Forest mammal roadkills as related to habitat connectivity in protected areas

Fragmentation of wildlife habitat by road development is a major threat to biodiversity. Hence, conservation and enhancement of habitat connectivity in roaded landscapes are crucial for effectively maintaining long-term persistence of ecological processes, such as gene flow and migration. Using multivariate statistical techniques combined with graph theoretical methods, we investigated the influence of road-crossing habitat connectivity and road-related features on roadkill abundance of forest mammals in protected areas of South Korea. Because species have different dispersal abilities and thus connectivity would differ between them, we explored three different groups of road-killed mammals, categorized as small, intermediate, and large ones. We found that in all three mammal groups, roadkills are increased on roads that intersect high-connectivity routes. Furthermore, the effect of habitat connectivity on roadkill abundance was scale-dependent. The roadkill abundances of small, intermediate, and large mammals were related with connectivity measured at scales ranging between 100 and 300 m, between 5 and 7 km, and between 10 and 25 km, respectively. Our finding with regard to scale-dependency highlights the importance of maintaining movement and connectivity across roads at multiple scales based on the dispersal potential of different species when planning conservation strategies for forest mammalian roadkill mitigation.     

Targeting habitat management in fragmented landscapes: a case study with forest vertebrates

Although improving the quality of habitat patches in fragmented landscapes is a main conservation target few studies have examined patch management in relation to the surrounding landscape. Tackling such an issue needs a cross-scale approach that takes the hierarchical nature of landscapes into account. Here I show the results of a cross-scale study focusing on the distribution patterns of ten forest vertebrate species (birds and mammals). The overarching goal of this study was to understand the strength of patch scale determinants of distribution, following the appropriate control for relevant landscape properties (e.g. habitat loss vs. habitat subdivision). I show how, after controlling for uncertainty in the detection of the species and for the role of landscape properties, patch scale variables still played an important role in determining occupancy patterns of forest vertebrates. For some species variation in the values of patch structure variables increased occurrence probability with only moderate levels of habitat loss, highlighting the fact that habitat management should be targeted towards precise landscape conditions. In other cases the effect of patch variables was strong therefore variation in their values always brought substantial increase/decrease of presence probability. Overall these results strongly suggest that habitat management should never be carried out irrespective of the properties of the surrounding landscape, rather, it should be carefully targeted towards specific landscape contexts (e.g. above a certain amount of habitat) where it is more likely to be effective.     

Shifts from native to invasive small mammals across gradients from tropical forest to urban habitat in Borneo

Urbanization has paved the way for the spread of commensal rodents at global scale. However, it is largely unknown how these species use tropical anthropogenic landscapes originally covered with forests and inhabited by diverse small mammal assemblages. We surveyed non-flying small mammals in various urban and suburban habitat types and adjacent forest in the tropical town of Kota Kinabalu in Borneo. We used occupancy and polynomial regression models to determine variation in species occurrences along gradients of land-use intensity. Müller’s sundamys (Sundamys muelleri) was the only native small mammal species found in urban and suburban landscapes with a continuous decrease in occurrence probability from forests to urban habitats. The invasive Asian black rat (Rattus rattus species complex) and the invasive Asian house shrew (Suncus murinus) had the highest occurrence probabilities in habitats of intermediate land-use intensity, but Asian black rats are also likely to occasionally invade forested habitats and occupied urban habitats in sympatry with the Norway rat (Rattus norvegicus). In urban and suburban habitats, fallow land possibly favoured the occurrence of S. muelleri and S. murinus. Other native small mammal species (Muridae, Sciuridae, Tupaiidae) were found only in forested areas. Our study shows that native small mammals found in forest are largely replaced by invasive species in urban and suburban habitats. Due to their occurrence in habitats of various land use intensities, S. muelleri and R. rattus comprise central links between forest wildlife and urban species, an association that is important to consider in studies of parasite and disease transmission dynamics.     

Summer habitat associations of bats between riparian landscapes and within riparian areas

The present study examines those features which promote bat feeding in agricultural riparian areas and the riparian habitat associations of individual species. Activity of Nathusius’ pipistrelle (Pipistrellus nathusii), common pipistrelle (Pipistrellus pipistrellus), soprano pipistrelle (Pipistrellus pygmaeus), Leisler’s bat (Nyctalus leisleri), and Myotis species (Myotis sp.) were recorded, and their habitat associations both “between” and “within” riparian areas were analyzed. General feeding activity was associated with reduced agricultural intensity, riparian hedgerow provision, and habitat diversity. Significant habitat associations for P. pipistrellus were observed only within riparian areas. Myotis species and P. pygmaeus were significantly related to indices of landscape structure and riparian hedgerow across spatial scales. Myotis species were also related to lower levels of riffle flow at both scales of analysis. The importance of these variables changed significantly, however, between analysis scales. The multi-scale investigation of species–habitat associations demonstrated the necessity to consider habitat and landscape characteristics across spatial scales to derive appropriate conservation plans.     

Dispersal and habitat connectivity in complex heterogeneous landscapes: an analysis with a GIS-based random walk model

A grid-based random walk model has been developed to simulate animal dispersal, taking landscape heterogeneity and linear barriers such as roads and rivers into account The model can be used to estimate connectivity and has been parametenzed for the badger in the central Netherlands The importance of key parameters was evaluated by means of sensitivity analysis Results agree with field observations, and give interesting insight into the isolation of populations and potential populations The model can be applied to obtain knowledge about dispersal processes in complex landscapes     

Aspen lichens in agricultural and forest landscapes: the importance of habitat quality

The amount of aspen Populus tremula , has declined in the boreal forest landscape. This decline is especially marked in young and intermediate stands due to the lack of regeneration. Aspen regeneration is nowadays mainly restricted to abandoned agricultural land. The decrease of aspen is of particular concern as it has more host-specific species than any other boreal tree species. The main question addressed is whether regenerating aspen stands in agricultural habitats can compensate for the deficiency of young stands in the forest. Data on epiphytic macrolichens show that cyanolichens increased, in number and frequency, with stand age in the forest landscape, and that there was a striking difference in species composition between stands in the two landscapes. Lichens with cyanobacterial and green-algal photobionts dominated in the forest and agricultural stands, respectively. Notably, cyanolichens were not found in stands younger than 50 yr in the forest, and stands younger than 100 yr in the agricultural landscape. This difference between the landscapes cannot be explained by stand age, stand size or isolation. Instead, differences in habitat quality, due to differences in the physical environment associated with the presence of conifers in the older forest stands, appear to be involved. We suggest that in order to conserve cyanolichens that are confined to aspen, active management practices have to be adopted that promote the regeneration of aspen in the forest landscape, and the establishment of conifers in areas where aspen regeneration is confined to the agricultural landscape. In addition, until new aspen stands with appropriate physical environments have been established, these measures must be combined with the preservation of existing old-growth stands, which can provide appropriate source populations.     

Increasing habitat connectivity in agricultural landscapes as a weed management strategy reconciling ecology and agronomy

While studies have explored how habitat amount drives weed assemblages in agroecosystems, knowledge remains limited of the effects of habitat connectivity. The response-effect trait framework provides insights into the mechanisms underpinning the relationship between landscape structure and the taxonomic diversity and abundance of weed assemblages. This study evaluated how habitat connectivity and habitat amount affect weed diversity and abundance in winter cereal fields, and whether these effects are driven by the functional composition of weed assemblages. We sampled weeds in 27 winter cereal fields. We measured habitat connectivity and habitat amount provided by wooded, grassland and cropland elements. We selected five traits related to the dispersal, establishment, and competitive abilities of weed species likely to respond to landscape structure: seed number per plant, type of reproduction, seed dry mass, plant vegetative height and seed germination rate. The functional composition of weed assemblages was assessed using community weighted mean trait values. Weed diversity and abundance were used as proxies of weed management. The taxonomic approach did not reveal any effect of landscape structure on weed diversity and abundance. Only the grassland elements that contributed to habitat connectivity, and to a lesser extent to habitat amount, drove the functional composition of weed assemblages. High habitat amount favoured species with many seeds, while high habitat connectivity favoured species with fewer seeds, a higher ability to reproduce vegetatively and higher seed germination rates. In turn, higher seed germination rates increased weed evenness and reduced weed abundance. Some of these relationships were influenced by the presence of rare species. Overall, high connectivity provided by grassland elements increases weed evenness and reduces weed abundance by shaping weed functional composition. Our study suggests that land-use planning policies that enhance the connectivity provided by grassland elements could be considered as a weed management strategy reconciling ecology and agronomy.     

Importance of high quality early-successional habitats in managed forest landscapes to rare beetle species

Species adapted to early-successional forest habitats are in managed landscapes largely confined to clearcuts. To improve habitat quality on clearcuts, green tree and dead wood retention is widely applied in forestry; however, its effects on rare early-successional species have rarely been shown. We repeatedly surveyed two red-listed beetle species (Upis ceramboides and Platysoma minus) on clearcuts in a managed boreal forest landscape. We found that U. ceramboides decreased its occupancy over time while P. minus increased, indicating that red-listed species vary in their ability to successfully utilise managed habitats. We found no effect of connectivity on probability of occurrence, colonisation or extinction per clearcut. Trees retained alive improved habitat quality of clearcuts, since both species were more frequent in dead wood of such trees, in comparison to logging residues. We suggest that retention can be improved by protecting and creating dead wood as intact trees during harvesting. Rare specialist species require habitat of high quality, and consequently it is impossible to meet the requirements of these species on every clearcut. To preserve all early-successional species at a regional scale, we recommend focusing retention of green trees and dead wood to one or a few trees species on each clearcut and in each landscape.     

Butterfly response to severe ENSO-induced forest fires in Borneo

Abstract.  1. Little is known about animal community response to severe El Niño Southern Oscillation (ENSO)-induced fire events. Here the response of butterflies to the 1997/98 ENSO-induced fire event in East Kalimantan (Indonesian Borneo) is assessed. In addition to the community-wide study, a detailed assessment of the lycaenid Jamides celeno is made.
2. Species richness declined significantly from 211 species pre-ENSO to 39 species post-ENSO and community composition changed significantly. Along with the decline in species richness there was a marked increase in dominance. Jamides celeno , for example, increased from 3% of the pre-ENSO assemblage to 58% of the post-ENSO assemblage. Like J. celeno , most of the species in the post-ENSO assemblage were generalists; most of the specialist species having disappeared from pre- to post-ENSO.
3. The major host plant used by J. celeno was the abundant resprouting Fordia splendidissima . Furthermore, significantly more eggs were laid on plants with the crazy ant, Anoplolepis gracilipes , present than on plants with other ants or no ant attendance. Caterpillar presence was significantly higher on plants tended by ants than on untended plants.
4. The median distance moved by J. celeno was 30 m with a maximum recorded distance of 290 m.
5. The abundance of J. celeno and other generalists in the post-ENSO assemblage at Wanariset was probably related to their ability to utilise the few available resources after the fire (e.g. F. splendidissima resprouts), their presence in degraded habitats surrounding the Wanariset forest, and their ability to disperse successfully by either being strong fliers (e.g. Euploea spp.) or being able to attain very high population sizes and thereby produce a surplus of dispersers (e.g. J. celeno ).     

Adjacent woodlands rather than habitat connectivity influence grassland plant,carabid and bird assemblages in farmland landscapes

One response to biodiversity decline is the definition of ecological networks that extend beyond protected areas and promote connectivity in human-dominated landscapes. In farmland, landscape ecological research has focused more on wooded than open habitat networks. In our study, we assessed the influence of permanent grassland connectivity, described by grassland amount and spatial configuration, on grassland biodiversity. We selected permanent grasslands in livestock farming areas of north-western France, which were sampled for plants, carabids and birds. At two spatial scales we tested the effects of amount and configuration of grasslands, wooded habitats and crops on richness and abundance of total assemblages and species ecological groups. Grassland connectivity had no significant effects on total richness or abundance of any taxonomic group, regardless of habitat affinity or dispersal ability. The amount of wooded habitat and length of wooded edges at the 200 m scale positively influenced forest and generalist animal groups as well as grassland plant species, in particular animal-dispersed species. However, for animal groups such as open habitat carabids or farmland bird specialists, the same wooded habitats negatively influenced richness and abundance at the 500 m scale. The scale and direction of biodiversity responses to landscape context were therefore similar among taxonomic groups, but opposite for habitat affinity groups. We conclude that while grassland connectivity is unlikely to contribute positively to biodiversity, increasing or maintaining wooded elements near grasslands would be a worthwhile conservation goal. However, the requirements of open farmland animal species groups must be considered, for which such action may be deleterious.     

Assessing the effectiveness of protected areas for conserving range-restricted rain forest butterflies in Sabah,Borneo

Rain forests on Borneo support exceptional concentrations of endemic insect biodiversity, but many of these forest-dependent species are threatened by land-use change. Totally protected areas (TPAs) of forest are key for conserving biodiversity, and we examined the effectiveness of the current TPA network for conserving range-restricted butterflies in Sabah (Malaysian Borneo). We found that mean diurnal temperature range and precipitation of the wettest quarter of the year were the most important predictors of butterfly distributions (N = 77 range-restricted species), and that species richness increased with elevation and aboveground forest carbon. On average across all species, TPAs were effective at conserving ~43% of species’ ranges, but encompassed only ~40% of areas with high species richness (i.e., containing at least 50% of our study species). The TPA network also included only 33%–40% of areas identified as high priority for conserving range-restricted species, as determined by a systematic conservation prioritization analysis. Hence, the current TPA network is reasonably effective at conserving range-restricted butterflies, although considerable areas of high species richness (6,565 km²) and high conservation priority (11,152–12,531 km²) are not currently protected. Sabah's remaining forests, and the range-restricted species they support, are under continued threat from agricultural expansion and urban development, and our study highlights important areas of rain forest that require enhanced protection.     

Vegetation responses to burning in a rain forest in Borneo

During the 1997/98 ENSO (El Niño Southern Oscillation) event more than 5 million ha of East Kalimantan, Indonesia burned. Here we quantify the initial stages of regeneration (19982001), both in forest that burned and in unburned controls. Sapling and seedling density and species richness remained significantly lower in burned than in unburned forest and community composition remained substantially different between both forest types throughout the sampling period. The only pronounced edge effect was a significantly higher density of seedlings in the interior of unburned forest. Sapling density increased and seedling density declined in both unburned and burned forest during the four-year study period. In the unburned forest, sapling and seedling species richness remained stable, but sapling species richness declined significantly with time in the burned forest. The pioneer community in the burned forest was, furthermore, characterised by higher growth and recruitment than in the unburned forest but mortality did not differ between both forest types. Differences in environment (burned versus unburned: 2965% of variation explained) and the distance between sample sites (1323% of variation explained) explained substantial amounts of variation in sapling and seedling community similarity. Similarity was, however, only marginally (< 1% explained) related to the edge position and temporal variation (difference among sample events). Our results, four years after the initial burn, indicate that burned forest still differed greatly from unburned forest in terms of density, species richness and community composition. There was also no clear trend of a return to pre-disturbance conditions, which indicates that the burned forest may remain in a severely degraded state for a prolonged period of time.     

Both spatiotemporal connectivity and habitat quality limit the immigration of forest plants into wooded corridors

Extensive afforestation of agricultural areas has increased the importance of green corridors as a dispersal network. We tested the effect of spatiotemporal connectivity, edge effect and habitat structural quality of wooded corridors on the long-term immigration success of forest specialist plants relative to the performance of forest generalists. In agricultural landscapes of central and southern Estonia, we sampled 28 historically connected and 52 isolated tree lines and alleys with a minimum age of 50 years, and 93 edges of ancient forests. The regional pool of common forest plants was compiled using species’ frequency data in 91 ancient forests. Both landscape connectivity and habitat quality affected the richness of response groups, but specialists and generalists responded to different drivers. Forest specialists required long-term neighbourhoods of ancient forest and benefited from a direct connection between forest and corridor. Habitat generalists reacted positively to the recently modified structure of the landscape. When a corridor was connected to forest, the dual edge in the corridor did not result in an increased negative edge effect on forest specialist arrival. Even if both specialists and generalists required wide corridors with optimum shade, forest specialists also benefited from mature overstorey and outward overhanging branches, whereas forest generalists used disturbance-created microhabitats. We conclude that only wooded corridors with long-term connectivity to seed source forests and widely branched tree canopies will function as a green infrastructure supporting forest-specific biodiversity.     

Application of random forest algorithm for studying habitat selection of colonial herons and egrets in human-influenced landscapes

Understanding the mechanisms of habitat selection is fundamental to the construction of proper conservation and management plans for many avian species. Habitat changes caused by human beings increase the landscape complexity and thus the complexity of data available for explaining species distribution. New techniques that assume no linearity and capable to extrapolate the response variables across landscapes are needed for dealing with difficult relationships between habitat variables and distribution data. We used a random forest algorithm to study breeding-site selection of herons and egrets in a human-influenced landscape by analyzing land use around their colonies. We analyzed the importance of each land-use variable for different scales and its relationship to the probability of colony presence. We found that there exist two main spatial scales on which herons and egrets select their colony sites: medium scale (4 km) and large scale (10–15 km). Colonies were attracted to areas with large amounts of evergreen forests at the medium scale, whereas avoidance of high-density urban areas was important at the large scale. Previous studies used attractive factors, mainly foraging areas, to explain bird-colony distributions, but our study is the first to show the major importance of repellent factors at large scales. We believe that the newest non-linear methods, such as random forests, are needed when modelling complex variable interactions when organisms are distributed in complex landscapes. These methods could help to improve the conservation plans of those species threatened by the advance of highly human-influenced landscapes.     

Using abundance and habitat variables to identify high conservation value areas for threatened mammals

The present study used abundance and habitat variables to design High Conservation Value Forests for wildlife protection. We considered great apes (Gorilla gorilla gorilla and Pan troglodytes troglodytes) as model species, and we used nest surveys, dietary analysis and botanical inventories to evaluate whether the traditional methods that use abundance data alone were consistent with the survival of the species. We assumed that setting a local priority area for animal conservation can be made possible if at least one variable (abundance or habitat variables) is spatially clustered and that the final decision for a species may depend on the pattern of spatial association between abundance, nesting habitat and feeding habitat. We used Kernel Density Estimation to evaluate the spatial pattern of each biological variable. The results indicate that all three variables were spatially clustered for both gorillas and chimpanzees. The abundance variables of both animal species were spatially correlated to their preferred nesting habitat variables. But while the chimpanzee feeding habitat variable was spatially correlated to the abundance and nesting habitat variables, the same pattern was not observed for gorillas. We then proposed different methods to be considered to design local priority areas for the conservation of each great ape species. Alone, the abundance variable does not successfully represent the spatial distribution of major biological requirements for the survival of wildlife species; we, therefore, recommend the integration of the spatial distribution of their food resources to overcome the mismatch caused by the existence of a biological interaction between congeneric species.     

Allocation to defense or growth in dipterocarp forest seedlings in Borneo

We quantified the allocation of net production to plant secondary metabolites (especially condensed tannins and lignins) to evaluate the investment into defense mechanisms of three tropical forest canopy species in Borneo vis-à-vis the resource availability hypothesis. In particular, we focused on Borneo ironwood (Eusideroxylon zwageri, Lauraceae), which seemed to employ an extreme defensive strategy. The wood of this species is extremely durable and has a high specific gravity with a very slow growth rate. The allocation to defense by Borneo ironwood was compared to two emergent species of Dipterocarpaceae, the dominant family in this forest community. We conducted shade-house experiments on seedlings under four controlled conditions (two light levels × two nutrient levels) and showed that the growth rate of E. zwageri was much lower than those of the other two species, and it allocated more of its net production to leaves and roots than to stems. The concentrations of condensed tannins and lignins were very high in the leaves and stems of this species, at about 20 and 30%, respectively. In total, E. zwageri allocated a maximum of about 35% of its net production to defensive substances (i.e., condensed tannins + lignins). In contrast, the two dipterocarp species allocated about 20–25% of their net production to defensive substances. The condensed tannins in E. zwageri help to prolong the lifespan of the leaves, and the lignins in the stems enhance the durability of the wood. Thus, although E. zwageri grows very slowly, the allocation to defensive substances seems to be an advantageous strategy for survival under dark conditions.An erratum to this article can be found at