoid edges because the altered vegetation structure in a Southeast Asian tropical forest. Our spe- structure or microclimate, higher anthropogenic noise cific objectives were to (1) determine how the forest levels, and/or higher predation pressure or brood para- edge affected the avian community and (2) identify the sitism, may affect their foraging habitats (Newton 1994; potential mechanisms that influence these edge effects. Menke et al. 2012). Based on previous studies in tropical regions (Watson The responses of avian communities to edge effects et al. 2004; Mammides et al. 2015), we predicted that vary from site to site in tropical forest habitats (Restrepo (1) edge effects would influence each avian guild dif - and Gomez 1998; Watson et al. 2004; Vetter et al. 2013). ferently, and (2) richness and abundance of some avian Certain guilds in tropical forests, particularly insecti- guilds, particularly understory, sallying, terrestrial, and vores, ground foragers, and bark gleaners, often occur at bark gleaning insectivores, would be higher in the for- lower densities close to edges (Lambert and Collar 2002; est interior than the forest edge because the interior is Mammides et al. 2015). In temperate forests, in contrast, likely to provide more suitable habitats. insectivorous birds make greater use of edges because higher plant productivity leads to higher abundance of arthropod larva (Barber and Marquis 2011; Terraube Methods et al. 2016). Edge habitats in tropical forests may also con- Study area tain more food resources, but other mechanisms, such The study was conducted in the Sakaerat Environmental as altered vegetation structure and microclimates, and Research Station (SERS), which is part of the Phuluang increased predation, may counter these positive effects Non-hunting Area, with a total forest area of 160 km . (Flaspohler et al. 2001; Pollock et al. 2015). Indeed, sev- SERS was declared a UNESCO Biosphere Reserve in eral studies, from both tropical and temperate regions, 1977. It is located in northeastern Thailand (14°30ʹN, have found that nectarivores and frugivores were com- 101°55ʹE), with an elevation of 280–762 m asl. The aver - monly closer to edges (Dale et al. 2000; Laurance 2004; age annual rainfall is approximately 1200 mm, with the Lindell et al. 2007). This may reflect vegetation structure dry season (average rainfall 220 mm) generally occur- at the edge which supports higher food availability (i.e. ring from November to April and the wet season (aver- fruit and nectar) compared with the interior (Bierregaard age rainfall 920 mm) from May to October. The average and Stouffer 1997; Restrepo et al. 1999). Utilization of temperature is 27 °C, ranging from 19 to 36 °C over the edge habitats by some species appears to be a trade-off year. SERS has a five-lane highway to the south, vil - between the risks of predation and/or the unfavourable lages and agricultural areas to the east and west, and the physical or biological conditions versus preferred forag- northeast is connected to the Lam Phra Phloeng Dam. It ing habitats. includes two major forest types, dry evergreen forest cov- Southeast Asia has the highest rate of forest loss in ering 70% of the area and dry dipterocarp forest covering the tropics (Stibig et al. 2014; Wilcove et al. 2013) as 20%, with the rest consisting of small patches of bamboo, result of industrial scale logging, the expansion of mon- plantations, and grassland. The study focused on an area oculture cash crops and, more recently, an explosion within dry evergreen forest ranging in elevation from 355 in infrastructure construction (Laurance et al. 2015; to 523 m asl within a kilometre of the forest edge (Fig. 1). Sloan and Sayer 2015). Although there have been sev- This edge is “hard”, in the sense that the forest ends eral studies of edge effects on avian communities, most abruptly at a five-lane highway (Route 304). The road has have been conducted in temperate regions (Bennett been in place for over 60 years but there has been recent 2017). Few studies have been conducted in the tropics, widening along part of its length with consequent loss of especially tropical Asia (Wong et al. 1998; Lambert and native vegetation on both sides. Khamcha et al. Avian Res (2018) 9:20 Page 3 of 13 Fig. 1 The location of the study area at the Sakaerat Environmental Research Station, Thailand in 2014–2015. Insets show the area’s location in Thailand (upper left), the boundary of study area and landscape context (upper right) and details of the study area including the road edge (Route 304) and survey points (dots) (below) Khamcha et al. Avian Res (2018) 9:20 Page 4 of 13 Bird surveys surveyed in 2014. Within each plot we counted the num- We conducted bird surveys using point-counts during ber of stems in three height classes (0.5–3 m, > 3–5 m the breeding season February–August 2014 and March– and > 5 m) in order to estimate stem density (stems/ha) August 2015. Our bird sampling units consisted of sur- and measured the DBH (diameter breast height) of all vey points (16 points during 2014, 24 points in 2015, trees with DBH > 5 cm in order to estimate the basal area 24 in total) set up along parallel transects (eight points/ (m /ha). The percentage vegetation cover was estimated transect). Transects were 700 m apart. The total area for five plant height classes: < 0.5 m, > 0.5–3 m, > 3–5 m, > sampled was approximately 140 ha of dry evergreen for- 5–10 m and > 10 m. est and was adjacent to the highway (Fig. 1). The points were arranged to sample a gradient of distances 0, 120, Traffic noise 240, 360, 480, 600, 720, and 840 m from the edge; thus, Ambient noise is expected to be high close to the road in total our design resulted in three points at each dis- edge, which could affect the bird community close to tance from the edge. Survey points were > 700 m from the road edge. To investigate the potential effects of traf - other forest types (Fig. 1). Each point was surveyed fic noise on avian guilds, ambient noise measurements for 10 min during each survey (Gale et al. 2009). To were conducted during bird surveys at the 24 bird sur- reduce the bias of time of the day, the starting point vey points during March–August 2015. At each of the was reversed each month. From February to May 2014 bird survey points we measured ambient noise in deci- we conducted surveys at 8 points (one point at each dis- bels (dB) for 10 min using the application Sound Meter tance from the edge) twice per month, and from June to version 1.6.1 on a smartphone (Lenovo A680) held 1.2 m August 2014 we surveyed 16 points (two points at each above the ground pointed towards the road. We recorded distance) once per month. During 2015 from March the average, minimum and maximum ambient noise for to August we conducted monthly surveys at 24 points each sample period. (three points at each distance) and to increase the num- ber of surveys in July and August 2015 we repeated the Data analysis surveys (a day after completing the initial 24 points) on To assess the effect of the forest edge on the richness eight of these points (one point at each distance), with (number of species observed) and abundance (counts) of the aim of accounting for every species present in our the avian community sampled from the 24 survey points, study area regardless of its detectability (Sliwinski et al. we categorized our avian data into guilds and con- 2016). The detections were recorded as either seen or structed generalized linear mixed models (GLMM) in R heard or both. Birds in flight were not counted. Points using package lme4 (Bolker et al. 2011). GLMM provides were surveyed in the early morning, starting at approxi- a flexible approach for analysing data with a variety of mately 06:00 a.m. (sunrise) and finishing before 09:00 sampling units with uneven and small sample sizes and a.m., as this time has the highest singing rates (Gale et al. repeated samples that allows for random effects. We 2009). All point-count surveys were conducted by one modelled abundance and richness using a Poisson dis- observer to minimize potential variance among observ- tribution. Our models included fixed effects of vegeta - ers (Ralph et al. 1993). tion (vegetation density, stem density of vegetation with We classified birds into 11 guilds following Johns heights between 0.5–3 m, 3–5 m, 5–10 m, and stem (1986) and Lambert and Collar (2002): arboreal frugi- density of trees with DBH > 5 cm, vegetation cover—per- vore-insectivore (AFI), arboreal nectarivore-insectivore centage cover of vegetation at < 0.5 m, 0.5–3 m, 3–5 m, (ANI), bark-gleaning insectivore (BGI), foliage-gleaning 5–10 m and > 10 m, and basal area), ambient noise, and insectivore (FGI), granivore-insectivore (GI), raptor (R), distance from forest edge. Although we collected veg- sallying insectivore (SaI), terrestrial insectivore (TI), ter- etation structure data for only 16 of the 24 bird survey restrial insectivore-faunivore (TIV), terrestrial insecti- points, our preliminary surveys and recent measure- vore-frugivore (TIF), and understory insectivore (UI) ments of vegetation structure in the same study area (Additional file 1: Table S1). found that the vegetation structure followed a consist- ent pattern along our interior-edge distance gradient (R. Vegetation measurements Angkeaw, unpubl. data; W. Petersen, unpubl. data). For Vegetation structure is expected to change along a forest the data analysis we used averaged values to represent edge, which has the potential to be a mechanism to influ - the vegetation at each of the 8 distances sampled from ence the composition of the bird community. To quantify the road. We used avian guilds as a random effect to vegetation structure, we measured the potentially influ - allow for different responses to the edge among guilds. ential vegetation variables; 10-m radius circular plots We used maximum ambient noise in this analysis, as this were established at the 16 bird survey points which were is likely to have the most effect on the avian community Khamcha et al. Avian Res (2018) 9:20 Page 5 of 13 relative to other noise measurements (Ware et al. 2015). such as migratory species; e.g. a group of warblers, cuck- We did not conduct ambient noise measurements in oos and the Siberian Blue Robin (Larvivora cyane). Most 2014, but a traffic survey by the Bureau of Highway species (69%) could be found at every distance from the Safety, Department of Highways, on Route 304 along our edge into the interior, but some species were recorded study area in 2014 (Department of Highway 2014) found at particular distances from the edge, including three a similar traffic volume to our study in 2015 (678 vs. 609 species of woodpeckers, Orange-breasted Trogon (Har- vehicles/h). Since traffic noise is determined by traffic pactes oreskios), Oriental Pied Hornbill (Anthracoceros volume (Arevalo and Newhard 2011), we used our 2015 albirostris), and Thick-billed Green Pigeon (Treron curvi - data to represent the ambient noise in both years. rostra), which were found only in the interior. On the We excluded year and spatial variables (transect and other hand, there were some species found only near the elevation) from this analysis as we found no support for edge, such as prinias (Prinia sp.), Asian Koel (Eudynamys either of them affecting counts or richness. One vari - scolopacea), Scaly-breasted Munia (Lonchura punctu- able of any pair that was highly correlated (r > 0.7) was lata), and Olive-backed Sunbird (Cinnyris jugularis). removed from the analysis. We expected that ambient Total bird species richness and abundance (including noise and vegetation variables would change with dis- both resident and migratory birds) both increased with tance from the edge, so when these factors were cor- distance from the edge, but this relationship was only related with distance, we regarded distance from edge marginally significant for abundance (richness: R = 0.95, as the best predictor of changes in species richness and p < 0.001; abundance: R = 0.69, p = 0.056) (Fig. 2). abundance of avian guilds. To evaluate the effect of the forest edge on each guild, Vegetation structure we generated models with random slopes and inter- Hopea ferrea (Dipterocarpaceae), Lagerstroemia duper- cepts. We generated AIC tables (R-package AICcmod- reana (Lythraceae), and Shorea henryana (Dipterocar- elavg) for model selection based on AIC values. We used paceae) were the dominant species in the canopy layer model averaging to estimate values of parameters if ≥ 1 (> 30 m). The dominant species in the understory layer model was within two ΔAIC of the top model. We con- (< 20 m) were Hydnocarpus ilicifolia (Flacourtiaceae), sidered the evidence of variable influence on species Walsura trichostemon (Meliaceae), Memecylon ova- richness and abundance using 85% confidence intervals, tum (Melastomataceae), and Memecylon geddesianum which are considered more suitable for model selection (Melastomataceae). The total basal area for trees with and parameter-evaluation criteria than the narrower DBH > 5 cm was 28.18 m /ha. The stem density was 95% confidence intervals (Arnold 2010). We then cre - 15,524 stems/ha for stems 0.5–3 m in height, 642 stems/ ated estimated regression lines and calculated estimated ha for stems 3–5 m, and 1176 stems/ha for stems > 5 m. coefficients, standard errors, and confidence intervals The stem density for trees with DBH > 5 cm was 1202 for each guild from the top-ranked model to indicate the stems/ha. Estimated mean vegetation cover by height responses of each guild to the possible edge effects. The layers was 44% < 0.5 m; 72% 0.5–3 m; 55% 3–5 m; predicted response, y (species richness or abundance), for 63% 5–10 m; and 81% > 10 m. each guild, g, to distance from forest edge, X, was calcu- When we considered the vegetation structure from the lated by holding the vegetation variable at its mean, V , a s forest edge into the interior, we found that the total basal shown by the following equation (Gelman and Hill 2007): area was larger further from the edge (Fig. 3). The stem density of trees was also higher at the interior. The stem y = α + α + β V + β X + β XV + β X r rg f ﬁ fx f ∗x density of saplings was higher at both edge and deeper where α = intercept of the fixed effects, f; α = intercept interior areas, relative to intermediate distances (Fig. 3). f r of the random effect, r; β = coefficient for the fixed veg - The percentage cover of the vegetation from the ground etation variable; β = coefficient for distance, X, from to the middle layer was higher closer to the edge in con- fx edge; β = coefficient of interaction between vegetation trast to the percentage cover of vegetation at the canopy f ∗x and distance; and β = coefficient of the random effect layer which was higher further from the edge (Fig. 3). rg for the guild of interest. Traffic noise Results From March–August 2015 the mean ambient noise was Bird species richness and abundance 45 dB (range: 29–82 dB). The mean maximum ambi - A total of 272 independent surveys were conducted dur- ent noise within 100 m of the forest edge was 75 dB ing our two-year study. We recorded 2781 detections and (range: 67–82 dB) and in the interior (at 800 m) was identified 70 species of birds (Additional file 1: Table S1). 58 dB (range: 42–69 dB) (Fig. 3). The mean traffic noise Some species were observed for relatively short periods, on low traffic days was 44 dB (range: 29–78 dB) and Khamcha et al. Avian Res (2018) 9:20 Page 6 of 13 Avian guild responses to forest edge We removed four variables that were correlated (r > 0.7) with distance from edge: basal area and the percentage cov- erage of vegetation at > 10 m height, which were positively correlated with distance from edge, and maximum ambi- ent noise and percentage coverage of vegetation at 3–5 m height, which were negatively correlated with distance from edge. We retained distance from edge as a fixed effect to represent effects of these four variables (Fig. 3 ). We generated 22 models for the effects of forest edge on richness and abundance of avian guilds. The top- ranked models for both species richness and abundance were interaction models that included two variables, dis- tance from edge and stem density of saplings (i.e. stems of height 0.5–3 m), while other models with strong support (ΔAICc < 2) also included the percentage cover of veg- etation at < 0.5 m and 5–10 m (Table 1). Following model averaging, the variables with a significant positive influence on species richness were stem density of saplings (β = 0.33) and the interaction between distance from edge and the percentage cover at < 0.5 m (β = 0.35), while the interaction between distance from edge and stem density of saplings (β = − 0.23) (Table 2) was significantly negatively associ - ated with species richness. For abundance, stem density of saplings (β = 0.49) was significantly positively associated with abundance, while the interaction between distance from edge and stem density of saplings (β = − 0.29) and the percentage cover at 5–10 m (β = − 0.32) had a significant negative influence (Table 2 ). For guild-specific predictions, the models suggested that there was an effect of the distance from edge and vegeta - Fig. 2 Total species richness and average abundance of birds tion structure on richness and abundance of avian guilds, in relation to distance from the forest edge at the Sakaerat and distance to edge also represented ambient noise and Environmental Research Station in 2014–2015 several characteristics of vegetation structure because of correlations (Fig. 3). However, each guild responded dif- ferently to the edge (Fig. 4). We found some evidence to on high traffic days was 45 dB (range: 33–82 dB). Total support negative effects of forest edge on richness of bark- traffic volume on Route 304 was 553–700 vehicles/h gleaning, sallying, terrestrial, and understory insectivores and average traffic volume was 609 vehicles/h. Average and we also found some evidence of negative edge effects traffic volume on the low traffic days was 564 vehicles/h on arboreal frugivore-insectivores (Table 3; Fig. 4a). The (range: 553–582 vehicles/h) and average traffic volume richness of these guilds were higher at greater distances on the high traffic days was 642 vehicles/h (range: 586– from the edge, where the density of saplings, basal area, and 700 vehicles/h). Ambient noise was significantly higher percentage cover of higher vegetation layers were higher, closer to the forest edge, especially within 100 m of but percentage cover of lower layers (0.5–5 m) and ambient the edge, and gradually decreased towards the interior noise were lower (Fig. 3). Edge effects seemed to have posi - (R = 0.90, p = 0.002). tive effects on richness of arboreal nectarivore-insectivores (See figure on next page.) Fig. 3 Means for selected variables at each sample distance from the edge to define potential effects on richness and abundance of avian guilds at the Sakaerat Environmental Research Station in 2014–2015 based on generalized linear mixed models. Selected variables (y-axes) included: stem density of vegetation with heights between 0.5–3 m (D0), tree basal area (BA), maximum ambient noise (Noise), percentage cover of vegetation at < 0.5 m (C0), 3–5 m (C3), 5–10 m (C5) and > 10 m (C10) Khamcha et al. Avian Res (2018) 9:20 Page 7 of 13 Khamcha et al. Avian Res (2018) 9:20 Page 8 of 13 Table 1 Candidate models (ΔAICc < 2) from generalized in higher layers were higher, but percentage cover of linear mixed models to explain the effects of forest edges vegetation in lower layers and ambient noise were lower on avian guilds at the Sakaerat Environmental Research (Fig. 3). There was very limited evidence to suggest there Station in 2014–2015 based on variables included were negative edge effects on the abundance of arboreal in the models frugivore-insectivores, foliage gleaning insectivores, and Model k AICc ΔAICc w LL raptors, although their abundances were higher further from the edge (Table 3; Fig. 4b). We found no evidence Richness to support edge effects on the abundance of granivore- Edge * D0 7 4653.38 0.00 0.20 − 2319.67 insectivores, terrestrial insectivore-faunivores, and ter- Edge + D0 6 4653.77 0.39 0.17 − 2320.87 restrial insectivore-frugivores (Table 3). Edge * C0 7 4654.60 1.22 0.11 − 2320.28 For most avian guilds, there were gradual shifts in Edge + D0 + C5 + C0 8 4654.70 1.32 0.10 − 2319.33 abundance and richness from the edge into the interior; Edge + D0 + C0 7 4654.75 1.37 0.10 − 2320.36 in contrast, the bark-gleaning insectivores were only Edge * D0 + C5 8 4655.31 1.93 0.08 − 2319.63 observed at distances > 600 m from the edge (Fig. 4), Edge 3 4719.15 67.24 0.00 − 2327.84 which appeared to be related to basal area and vegeta- D0 3 4728.79 76.87 0.00 − 2361.39 tion cover at the canopy level (Fig. 3). The effect of traf - NULL 2 7681.65 3029.74 0.00 − 3838.82 fic/ambient noise seemed to occur mostly within 100 m Abundance (Fig. 3) (see the results on the effect of traffic/ambient Edge * D0 7 5231.89 0.00 0.26 − 2608.93 noise below). The vegetation structure near the forest Edge + D0 + C5 + C0 8 5233.28 1.39 0.13 − 2608.61 edge (< 350 m) was significantly different from the forest Edge * D0 + C0 8 5233.70 1.81 0.11 − 2608.83 interior. The difference was most notable within 100 m Edge 3 5325.91 94.02 0.00 − 2659.95 from the edge (Fig. 3). D0 3 5330.18 98.29 0.00 − 2261.09 We generated 15 models to evaluate the influence of the NULL 2 9245.03 4013.14 0.00 − 4620.51 potential variables on richness and abundance of avian Variables in our models included: distance to forest edge (Edge), stem density of guilds that were correlated with distance from edge. The vegetation with height between 0.5–3 m (D0), percentage cover of vegetation at models with strong support (ΔAICc < 2) included maxi- < 0.5 m (C0) and 5–10 m (C5) mum ambient noise and largely the same variables (stem AICc Akaike’s information criterion values, ∆AICc the difference in AIC rank relative to the top model, w the relative model weights, k the number of density of saplings, percentage cover of vegetation at parameters in the model, LL log-likelihood < 0.5 m and 5–10 m) as the models with strong support Full set of models for richness and abundance modelled: Edge, Edge + D0, (ΔAICc < 2) from the analysis which included “distance Edge + C0, Edge + D0 + C0, Edge + D3, Edge + D5, Edge + DDBH, Edge + C05, Edge + C5, Edge + D0 + C5, Edge + D0 + C5 + C0, Edge * C3, Edge * D3 + D0, from edge” without ambient noise. Maximum ambient Edge * D0, Edge * D0 + C5, Edge * D0 + C5 + C0, Edge * C5, Edge * C0, noise was negatively associated with species richness Edge * D0 + C0, Edge * DDBH, Edge * C5, Edge * D5. D3 is stem density of (β = − 0.04) and abundance (β = − 0.06). Other vari- vegetation with height between 3-5 m, D5 is stem density of vegetation with height between 5-10 m and DDBH is stem density of trees with DBH > 5 cm ables in the top-ranked models and models with strong support that had influence on species richness included stem density of saplings (β = 0.22) and percentage cover (Table 3; Fig. 4a). We found only limited evidence to sug- at 5–10 m (β = − 14). For abundance, stem density of sap- gest there were negative effects of forest edge on the rich - lings (β = 0.25), the interaction between maximum ambi- ness of foliage gleaning insectivores and raptors, although ent noise and stem density of saplings (β = − 0.13), and their richness were higher further from the edge (Table 3; percentage cover at 5–10 m (β = − 12) were the impor- Fig. 4a). We found no evidence to support effects of forest tant factors, with influence in the same direction as the edge on the richness of granivore-insectivores, terrestrial models with distance from edge. Some bird guilds (sal- insectivore-faunivores, and terrestrial insectivore-frugi- lying, terrestrial, and understory insectivores) responded vores (Table 3). negatively to ambient noise opposite to the responses Similar to richness, we found some evidence of nega- to distance to edge; there was a strong negative correla- tive effects of edge on the abundance of bark-gleaning, tion (r = − 0.8) between ambient noise and distance from sallying, terrestrial, and understory insectivores (Table 3; edge. Fig. 4b). We also found some evidence suggesting positive effects of edges on abundance of arboreal nectarivore- Discussion insectivores (Table 3; Fig. 4b). The abundances of bark- Loss and fragmentation of intact forest landscapes gleaning, sallying, terrestrial, and understory insectivores caused by roads and other infrastructure development were higher further away from the edge where the density leads to landscape transformation and loss of its conser- of saplings, basal area, and percentage cover of vegetation vation value (Potapov et al. 2017). We found support for Khamcha et al. Avian Res (2018) 9:20 Page 9 of 13 Table 2 Estimates of coefficients of variables that suggest vegetation cover near the ground but lower cover in the significant influence on avian guild species richness canopy, as well as lower total basal area and lower density and abundance, standard errors (SE) and their 85% of larger trees (Dale et al. 2000; Watson et al. 2004). How- confidence intervals (CI) at the Sakaerat Environmental ever, the density of saplings in our study area was higher Research Station in 2014–2015 derived from model at both edge and deeper interior areas, but relatively averaging lower at intermediate distances. Furthermore, the edge in Variables Coefficient SE Lower 85% CI Upper 85% CI our study was a busy (approximately 950 cars/h) five-lane estimated highway with substantial traffic noise, especially within 100 m of the edge, which may have further negatively Richness impacted at least some guilds as demonstrated elsewhere Edge 0.09 1.22 − 1.67 1.84 (Arevalo and Newhard 2011; Ware et al. 2015), and is fur- D0 0.33 0.19 0.05 0.61 ther explained below. Edge * D0 − 0.23 0.15 − 0.46 − 0.01 C0 − 0.20 0.20 − 0.50 0.08 Edge * C0 0.35 0.09 0.21 0.49 Influence of vegetation structure on avian community C5 − 0.18 0.22 − 0.50 0.13 The lower richness and abundance of insectivores near Abundance the edge in our study was similar to other tropical stud- Edge 0.97 0.86 − 0.26 2.21 ies (Laurance et al. 2004; Mammides et al. 2015), and may D0 0.49 0.16 0.26 0.72 be explained by the relatively unsuitable microhabitats Edge * D0 − 0.29 0.14 − 0.49 − 0.08 created by the increased understory density and reduced C5 − 0.32 0.19 − 0.59 − 0.05 canopy cover, which may lead to higher light intensity, C0 − 0.25 0.25 − 0.62 0.11 higher temperatures, and lower humidity (Pollock et al. 2015). Vegetation structure had the strongest effect on Edge represents distance to forest edge, D0 is stem density of vegetation with height between 0.5–3 m, C0 is percentage cover of vegetation at < 0.5 m, and C5 species richness and abundance of insectivorous birds, is percentage cover of vegetation at 5–10 m with higher complexity of vegetation structure positively Indicates estimated coefficients of variables that had a significant influence on associated with species richness and abundance of most avian guild richness and abundance insectivores, especially bark-gleaning, sallying, terrestrial, and understory insectivores (Ferger et al. 2014; Mam- our hypothesis that edge effects influenced at least some mides et al. 2015). In our study area, vegetation struc- avian guilds differently and that certain groups, including ture within 100 m from the edge was generally simpler, understory, sallying, terrestrial, and bark-gleaning insec- with only one or two layers of particularly dense small tivores, were observed to be higher in richness and abun- trees and saplings and a greater cover of vegetation near dance in the forest interior. Our study revealed complex ground level. These structural changes may have nega - spatial patterns in the bird community near a roadside tively affected the foraging microhabitats for understory forest edge at SERS in Thailand. Although most (69%) insectivorous birds (Pollock et al. 2015). The reduced species were found at all distances from the edge, in layers of small trees may limit shelter and foraging sub- terms of avian guilds, proximity to the edge appeared to strates for sallying and understory insectivores, such have negative effects on most avian guilds, but arboreal as the Black-naped Monarch (Hypothymis azurea) and nectarivore-insectivores responded positively. Orange-breasted Trogon (Harpactes oreskios); whereas the greater cover at ground level may be an obstacle to Edge effects on avian guilds terrestrial insectivores such as the Puff-throated Bab - Overall, species richness and abundance of most avian bler (Pellorneum ruficeps). In contrast, the forest interior guilds were reduced close to the edge, similar to findings had a more complex vegetation structure, with a higher of other tropical studies (Watson et al. 2004; Deikumah basal area, higher density of larger trees, more vegetation et al. 2014). Distance from the edge, density of saplings, layers, and less ground cover, which probably provided basal area, and percentage cover of vegetation > 10 m had more diverse arthropod resources and foraging habitats positive effects on most guilds. In contrast, increased (Ferger et al. 2014). vegetation cover in low and middle layers and ambient Bark-gleaning woodpeckers and understory insec- noise seemed to have negative effects. The vegetation tivores (trogons and kingfishers) also were negatively structure near the forest edge was different from the for - affected by the edge. Their richness and abundance were est interior. Differences extended to 350 m from the edge, higher in the interior and positively related to the basal but the clearest differences were within the first 100 m. area, density of large trees and canopy cover, which Similar to previous studies, forest edge areas had a higher has also been reported by Mahmoudi et al. (2016), and density of small trees and saplings, resulting in higher Whelan and Maina (2005). This group of birds may also Khamcha et al. Avian Res (2018) 9:20 Page 10 of 13 Fig. 4 The estimated regression lines for eight avian guilds and their responses relative to distance from the edge at the Sakaerat Environmental Research Station in 2014–2015. a estimated regression lines for richness of eight avian guilds, b estimated regression lines for abundance of eight avian guilds. Arboreal frugivore-insectivores (AFI), arboreal nectarivore-insectivores (ANI), bark-gleaning insectivores (BGI), foliage-gleaning insectivores (FGI), raptors (R), sallying insectivores (SaI), terrestrial insectivores ( TI), and understory insectivores (UI) avoid edge areas due to the rarity of particular foraging small-bodied frugivores such as the Yellow White-eye and nesting substrates, including larger trees and stumps (Zosterops senegalensis) and Common Bulbul (Pycnono- (Newton 1994; Lindell et al. 2004). tus barbatus) that occur at soft edges with agriculture There was a small, but significantly negative response areas. On the other hand, richness and abundance of for- to edges by arboreal frugivore-insectivores, similar to est specialists and large-bodied frugivores, such as horn- some other studies in the tropics (Watson et al. 2004; bills and pigeons, decreased at the edges (Deikumah et al. Mammides et al. 2015). Slightly higher richness and 2014; Mammides et al. 2015). Sixty percent of our birds in abundance of frugivores in the forest interior may be this guild were forest interior species (Round et al. 2011) associated with a higher basal area and density of large and some had relatively larger body sizes, such as the trees (Deikumah et al. 2014) and more potential food Asian Fairy-bluebird (Irena puella; 70 g), Oriental Pied resources (Ferger et al. 2014). Higher basal area, density Hornbill (Anthracoceros albirostris; 750 g), and Thick- of large trees and vegetation cover at the canopy level of billed Green-pigeon (Treron curvirostra; 150 g). Moreo- our forest interior could provide greater fruit availability ver, the roadside habitat adjacent to our forest edge was (Lindenmayer et al. 2012). Some other studies have found unlikely to attract frugivores in the way that agricultural frugivores to be tolerant to edges (Menke et al. 2012), but habitats often do. the tolerant species were mostly generalists, and included Khamcha et al. Avian Res (2018) 9:20 Page 11 of 13 Table 3 Estimates of coefficients of 11 avian guilds for species richness and abundance response to edge effects, SE and their 95% confidence intervals (CI) at the Sakaerat Environmental Research Station in 2014–2015 Guilds Richness Abundance Coefficients SE Lower 95% CI Upper 95% CI Coefficients SE Lower 95% CI Upper 95% CI AFI 0.17 0.07 0.01 0.32 0.12 0.06 − 0.01 0.25 ANI − 1.25 0.28 − 1.81 − 0.68 − 1.38 0.28 − 1.94 − 0.81 BGI 2.92 0.56 1.78 4.05 2.95 0.58 1.78 4.11 FGI 0.11 0.07 − 0.03 0.25 0.10 0.06 − 0.03 0.23 GI − 1.17 0.91 − 3.01 0.66 -0.33 0.91 − 2.15 1.49 R 0.83 0.45 − 0.08 1.74 0.81 0.44 − 0.07 1.69 SaI 0.54 0.11 0.31 0.76 0.53 0.10 0.32 0.73 TI 0.51 0.13 0.24 0.77 0.64 0.11 0.40 0.87 TIF 0.20 0.32 − 0.45 0.85 0.24 0.32 − 0.41 0.89 TIV − 1.16 0.68 − 2.53 0.21 − 0.66 0.61 − 1.88 0.56 UI 0.62 0.14 0.32 0.91 0.68 0.13 0.41 0.94 Positive coefficients indicate guilds that were more likely to be found away from the edge; negative coefficients indicate guilds that were more likely found near the edge Arboreal frugivore-insectivores (AFI), arboreal nectarivore-insectivores (ANI), bark-gleaning insectivores (BGI), foliage-gleaning insectivores (FGI), granivore- insectivores (GI), raptors (R), sallying insectivores (SaI), terrestrial insectivores (TI), terrestrial insectivore-frugivores (TIF), terrestrial insectivore-faunivores (TIV ), and understory insectivores (UI) The nectarivore-insectivore guild, such as sunbirds, assessment is correlative, we encourage future studies to was the only one to show a positive response to forest further evaluate noise at edges in an experimental fash- edge, similar to other studies (Laurance 2004; Watson ion to confirm this mechanism (e.g. Ware et al. 2015). et al. 2004). This guild depends greatly on nectar which Finally, the lack of an edge effect on raptors may be is likely more abundant closer to the edge where higher explained by their wide-ranging behaviour, with move- light levels have been suggested to cause understory ments likely larger than the distance we surveyed from shrubs and lianas to produce more flowers, especially lia - the forest edge. We reported few raptors (0.02 indi- nas which are also denser closer to the edge at all height viduals/km ) and they typically occur at low densities levels (Levey 1988; Barber and Marquis 2011). (Andersen 2007). Minimal edge effects on foliage-glean - ing insectivores may also be a result of life history traits Influence of traffic noise on avian community of members in this guild, of which around half were A decline in richness and abundance of birds close to small, non-forest specialists, including the Common Tai- road edges has been observed in many studies and noise lorbird (Orthotomus sutorius), Common Iora (Aegithina may have significant effects on avian communities in gen - tiphia), and Plain Prinia (Prinia inornata) (Round et al. eral (Arevalo and Newhard 2011; Polak et al. 2013; Ware 2011). In the case of granivore-insectivores, terrestrial et al. 2015). Traffic noise from Route 304 was high (aver - insectivore-faunivores, and terrestrial insectivore-frugi- age = 67 dB) within 100 m from the edge and then rapidly vores, we found no evidence of edge effects on their rich - decreased into the interior. Noise disturbance from this ness and abundance. However, each of these guilds had busy five-lane highway could therefore be the primary only one species and their sample sizes were small. explanation for the decline in richness and abundance for most avian guilds at the edge (Arevalo and Newhard Conclusions 2011), although this effect was confounded with vegeta - The responses of avian guilds to edges can vary both tion changes in our study. Traffic noise distracts birds, between and within regions (Watson et al. 2004; Lin- making them more vulnerable to predation, and disrupts dell et al. 2007), which creates a need for region-specific the singing for pairing during the breeding season, which assessments. Thus, while it is difficult to extrapolate may have resulted in edge avoidance by some avian guilds the effects of forest edges on avian communities from (Arevalo and Newhard 2011; Polak et al. 2013), such one relatively small study area and one type of edge to as the terrestrial insectivores, Green-legged Partridge other forests, this is the first detailed study of the edge (Arborophila chloropus) and Puff-throated Babbler (Pel responses of birds in Southeast Asia, and the only study to investigate a roadside edge. Differences in vegetation lorneum ruficeps), that frequently vocally communi - structure and greater traffic noise made the forest edge cate especially during the breeding season. Because our Khamcha et al. Avian Res (2018) 9:20 Page 12 of 13 Angkeaw for her full-time help in the field. We are also grateful to T. Klubchum, less suitable for most avian guilds. However, traffic noise the Sakaerat bird team and the Conservation Ecology Program staff and stu- and vegetation structure are confounded in most analy- dents who provided support for this study. DK thanks S. Bumrungsri for advice ses in our study, so we encourage future studies in which on this project. We also thank D. Ngoprasert and W. Chutipong for their advice on the statistical analysis. sound levels and canopy openness and/or understory vegetation density are experimentally varied. Competing interests Laurance et al. (2015) noted that “we are living in the The authors declare that they have no competing interests. most explosive era of infrastructure explosion in human Consent for publication history” and Southeast Asia is the epicentre for this Not applicable. expansion. Consistent with this observation, our study Ethics approval and consent to participate revealed impacts of a five-lane highway (Route 304), All procedures involving animals in this study complied with the current act which runs adjacent to an intact forest landscape com- regarding Animals for Scientific Purposes in Thailand and had the approval prising three protected areas in Thailand, including our of the Animal Ethics Committee of King Mongkut’s University of Technology Thonburi, Bangkok, Thailand. This study was permitted to do fieldwork in study area. Planners should avoid road development in Sakaerat Environmental Research Station under the permission from the direc- forests of high conservation value to reduce impacts on tor of Sakaerat Environmental Research Station. biodiversity; such development also effects carbon stor - Funding age, soil erosion and water catchments. Where avoid- Our research was supported by King Mongkut’s University of Technology ance is unavoidable, it may be possible to mitigate the Thonburi ( Thailand) and the National Science and Technology Development impacts on forest bird communities by well-designed Agency (CPMO P-14-51347). DK was supported by the Royal Golden Jubilee Ph.D. Program, Thailand (PHD/0036/2556). plantings of a dense buffer of fast-growing trees (Van Renterghem et al. 2012), restoration of vegetation Received: 25 October 2017 Accepted: 25 May 2018 (Arnold and Weeldenburg 1990), or use of earth berms along the roadside in sensitive areas. Natural barriers can reduce traffic noise as successfully as artificial noise barriers such as concrete walls, and natural noise barri- References Andersen DE. Survey techniques. In: Bird DM, Bildstein KM, editors. Raptor ers are likely to have less impact on the avian community research and management techniques. Surrey: Hancock House Publish- in roadside habitats, although more detailed assessments ers; 2007. p. 89–100. of mitigation measures would likely be beneficial in this Arevalo E, Newhard K. Traffic noise affects forest bird species in a protected tropical forest. Rev Biol Trop. 2011;59:969–80. region. Trade-offs between development and conserva - Arnold GW, Weeldenburg JR. Factors determining the number and species of tion are unavoidable in many cases (Bennett 2017), so we birds in road verges in the wheatbelt of Western Australia. Biol Conserv. encourage conservation planners to work with govern- 1990;53:295–315. Arnold TW. Uninformative parameters and model selection using Akaike’s ment agencies involved with road development and road information criterion. J Wildl Manag. 2010;74:1175–8. maintenance projects to proactively minimize adverse Barbaro L, van Halder I. Linking bird, carabid beetle and butterfly life-history impacts on avian communities in protected areas. traits to habitat fragmentation in mosaic landscapes. Ecography. 2009;32:321–33. Additional file Barber NA, Marquis RJ. Light environment and the impacts of foliage quality on herbivorous insect attack and bird predation. Oecologia. 2011;166:401–9. Additional file 1. Table S1. List of bird species and their feeding guilds in Barber CP, Cochrane MA, Souza CM Jr, Laurance WF. Roads, deforestation, and the Sakaerat Environmental Research Station in 2014–2015. the mitigating effect of protected areas in the Amazon. Biol Conserv. 2014;177:203–9. Bennett VJ. Eec ff ts of road density and pattern on the conservation of species and biodiversity. Curr Landsc Ecol Rep. 2017;2:1–11. Authors’ contributions Bierregaard RO, Stouffer PC. Understory birds and dynamic habitat mosaics in DK, GAG, RTC, TS and AJL conceived and designed the study. DK performed Amazonian rainforests. In: Laurance WF, Bierregaard RO, editors. Tropical the fieldwork. DK and GAG analyzed data and interpreted the results. DK wrote forest remnants: ecology, management, and conservation of fragmented the original draft of this manuscript. GAG, LAP, RTC, TS and AJL reviewed and communities. Chicago: University of Chicago Press; 1997. p. 138–55. made comments to the manuscript. DK, GAG and LAP revised the final draft. Bolker BA, Brooks ME, Clark CJ, Geange SW, Poulsen JR, Stevens MHH, White JS. All authors read and approved the final manuscript. GLMMs in action: gene-by-environment interaction in total fruit produc- tion of wild populations of Arabidopsis thaliana. Revised version, part 2. Author details 2011. http://www.cell.com/cms/attac hment /60162 3/47424 53/mmc2. Conservation Ecology Program, School of Bioresources and Technology, King pdf. Accessed 2 Nov 2015. Mongkut’s University of Technology Thonburi, Bangkok, Thailand. Center Dale S, Mork K, Solvang R, Plumptre AJ. Edge effects on the understory bird for Integrative Conservation, Xishuangbanna Tropical Botanical Garden, community in a logged forest in Uganda. Conserv Biol. 2000;14:265–76. Chinese Academy of Sciences, Yunnan, China. School of Natural Resources, Dayananda SK, Goodale E, Lee MB, Liu JJ, Mammides C, Quan RC, Sreekar R, University of Nebraska-Lincoln, Lincoln, NE, USA. Wildlife Conservation Soci- Yasuda M. Eec ff ts of forest fragmentation on nocturnal Asian birds: a case ety, Global Conservation Program, New York, USA. study from Xishuangbanna, China. Zool Res. 2016;37:151–8. Deikumah JP, McAlpine CA, Maron M. Mining matrix effects on West African Acknowledgements rainforest birds. Biol Conserv. 2014;169:334–43. We would like to thank T. Artchawakom, director of Sakaerat Environmental Research Station for permissions to conduct this study. Special thanks go to R. Khamcha et al. Avian Res (2018) 9:20 Page 13 of 13 Department of Highways. Annual Average Daily Traffic on Highways 2014, Murcia C. Edge effect in fragmented forest: implications for conservation. Bangkok. 2014. http://www.doh.go.th:88/dohna khonr atcha sima/attac Trends Ecol Evol. 1995;10:58–62. hment s/artic le/30/repor t2557 .pdf. Accessed 10 Oct 2016. Newmark WD, Stanley TR. Habitat fragmentation reduces nest survival in an Ewers RM, Didham RK. Confounding factors in the detection of spe- Afrotropical bird community in a biodiversity hotspot. Proc Natl Acad Sci cies responses to habitat fragmentation. Biol Rev Camb Philos Soc. USA. 2011;108:11488–93. 2006;81:117–42. Newton I. The role of nest sites in limiting the number of hole-nesting birds: a Ferger SW, Schleuning M, Hemp A, Howell KM, Böhning-Gaese K. Food review. Biol Conserv. 1994;70:265–76. resources and vegetation structure mediate climatic effects on species Polak M, Wiącek J, Kucharczyk M, Orzechowski R. The effect of road traffic on a richness of birds. Glob Ecol Biogeogr. 2014;23:541–9. breeding community of woodland birds. Eur J For Res. 2013;132:931–41. Flaspohler DJ, Temple SA, Rosenfield RN. Species-specific edge effects on nest Pollock HS, Cheviron ZA, Agin TJ, Brawn JD. Absence of microclimate selec- success and breeding bird density in a forested landscape. Ecol Appl. tivity in insectivorous birds of the Neotropical forest understory. Biol 2001;11:32–46. Conserv. 2015;188:116–25. Gale GA, Round PD, Pierce AJ, Nimnuan S, Pattanavibool A, Brockelman WY. Potapov P, Hansen MC, Laestadius L, Turubanova S, Yaroshenko A, Thies C, A field test of distance sampling methods for a tropical forest bird com- Smith W, Zhuravleva I, Komarova A, Minnemeyer S, Esipova E. The last munity. Auk. 2009;126:439–48. frontiers of wilderness: tracking loss of intact forest landscapes from 2000 Gelman A, Hill J. Data analysis using regression and multilevel/hierarchical to 2013. Sci Adv. 2017;3:e1600821. models. Cambridge: Cambridge University Press; 2007. Ralph CJ, Geupel GR, Pyle P, Martin TE, DeSante DF. Handbook of field methods Haddad NM, Brudvig LA, Clobert J, Davies KF, Gonzalez A, Holt RD, Lovejoy for monitoring landbirds. Albany: U.S. Department for Agriculture; 1993. TE, Sexton JO, Austin MP, Collins CD, Cook WM, Damschen EI, Ewers Restrepo C, Gomez N. Responses of understory birds to anthropogenic edges RM, Foster BL, Jenkins CN, King AJ, Laurance WF, Levey DJ, Margules CR, in a neotropical montane forest. Ecol Appl. 1998;8:170–83. Melbourne BA, Nicholls AO, Orrock JL, Song DX, Townshend JR. Habitat Restrepo C, Gomez N, Heredia S. Anthropogenic edges, treefall gaps, and fragmentation and its lasting impact on Earth’s ecosystems. Sci Adv. 2015. fruit-fruigivore interactions in a neotropical montane forest. Ecology. https ://doi.org/10.1126/sciad v.15000 52. 1999;80:668–85. Halfwerk W, Holleman LJM, Lessells CM, Slabbekoorn H. Negative impact of Round PD, Pierce AJ, Sankamethawee W, Gale GA. The Avifauna of the Mo traffic noise on avian reproductive success. J Appl Ecol. 2011;48:210–9. singto forest dynamics plot, Khao Yai national park, Thailand. Nat Hist Bull Jack J, Rytwinski T, Fahrig L, Francis CM. Influence of traffic mortality on forest Siam Soc. 2011;57:57–80. bird abundance. Biodivers Conserv. 2015;24:1507–29. Sliwinski M, Powell P, Koper N, Giovanni M, Schacht W. Research design Johns AD. Eec ff t of selective logging on the ecology organization of a penin- considerations to ensure detection of all species in an avian community. sular Malaysian rain forest avifauna. Forktail. 1986;1:65–79. Methods Ecol Evol. 2016;7:456–62. Lambert FR, Collar NJ. The future for Sundaic lowland forest birds: long- Sloan S, Sayer J. Forest Resources Assessment of 2015 shows positive global term effect of commercial logging and fragmentation. Forktail. trends but forest loss and degradation persist in poor tropical countries. 2002;18:127–46. For Ecol Manag. 2015;352:134–45. Laurance SGW. Responses to understory rain forest birds to road edges in Stibig HJ, Achard F, Carboni S, Raši R, Miettinen J. Changes in tropical central Amazonia. Ecol Appl. 2004;14:1344–57. forest cover of Southeast Asia from 1990 to 2010. Biogeosciences. Laurance SG, Stouffer PC, Laurance WF. Eec ff ts of road clearings on movement 2014;11:247–58. patterns of understory rainforest birds in central Amazonia. Conserv Biol. Terraube J, Archaux F, Deconchat M, Halder I, Jactel H, Barbaro L. Forest edges 2004;18:1099–109. have high conservation value for bird communities in mosaic landscapes. Laurance WF, Peletier-Jellema A, Geenen B, Koster H, Verweij P, Van Dijck P, Ecol Evol. 2016;6:5178–89. Lovejoy TE, Schleicher J, Van Kuijk M. Reducing the global environmental Van Renterghem T, Botteldooren D, Verheyen K. Road traffic noise shielding by impacts of rapid infrastructure expansion. Curr Biol. 2015;25:259–62. vegetation belts of limited depth. J Sound Vib. 2012;331:2404–25. Lee TM, Soh MC, Sodhi N, Koh LP, Lim SLH. Eec ff ts of habitat disturbance Vetter D, Hansbauer MM, Végvári Z, Storch I. Predictors of forest fragmentation on mixed species bird flocks in a tropical sub-montane rainforest. Biol sensitivity in neotropical vertebrates: a quantitative review. Ecography. Conserv. 2005;122:193–204. 2011;34:1–8. Levey DJ. Tropical wet forest treefall gaps and distributions of understory birds Vetter D, Rücker G, Storch I. A meta-analysis of tropical forest edge effects and plants. Ecology. 1988;69:1076–89. on bird nest predation risk: edge effects in avian nest predation. Biol Lindell CA, Chomentowski WH, Zook JR. Characteristics of bird species using Conserv. 2013;159:382–95. forest and agricultural land covers in southern Costa Rica. Biodivers Ware HE, McClure CJW, Carlisle JD, Barber J. A phantom road experiment Conserv. 2004;13:2419–41. reveals traffic noise is an invisible source of habitat degradation. Proc Natl Lindell CA, Riffell SK, Kaiser SA, Battin AL, Smith ML, Sisk TD. Edge responses of Acad Sci USA. 2015;112:12105–9. tropical and temperate birds. Wilson J Ornithol. 2007;119:205–20. Watson JEM, Whittaker RJ, Dawson TP. Habitat structure and proximity to Lindenmayer DB, Laurance WF, Franklin JF. Global decline in large old trees. forest edge affect the abundance and distribution of forest-dependent Science. 2012;338:1305–6. birds in tropical coastal forests of southern Madagascar. Biol Conserv. Mammides C, Schleuning M, Böhning-Gaese K, Schaab G, Farwig N, Kadis C, 2004;120:311–27. Coulson T. The indirect effects of habitat disturbance on the bird commu- Whelan CJ, Maina GG. Eec ff ts of season, understorey vegetation density, habi- nities in a tropical African forest. Biodivers Conserv. 2015;24:3083–107. tat edge and tree diameter on patch-use by bark-foraging birds. Funct Mahmoudi S, Ilanloo SS, Shahrestanaki AK, Valizadegan N, Yousefi M. Eec ff t Ecol. 2005;19:529–36. of human-induced forest edges on the understory bird community in Wilcove DS, Giam X, Edwards DP, Fisher B, Koh LP. Navjot’s nightmare revisited: Hyrcanian forests in Iran: implication for conservation and management. logging, agriculture, and biodiversity in Southeast Asia. Trends Ecol Evol. For Ecol Manag. 2016;382:120–8. 2013;28:531–40. Menke S, Böhning-Gaese K, Schleuning M. Plant-frugivore networks are less Wong TC, Sodhi NS, Turner IM. Artificial nest and seed predation experiments specialized and more robust at forest-farmland edges than in the interior in tropical lowland rainforest remnants of Singapore. Biol Conserv. of a tropical forest. Oikos. 2012;121:1553–66. 1998;85:97–104. Moradi HV, Zakaria M, Mohd A, Yusof E. Insectivorous birds and environmental factors across an edge-interior gradient in tropical rainforest of Malaysia. Int J Zool Res. 2009;5:27–41.
– Springer Journals
Published: Jun 2, 2018