Livestock-driven land use change to model species distributions: Egyptian vulture as a case study

Species distribution models (SDMs) are increasingly used to predict species ranges and their shifts under future scenarios of global environmental change (GEC). SDMs are thus incorporating key drivers of GEC (e.g. climate, land use) to improve predictions of species’ habitat suitability (i.e. as an indicator of species occurrence). Yet, most SDMs incorporating land use only consider dominant land cover types, largely ignoring other key aspects of land use such as land management intensity and livestock. We developed SDMs including main land use components (i.e. land cover, livestock and its management intensity) to assess their relative importance in shaping habitat suitability for the Egyptian vulture, an endangered raptor linked to livestock presence. We modelled current and future (2020 and 2050) habitat suitability for this vulture using an organism-centred approach. This allowed us to account for basic species’ habitat needs (i.e. nesting cliff) while gaining insight into our variables of interest (i.e. livestock and land cover). Once nest-site requirements were fulfilled, land use variables (i.e. openland and sheep and goat density) were the main factors determining species’ habitat suitability. Current suitable area could decrease by up to 6.81% by 2050 under scenarios with rapid economic growth but no focus on environmental conservation and rural development. Local solutions to environmental sustainability and rural development could double current habitat suitability by 2050. Land use is expected to play a key role in determining Egyptian vulture's distribution through land cover change but also through changes in livestock management (i.e. species and stocking density). Change in stocking densities (sheep and goats/km²) becomes thus an indicator of habitat suitability for this vulture in our study area. Abandonment of agro-pastoral practises (i.e. below ∼15–20 sheep and goats/km²) will negatively influence the species distribution. Nonetheless, livestock densities above these values will not further increase habitat suitability. Given the widespread impacts of livestock on ecosystems, the role of livestock and its management intensity in SDMs for other (non-livestock-related) species should be further explored.     

Forest bees are replaced in agricultural and urban landscapes by native species with different phenologies and life‐history traits

Anthropogenic landscapes are associated with biodiversity loss and large shifts in species composition and traits. These changes predict the identities of winners and losers of future global change, and also reveal which environmental variables drive a taxon's response to land use change. We explored how the biodiversity of native bee species changes across forested, agricultural, and urban landscapes. We collected bee community data from 36 sites across a 75,000 km² region, and analyzed bee abundance, species richness, composition, and life‐history traits. Season‐long bee abundance and richness were not detectably different between natural and anthropogenic landscapes, but community phenologies differed strongly, with an early spring peak followed by decline in forests, and a more extended summer season in agricultural and urban habitats. Bee community composition differed significantly between all three land use types, as did phylogenetic composition. Anthropogenic land use had negative effects on the persistence of several life‐history strategies, including early spring flight season and brood parasitism, which may indicate adaptation to conditions in forest habitat. Overall, anthropogenic communities are not diminished subsets of contemporary natural communities. Rather, forest species do not persist in anthropogenic habitats, but are replaced by different native species and phylogenetic lineages preadapted to open habitats. Characterizing compositional and functional differences is crucial for understanding land use as a global change driver across large regional scales.     

Impacts of cattle grazing on forest structure and raptor distribution within a neotropical protected area

Uncontrolled cattle grazing is frequent in protected areas across the tropics but its effect on habitat structure and biodiversity is poorly known. We used generalized additive models to examine the precise relationships between cattle grazing intensity, vegetation structure, and raptor occurrence and richness across 39 1-km² plots within the North West Biosphere Reserve, Peru. Cattle grazing was widespread and intensity was negatively correlated with average canopy height and percentage of vegetation cover at 5–15 m. Raptors were influenced by cattle density but they were probably more strongly influenced by canopy and sub-canopy characteristics. Raptor species responded differently to increasing cattle density: presence of species that hunted or searched for food in open habitats increased with cattle density, while presence of range restricted species that hunted from perches, and declining species decreased. Moderate cattle densities (20–60 cows km^?2) may actually benefit some raptor species and help to maintain the high raptor diversity in the study area. Reserve authorities should, however, act to reduce cattle densities in some areas, which can exceed 120 cows km^?2, that seriously affect some key raptor species, and cause forest to be converted to permanent cattle pasture.     

Anticipated climate and land‐cover changes reveal refuge areas for Borneo's orang‐utans

Habitat loss and climate change pose a double jeopardy for many threatened taxa, making the identification of optimal habitat for the future a conservation priority. Using a case study of the endangered Bornean orang‐utan, we identify environmental refuges by integrating bioclimatic models with projected deforestation and oil‐palm agriculture suitability from the 1950s to 2080s. We coupled a maximum entropy algorithm with information on habitat needs to predict suitable habitat for the present day and 1950s. We then projected to the 2020s, 2050s and 2080s in models incorporating only land‐cover change, climate change or both processes combined. For future climate, we incorporated projections from four model and emission scenario combinations. For future land cover, we developed spatial deforestation predictions from 10 years of satellite data. Refuges were delineated as suitable forested habitats identified by all models that were also unsuitable for oil palm – a major threat to tropical biodiversity. Our analyses indicate that in 2010 up to 260 000 km² of Borneo was suitable habitat within the core orang‐utan range; an 18–24% reduction since the 1950s. Land‐cover models predicted further decline of 15–30% by the 2080s. Although habitat extent under future climate conditions varied among projections, there was majority consensus, particularly in north‐eastern and western regions. Across projections habitat loss due to climate change alone averaged 63% by 2080, but 74% when also considering land‐cover change. Refuge areas amounted to 2000–42 000 km² depending on thresholds used, with 900–17 000 km² outside the current species range. We demonstrate that efforts to halt deforestation could mediate some orang‐utan habitat loss, but further decline of the most suitable areas is to be expected given projected changes to climate. Protected refuge areas could therefore become increasingly important for ongoing translocation efforts. We present an approach to help identify such areas for highly threatened species given environmental changes expected this century.     

Scale dependency of two endangered charismatic species as biodiversity surrogates

Charismatic megafauna have been used as icons and financial drivers of conservation efforts worldwide given that they are useful surrogates for biodiversity in general. However, tests of this premise have been constrained by data limitations, especially at large scales. Here we overcome this problem by combining large-scale citizen-sourced data with intensive expert observations of two endangered charismatic species, Blakiston’s fish owl (forest specialist) and the red-crowned crane (wetland specialist). We constructed large-scale maps of species richness for 52 forest and 23 grassland/wetland bird species using hierarchical community modeling and citizen-sourced data at 1, 2, 5, and 10-km grid resolutions. We compared the species richness of forest and grassland/wetland birds between the breeding and non-breeding sites of the two charismatic birds at each of the four spatial resolutions, and then assessed the scale dependency of the biodiversity surrogates. Regardless of the habitat amounts, owl and crane breeding sites had higher forest and grassland/wetland bird species richness, respectively. However, this surrogacy was more effective at finer scales (1–2-km resolutions), which corresponds to the charismatic species’ home range sizes (up to 9.4 ± 2.0 km² for fish owls, and 3–4 km² for cranes). Species richness showed the highest spatial variations at 1–2-km resolutions. We suggest that the agreement of functional scales between surrogate species and broader biodiversity is essential for successful surrogacy, and that habitat conservation and restoration targeting multiple charismatic species with different specialties can complement to biodiversity conservation.     

Breeding bird species diversity across gradients of land use from forest to agriculture in Europe

Loss, fragmentation and decreasing quality of habitats have been proposed as major threats to biodiversity world‐wide, but relatively little is known about biodiversity responses to multiple pressures, particularly at very large spatial scales. We evaluated the relative contributions of four landscape variables (habitat cover, diversity, fragmentation and productivity) in determining different components of avian diversity across Europe. We sampled breeding birds in multiple 1‐km² landscapes, from high forest cover to intensive agricultural land, in eight countries during 2001?2002. We predicted that the total diversity would peak at intermediate levels of forest cover and fragmentation, and respond positively to increasing habitat diversity and productivity; forest and open‐habitat specialists would show threshold conditions along gradients of forest cover and fragmentation, and respond positively to increasing habitat diversity and productivity; resident species would be more strongly impacted by forest cover and fragmentation than migratory species; and generalists and urban species would show weak responses. Measures of total diversity did not peak at intermediate levels of forest cover or fragmentation. Rarefaction‐standardized species richness decreased marginally and linearly with increasing forest cover and increased non‐linearly with productivity, whereas all measures increased linearly with increasing fragmentation and landscape diversity. Forest and open‐habitat specialists responded approximately linearly to forest cover and also weakly to habitat diversity, fragmentation and productivity. Generalists and urban species responded weakly to the landscape variables, but some groups responded non‐linearly to productivity and marginally to habitat diversity. Resident species were not consistently more sensitive than migratory species to any of the landscape variables. These findings are relevant to landscapes with relatively long histories of human land‐use, and they highlight that habitat loss, fragmentation and habitat‐type diversity must all be considered in land‐use planning and landscape modeling of avian communities.     

Bird Assemblages in Anthropogenic Habitats: Identifying a Suitability Gradient for Native Species in the Atlantic Forest

Traditional approaches to the study of species persistence in fragmented landscapes generally consider a binary classification of habitat being suitable or unsuitable; however, the range of human‐modified habitats within a region may offer a gradient of habitat suitability (or conservation value) for species. We identified such a gradient by comparing bird assemblages among contrasting land uses (pine plantations of different age, annual crops, clear cuts and cattle pastures) in the Upper Parana Atlantic forest. Bird assemblages and vegetation structure were characterized in an extensive area of 4400 km² in Argentina and Paraguay during the breeding seasons of 2005–2010. Similarity of bird assemblages between anthropogenic habitats and the native forest and the proportion of forest species increased with vegetation vertical structure, while the proportion of open‐area species decreased. As a consequence, mature tree plantations were the most suitable habitats for forest species and were mainly used by frugivores and bark insectivores. In contrast, open habitats were the least suitable habitat for forest species and were used primarily by insectivores. Human‐created habitats that are structurally complex can be used by a subset of forest species, and may improve functional connectivity and mitigate edge effects. The conservation of large tracks of native forests, however, is critical for the long‐term persistence of the entire bird assemblage, especially for native forest dependent species.     

拉萨市城关区近50年城市扩展过程对自然生境质量的综合影响

拉萨市城关区建设用地紧邻拉鲁湿地国家级自然保护区,导致城市发展与自然生境保护间的矛盾十分突出。全面揭示该地区城市扩展过程对自然生境质量的影响对于优化城市景观格局、保护生物多样性具有重要意义。为此,定量评价了该地区1968—2019年城市扩展过程对自然生境质量的综合影响。基于锁眼卫星遥感数据、Landsat系列遥感影像和Google Earth遥感数据获取了该地区土地利用/覆盖变化信息。结合空间叠加和模型模拟量化了城市扩展过程对自然生境质量的影响。结果表明,拉萨市城关区近50年城市扩展过程不断加快,城镇建设用地面积从9.78 km~2增加到67.67 km~2,增加了近6倍。城市扩展过程导致自然生境质量加速下降,近50年下降了12.69%。自然生境质量同时受到城市扩展过程的直接影响和间接影响。在各类自然生境中,林地和湿地的生境质量下降尤为明显,分别下降了96.59%和49.21%。其中,林地主要被城镇建设用地直接占用,湿地则主要受城市扩展过程的间接影响。因此,需要同时关注城市扩展过程对自然生境的直接占用和间接影响,采取有效措施协调城市建设与生物多样性保护,促进区域可持续发展。     

Habitat loss in the restricted range of the endemic Ghanaian cichlid Limbochromis robertsi

Remote sensing has become an integral and invaluable tool to inform biodiversity conservation and monitoring of habitat degradation and restoration over time. Despite the disproportionately high levels of biodiversity loss in freshwater ecosystems worldwide, ichthyofauna are commonly overlooked in favor of other keystone species. Freshwater fish, as indicators of overall aquatic ecosystem health, can also be indicators of larger scale problems within an ecosystem. As a case study with multi-temporal, multi-resolution satellite imagery, we examined deforestation and forest fragmentation around the Atewa Forest Reserve, Ghana. Within small creeks, Limbochromis robertsi, a unique freshwater cichlid with an extremely limited distribution range, can be found. Historically, the land cover in the area has undergone substantial deforestation for agriculture and artisanal small-scale mining. In the 1389-km² study area, we found deforestation accelerated along with increased forest fragmentation in the 2014–2017 period (167.4 km² of deforestation) with the majority of the forest loss along the river and creek banks due to small-scale mining operations and increased agriculture. Field visits indicated a decrease in the total L. robertsi population by approximately 90% from the early 1990s to 2018. Its distribution has been reduced to higher elevations by anthropogenic habitat barriers at low elevations and the presence of predatory species. Loss of riparian forest through land use and cover change to mining and agriculture contributes to the habitat degradation for this endemic species. Fine spatial- and temporal-scale studies are required to assess habitat characteristics are not captured by global- or continental-scale datasets.     

Habitat fragmentation rather than habitat amount or habitat split reduces the diversity and abundance of ground-dwelling anurans within forest remnants of the Brazilian Cerrado

Rampant deforestation has caused the loss and fragmentation of natural habitats, which has precipitated a global biodiversity crisis. Research on how land-use change contributes to a loss of biodiversity is urgently needed, especially in ecosystems that have undergone rapid anthropogenic changes. We sought to investigate the extent to which habitat loss, fragmentation, and habitat split (the separation of forest and aquatic habitats) negatively influenced taxonomic diversity, functional diversity, total abundance, and the individual abundances of five anuran species in the Brazilian Cerrado. We sampled anurans between December 2017 and March 2018 using pitfall traps at sites distributed along a gradient of habitat fragmentation/habitat split: unfragmented forest, forest fragments without habitat split, and forest fragments with habitat split. Forest cover was measured within a 1-km radius of each site. Sites within unfragmented forests had higher taxonomic and functional diversities than either fragment type. Taxonomic diversity was highly correlated with functional diversity, but we did not find a pattern to the loss of functional traits. Total anuran abundance and the abundances of Chiasmocleis albopunctata, Physalaemus cuvieri, and Rhinella diptycha were higher in unfragmented forests compared to forest fragments. No species was more abundant in fragments than in unfragmented forests. Our results indicate that the fragmentation of forests by agricultural land use is directly and indirectly responsible for the loss of taxonomic and functional diversity, as well as for reducing population sizes of ground-dwelling anurans. Although we did not find a distinct effect of habitat split on ground-dwelling anurans, our study underscores the importance of preserving continuous forest habitats for the maintenance of anuran diversity in the Cerrado.     

Habitat use,abundance, and persistence of Neotropical migrant birds in a habitat matrix in northeast Belize

ABSTRACT To ensure adequate protection of nonbreeding habitats used by Neotropical migratory landbirds, we must first address questions about habitat use and quality. On the Yucatan peninsula, migrants use many habitats, several of which remain unstudied, and methodological differences preclude interhabitat comparisons based on studies to date. We used distance sampling along line transects in six habitats in northeast Belize to examine use of previously unstudied habitats (e.g., salt marsh) by Neotropical migrants and to permit comparison across habitats. We calculated unadjusted and adjusted (for detectability) density estimates for individual migrant species and for all species combined to generate hypotheses about habitat quality based on the assumption that density and quality are positively correlated. Adjusted density estimates for all migrants were highest in black mangrove habitat (1799 ± 110 ind/km²), intermediate in three forest types and milpa (range 598–802 ind/km²), and lowest in salt marsh (207 ± 32.3 ind/km²). By combining density estimates with habitat availability in our study region, we estimated that evergreen forest and black mangrove supported 70% and 9% of the region's migrant population, respectively. At the species level, five of the 10 most common species had habitat preferences (>50% detections in one habitat). Given the diversity of habitat preferences among species and apparent seasonal movements, our results indicate that Neotropical migrants in northeast Belize are dependent on a matrix of interconnected habitats.     

Land cover change and carbon emissions over 100 years in an African biodiversity hotspot

Agricultural expansion has resulted in both land use and land cover change (LULCC) across the tropics. However, the spatial and temporal patterns of such change and their resulting impacts are poorly understood, particularly for the presatellite era. Here, we quantify the LULCC history across the 33.9 million ha watershed of Tanzania's Eastern Arc Mountains, using geo‐referenced and digitized historical land cover maps (dated 1908, 1923, 1949 and 2000). Our time series from this biodiversity hotspot shows that forest and savanna area both declined, by 74% (2.8 million ha) and 10% (2.9 million ha), respectively, between 1908 and 2000. This vegetation was replaced by a fivefold increase in cropland, from 1.2 million ha to 6.7 million ha. This LULCC implies a committed release of 0.9 Pg C (95% CI: 0.4–1.5) across the watershed for the same period, equivalent to 0.3 Mg C ha^?1 yr^?1. This is at least threefold higher than previous estimates from global models for the same study area. We then used the LULCC data from before and after protected area creation, as well as from areas where no protection was established, to analyse the effectiveness of legal protection on land cover change despite the underlying spatial variation in protected areas. We found that, between 1949 and 2000, forest expanded within legally protected areas, resulting in carbon uptake of 4.8 (3.8–5.7) Mg C ha^?1, compared to a committed loss of 11.9 (7.2–16.6) Mg C ha^?1 within areas lacking such protection. Furthermore, for nine protected areas where LULCC data are available prior to and following establishment, we show that protection reduces deforestation rates by 150% relative to unprotected portions of the watershed. Our results highlight that considerable LULCC occurred prior to the satellite era, thus other data sources are required to better understand long‐term land cover trends in the tropics.     

Predicted declines in suitable habitat for greater one‐horned rhinoceros (Rhinoceros unicornis) under future climate and land use change scenarios

Rapidly changing climate is likely to modify the spatial distribution of both flora and fauna. Land use change continues to alter the availability and quality of habitat and further intensifies the effects of climate change on wildlife species. We used an ensemble modeling approach to predict changes in habitat suitability for an iconic wildlife species, greater one‐horned rhinoceros due to the combined effects of climate and land use changes. We compiled an extensive database on current rhinoceros distribution and selected nine ecologically meaningful environmental variables for developing ensemble models of habitat suitability using ten different species distribution modeling algorithms in the BIOMOD2 R package; and we did this under current climatic conditions and then projected them onto two possible climate change scenarios (SSP1‐2.6 and SSP5‐8.5) and two different time frames (2050 and 2070). Out of ten algorithms, random forest performed the best, and five environmental variables—distance from grasslands, mean temperature of driest quarter, distance from wetlands, annual precipitation, and slope, contributed the most in the model. The ensemble model estimated the current suitable habitat of rhinoceros to be 2610 km², about 1.77% of the total area of Nepal. The future habitat suitability under the lowest and highest emission scenarios was estimated to be: (1) 2325 and 1904 km² in 2050; and (2) 2287 and 1686 km² in 2070, respectively. Our results suggest that over one‐third of the current rhinoceros habitat would become unsuitable within a period of 50 years, with the predicted declines being influenced to a greater degree by climatic changes than land use changes. We have recommended several measures to moderate these impacts, including relocation of the proposed Nijgad International Airport given that a considerable portion of potential rhinoceros habitat will be lost if the airport is constructed on the currently proposed site.     

A small mammal’s map: identifying and improving the large-scale and cross-border habitat connectivity for the hazel dormouse <Emphasis Type="Italic">Muscardinus avellanarius</Emphasis> in a fragmented agricultural landscape

Fragmentation and habitat loss pose major threats to global biodiversity. Especially forest dwelling species with small ranges and high habitat specialisation are affected by ongoing land use change. Building projects for infrastructural purposes, expanding settlements, and extensive agricultural areas are assumed to have a high impact on these species. The European Habitat’s Directive aims to conserve and restore habitat networks to lower these impacts. We propose that the idea of securing habitat networks for protected species should be incorporated within large scale landscape planning e.g. by modelling and improving corridors for umbrella species. Within a Danish-German project we developed a model demonstrating potential connecting corridors for Muscardinus avellanarius, a specialised forest dwelling rodent species with low dispersal ability. We used presence data and eco-geographical variables to find a data based time-efficient procedure which may be applied also in other species for future landscape planning. The habitat suitability model shows that the hazel dormouse occupies a narrow niche with highly suitable habitats comprising edge habitats. It indicates the preference of forest patches, linear structures and networks of the mentioned habitats. Along with connectivity the diversity of suitable habitats diversity is the major factor predicting hazel dormouse presence. For conservation management, we calculated habitat corridors and highlight sections with missing connections. This allowed us to accentuate regions of high management interest. The results represent the foundation not only for this cross-border conservation project, but also for long-term dormouse conservation on a federal-state level.     

Land Cover Change in Colombia: Surprising Forest Recovery Trends between 2001 and 2010

Background

Monitoring land change at multiple spatial scales is essential for identifying hotspots of change, and for developing and implementing policies for conserving biodiversity and habitats. In the high diversity country of Colombia, these types of analyses are difficult because there is no consistent wall-to-wall, multi-temporal dataset for land-use and land-cover change.

Methodology/Principal Findings

To address this problem, we mapped annual land-use and land-cover from 2001 to 2010 in Colombia using MODIS (250 m) products coupled with reference data from high spatial resolution imagery (QuickBird) in Google Earth. We used QuickBird imagery to visually interpret percent cover of eight land cover classes used for classifier training and accuracy assessment. Based on these maps we evaluated land cover change at four spatial scales country, biome, ecoregion, and municipality. Of the 1,117 municipalities, 820 had a net gain in woody vegetation (28,092 km²) while 264 had a net loss (11,129 km²), which resulted in a net gain of 16,963 km² in woody vegetation at the national scale. Woody regrowth mainly occurred in areas previously classified as mixed woody/plantation rather than agriculture/herbaceous. The majority of this gain occurred in the Moist Forest biome, within the montane forest ecoregions, while the greatest loss of woody vegetation occurred in the Llanos and Apure-Villavicencio ecoregions.

Conclusions

The unexpected forest recovery trend, particularly in the Andes, provides an opportunity to expand current protected areas and to promote habitat connectivity. Furthermore, ecoregions with intense land conversion (e.g. Northern Andean Páramo) and ecoregions under-represented in the protected area network (e.g. Llanos, Apure-Villavicencio Dry forest, and Magdalena-Urabá Moist forest ecoregions) should be considered for new protected areas.     

The potential of military training areas for bird conservation in a central European landscape

European biodiversity has suffered from serious declines during the past few decades, with alterations of land use practices resulting in a loss of fine-scale habitat heterogeneity being a dominant driver. This heterogeneity was maintained by extensive landscape management, which has gradually been replaced by either intensive exploitation or land abandonment. It has been suggested that military training can generate habitat heterogeneity that may support the existence of species of conservation concern, but studies rigorously testing the real importance of military training areas for biodiversity are lacking. Here we address this issue by analyses of two datasets. First, we compared land cover classes between all large military training areas (MTAs) and surrounding control areas (CAs) of the same size in the Czech Republic using multivariate redundancy analysis. We found that the difference in land cover between MTAs and CAs was significant and represented the strongest gradient in land cover classes: from various farmland and artificial habitats typical for CAs to forest and scrubland-grassland mosaic typical for MTAs. Second, we selected one of these areas and compared bird species richness between the MTA and the nearby CA using generalized linear mixed effects models. We found that the number of species of conservation concern was significantly higher in the MTA than in the CA. With respect to habitats, bird species richness was significantly higher in the MTA than in the CA for open habitats, but not for forest habitats. Our results are thus consistent with the view that military training creates areas that are different from the surrounding landscape in terms of land cover, and that this difference translates to a suitability for species of conservation concern. It is remarkable that the positive influence of military training is confined to open habitats, which are subject to the most intensive military activities and also suffer the highest degree of deterioration in European landscapes.     

The white-backed woodpecker: umbrella species for forest conservation planning?

In northern Europe, a long history of land use has led to profound changes within forest ecosystems. The white-backed woodpecker (Dendrocopos leucotos) is one of several specialised forest species whose populations have declined. Conservation management directed at this species’ habitat has made it a de facto umbrella species for conservation of the biodiversity associated with forests rich in deciduous trees and dead wood. We assessed empirically the value of the white-backed woodpecker as an indicator and umbrella species in central Sweden. Occurrence of the woodpecker in breeding bird atlas squares (5 × 5 km²) indicated high species richness of forest birds, particularly species of special conservation concern, which included on average 13% more species in squares with than without the woodpecker. The number of red-listed cryptogam species expected to benefit from conservation actions directed at white-backed woodpecker habitats was higher in squares where the woodpecker bred compared to where is was absent. However, no such pattern was found for red-listed beetles, a group with very few records in the studied squares. White-backed woodpecker occurrence was positively associated with the current area of deciduous and mixed forest of high conservation value. Considering its indicator value, its specialised habitat requirements and its potential as a communication tool, using the white-backed woodpecker as an umbrella species may provide a coarse filter for the conservation of several other deciduous forest species. However, focusing solely on white-backed woodpecker habitat may not provide for the conservation of all such species, which stresses the need for a suite of complementary planning approaches.     

亚洲象生境长时序变化及其对人象冲突的影响——基于融合MaxEnt与HSI模型的生境适宜性评价方法

生境变化直接关系到物种种群维持与人类安全,揭示其长期变化规律及其对人类的影响,可为物种保护与生境恢复提供科学支撑。但因受物种活动点数据获取与位置精度的局限,鲜见濒危、危险物种的长时序生境变化研究。以人象冲突频发的西双版纳勐海—普洱澜沧地区亚洲象种群(勐海—澜沧象群)活动区为例,提出融合MaxEnt与HSI模型的亚洲象长时序生境适宜性评价方法,即基于荟萃分析筛选出的15个亚洲象生境评价因子,结合近期有限的物种活动点监测数据,利用MaxEnt得到生境评价因子的贡献率,再运用HSI模型计算生境适宜性指数;利用该方法制作出研究区1988—2020年逐年时序的亚洲象生境适宜性图,以分析亚洲象生境的时空变化,将其与亚洲象肇事数据结合,进而分析人象冲突与生境变化的关联。结果表明：(1)基于物种生境偏好不变的前提,融合MaxEnt模型与HSI模型的生境适宜性评价方法可应用于物种的长时序生境评价,且基于亚洲象活动点数据从动物对生境利用的生态学视角定量获取亚洲象对各生境评价因子的偏好程度,使生境评价结果具有良好的生态可解释性;(2)目前亚洲象适宜生境面积占研究区面积三分之一(4039.76 km^...     

How to deal with ground truthing affected by human‐induced habitat change?: Identifying high‐quality habitats for the Critically Endangered Red Siskin

Species distribution models (SDM) can be valuable for identifying key habitats for conservation management of threatened taxa, but anthropogenic habitat change can undermine SDM accuracy. We used data for the Red Siskin (Spinus cucullatus), a critically endangered bird and ground truthing to examine anthropogenic habitat change as a source of SDM inaccuracy. We aimed to estimate: (1) the Red Siskin's historic distribution in Venezuela; (2) the portion of this historic distribution lost to vegetation degradation; and (3) the location of key habitats or areas with both, a high probability of historic occurrence and a low probability of vegetation degradation. We ground‐truthed 191 locations and used expert opinion as well as landscape characteristics to classify species' habitat suitability as excellent, good, acceptable, or poor. We fit a Random Forest model (RF) and Enhanced Vegetation Index (EVI) time series to evaluate the accuracy and precision of the expert categorization of habitat suitability. We estimated the probability of historic occurrence by fitting a MaxLike model using 88 presence records (1960–2013) and data on forest cover and aridity index. Of the entire study area, 23% (20,696 km²) had a historic probability of Red Siskin occurrence over 0.743. Furthermore, 85% of ground‐truthed locations had substantial reductions in mean EVI, resulting in key habitats totaling just 976 km², in small blocks in the western and central regions. Decline in Area of Occupancy over 15 years was between 40% and 95%, corresponding to an extinction risk category between Vulnerable and Critically Endangered. Relating key habitats with other landscape features revealed significant risks and opportunities for proposed conservation interventions, including the fact that ongoing vegetation degradation could limit the establishment of reintroduced populations in eastern areas, while the conservation of remaining key habitats on private lands could be improved with biodiversity‐friendly agri‐ and silviculture programs.     

Neotropical moth assemblages degrade due to oil palm expansion

Oil palm is one of the most rapidly expanding crops throughout the tropics, yet little is known about its impacts on Neotropical invertebrate biodiversity. Responses of insect assemblages to land conversion may substantially vary among taxa. We assessed geometrid and arctiine moth assemblages in a Costa Rican human dominated landscape, where oil palm plantations are now the second most common land cover. Moths were sampled during 6 months with automatic traps in the interior and margin of old-growth forests, young secondary forests and oil palm plantations in a 30 km² area. Our results show that richness and diversity of both taxa were severely reduced in oil palm compared to all other habitats. Geometrid abundance was highest in forest interiors and lowest in oil palm, while arctiine numbers did not differ between habitats. Dominance was highest in oil palm plantations, where one arctiine species and one geometrid species accounted for over 40% of total abundance in each of their respective taxa. Species composition was distinct in oil palm and forest interior sites, and depicted a gradient of habitat disturbance in ordination space that was strongly related to vegetation diversity and structure. This study demonstrates that oil palm plantations are not a suitable habitat for these moth taxa. Whilst some arctiine species seem adapted to disturbed habitats, geometrids were more dependent on old-growth forests, showing higher bioindicator potential. In the face of accelerated oil palm expansion, conservation strategies should focus on protecting old-growth forest remnants, as well as increasing species diversity and structural complexity of degraded habitats.