scholarly journals Exploration in Mapping Kernel-Based Home Range Models from Remote Sensing Imagery with Conditional Adversarial Networks

2018 ◽  
Vol 10 (11) ◽  
pp. 1722
Author(s):  
Ruobing Zheng ◽  
Guoqiang Wu ◽  
Chao Yan ◽  
Renyu Zhang ◽  
Ze Luo ◽  
...  
Keyword(s):  
Remote Sensing ◽  
Home Range ◽  
Large Scale ◽  
Qinghai Lake ◽  
Conservation Strategies ◽  
Home Ranges ◽  
Lake Area ◽  
Range Analysis ◽  
Remote Sensing Imagery ◽  
Adversarial Network

Kernel-based home range models are widely-used to estimate animal habitats and develop conservation strategies. They provide a probabilistic measure of animal space use instead of assuming the uniform utilization within an outside boundary. However, this type of models estimates the home ranges from animal relocations, and the inadequate locational data often prevents scientists from applying them in long-term and large-scale research. In this paper, we propose an end-to-end deep learning framework to simulate kernel home range models. We use the conditional adversarial network as a supervised model to learn the home range mapping from time-series remote sensing imagery. Our approach enables scientists to eliminate the persistent dependence on locational data in home range analysis. In experiments, we illustrate our approach by mapping the home ranges of Bar-headed Geese in Qinghai Lake area. The proposed framework outperforms all baselines in both qualitative and quantitative evaluations, achieving visually recognizable results and high mapping accuracy. The experiment also shows that learning the mapping between images is a more effective way to map such complex targets than traditional pixel-based schemes.

scholarly journals Changes in landscape composition influence the decline of a threatened woodland caribou population

2020 ◽  
Author(s):  
Heiko Wittmer ◽  
B McLellan ◽  
R Serrouya ◽  
C Apps
Keyword(s):  
Large Scale ◽  
Spatial Scales ◽  
Apparent Competition ◽  
Timber Harvesting ◽  
Conservation Strategies ◽  
Home Ranges ◽  
Vital Rates ◽  
Population Declines ◽  
Woodland Caribou ◽  
Habitat Alterations

Large-scale habitat loss is frequently identified with loss of biodiversity, but examples of the direct effect of habitat alterations on changes in vital rates remain rare. Quantifying and understanding the relationship between habitat composition and changes in vital rates, however, is essential for the development of effective conservation strategies. It has been suggested that the decline of woodland caribou Rangifer tarandus caribou populations in North America is precipitated by timber harvesting that creates landscapes of early seral forests. Such habitat changes have altered the predator-prey system resulting in asymmetric predation, where predators are maintained by alternative prey (i.e. apparent competition). However, a direct link between habitat condition and caribou population declines has not been documented. We estimated survival probabilities for the threatened arboreal lichen-feeding ecotype of woodland caribou in British Columbia, Canada, at two different spatial scales. At the broader scale, observed variation in adult female survival rates among 10 distinct populations (range = 0.67-0.93) was best explained by variation in the amount of early seral stands within population ranges and population density. At the finer scale, home ranges of caribou killed by predators had lower proportions of old forest and more mid-aged forest as compared with multi-annual home ranges where caribou were alive. These results are consistent with predictions from the apparent competition hypothesis and quantify direct fitness consequences for caribou following habitat alterations. We conclude that apparent competition can cause rapid population declines and even extinction where changes in species composition occur following large scale habitat change. © 2007 The Authors. Journal compilation © 2007 British Ecological Society.

scholarly journals Improved Winter Wheat Spatial Distribution Extraction Using A Convolutional Neural Network and Partly Connected Conditional Random Field

2020 ◽  
Vol 12 (5) ◽  
pp. 821 ◽  
Author(s):  
Shouyi Wang ◽  
Zhigang Xu ◽  
Chengming Zhang ◽  
Jinghan Zhang ◽  
Zhongshan Mu ◽  
...  
Keyword(s):  
Remote Sensing ◽  
Spatial Distribution ◽  
Winter Wheat ◽  
Random Field ◽  
Large Scale ◽  
Conditional Random Field ◽  
Pixel Classification ◽  
Classification Result ◽  
Edge Pixel ◽  
Remote Sensing Imagery

Improving the accuracy of edge pixel classification is crucial for extracting the winter wheat spatial distribution from remote sensing imagery using convolutional neural networks (CNNs). In this study, we proposed an approach using a partly connected conditional random field model (PCCRF) to refine the classification results of RefineNet, named RefineNet-PCCRF. First, we used an improved RefineNet model to initially segment remote sensing images, followed by obtaining the category probability vectors for each pixel and initial pixel-by-pixel classification result. Second, using manual labels as references, we performed a statistical analysis on the results to select pixels that required optimization. Third, based on prior knowledge, we redefined the pairwise potential energy, used a linear model to connect different levels of potential energies, and used only pixel pairs associated with the selected pixels to build the PCCRF. The trained PCCRF was then used to refine the initial pixel-by-pixel classification result. We used 37 Gaofen-2 images obtained from 2018 to 2019 of a representative Chinese winter wheat region (Tai’an City, China) to create the dataset, employed SegNet and RefineNet as the standard CNNs, and a fully connected conditional random field as the refinement methods to conduct comparison experiments. The RefineNet-PCCRF’s accuracy (94.51%), precision (92.39%), recall (90.98%), and F1-Score (91.68%) were clearly superior than the methods used for comparison. The results also show that the RefineNet-PCCRF improved the accuracy of large-scale winter wheat extraction results using remote sensing imagery.

Spatial ecology of the Mariana Crow Corvus kubaryi: Implications for management strategies

2018 ◽  
Vol 29 (4) ◽  
pp. 527-541 ◽  
Author(s):  
SARAH K. FAEGRE ◽  
LINDSEY NIETMANN ◽  
DYLAN HUBL ◽  
JAMES C. HA ◽  
RENEE R. HA
Keyword(s):  
Home Range ◽  
Spatial Ecology ◽  
Large Scale ◽  
Life Stage ◽  
Management Strategies ◽  
Home Ranges ◽  
Spatial Behaviour ◽  
Island Endemic ◽  
Species Specific ◽  
Over Time

SummaryKnowledge of species-specific spatial ecology is critical for applying appropriate management strategies to maximise conservation outcomes. We used radio-telemetry to describe spatial behaviour of the critically endangered, island-endemic Mariana Crow Corvus kubaryi. To determine whether management strategies should reflect life stage, we measured the home ranges and daily movements of 22 Mariana Crows. Fledgling mobility was low during the first 31 days post-fledging and effects of age (fledgling or sub-adult) and time (months post-fledging or post-dispersal) were often driven entirely by this period. After controlling for reduced fledgling mobility, cumulative home range size increased over time for both age classes and was, on average, more than twice the area for sub-adults than fledglings. Sub-adults also tended to make longer daily movements than fledglings. Non-cumulative, monthly home range areas did not increase over time but the average overlap in home range area between consecutive months was only 63%, suggesting large shifts in space use each month. These results highlight the dynamic nature of Mariana Crow home ranges and suggest that large-scale management efforts are critical for protecting both breeding and non-breeding individuals. The application of the traditional home range concept to Mariana Crows and other wide-ranging passerine birds may result in sub-optimal management strategies. Instead, we recommend that the spatial and temporal scale of conservation efforts be informed by species-specific spatial behaviour across all relevant life stages.

Effects of food and predators on the home-range sizes of Arctic ground squirrel (Spermophilus parryii)

1998 ◽  
Vol 76 (3) ◽  
pp. 592-596 ◽  
Author(s):  
Anne H Hubbs ◽  
Rudy Boonstra
Keyword(s):  
Home Range ◽  
Ground Squirrel ◽  
Large Scale ◽  
Yukon Territory ◽  
Ground Squirrels ◽  
Home Ranges ◽  
Spermophilus Parryii ◽  
Mammalian Predators ◽  
Core Areas ◽  
Arctic Ground Squirrels

We used radiotelemetry to study the effects of food addition and predator reduction on the home-range sizes of adult Arctic ground squirrels (Spermophilus parryii) on large-scale experimental grids in the boreal forest of the southwestern Yukon Territory. Home ranges were 2-7 times smaller on food-supplemented grids than on nonsupplemented grids, regardless of whether large mammalian predators were present. Similarly, core areas (where 50% of activities occur) were 8-11 times smaller on food-supplemented grids. Food availability rather than predator presence primarily determined the sizes of home ranges and core areas of Arctic ground squirrels.

The Dynamic of Qinghai Lake Area for Remote Sensing Monitoring and the Analysis of its Driving Force

2014 ◽  
Vol 971-973 ◽  
pp. 2162-2167
Author(s):  
Bao Kang Liu ◽  
Yu E Du ◽  
Ai Jun Hu
Keyword(s):  
Remote Sensing ◽  
Water Body ◽  
Driving Force ◽  
Qinghai Lake ◽  
Lake Area ◽  
Body Area ◽  
Large Lakes ◽  
Important Means ◽  
Area Variation ◽  
Environmental Mitigation

Remote sensing (RS) and geographic information system (GIS) are an important means of monitoring water body areas for large lakes and water reservoirs. This study is based on environment and mitigation satellite CCD data characteristics of China's own R & D, using the combining method of multi-band and single-band to construct a water body identification patterns based on environmental mitigation satellite and extracting the water body area of Qinghai Lake since the September of 2008 to 2011 during April - November each year. Then the variations characters of the annual, seasonal and monthly were analyzed during 2008 and 2011. At last, induced its driving force of area variation of Qinghai Lake.

Natal dispersal of grizzly bears

2001 ◽  
Vol 79 (5) ◽  
pp. 838-844 ◽  
Author(s):  
Bruce N McLellan ◽  
Frederick W Hovey
Keyword(s):  
Home Range ◽  
Ursus Arctos ◽  
Natal Dispersal ◽  
Conservation Strategies ◽  
Grizzly Bears ◽  
Home Ranges ◽  
Adult Males ◽  
Gradual Process ◽  
The Social ◽  
Dispersal Distances

We studied natal dispersal of grizzly bears (Ursus arctos), a solitary nonterritorial carnivore with a promiscuous mating system, between 1979 and 1998. Dispersal distances for 2-year-olds did not differ between males and females, but by 3 years of age, males had dispersed farther than females, and farther still by 4 years of age. Dispersal of both sexes was a gradual process, occurring over 1–4 years. From the locations of death, or last annual ranges, it was estimated that 18 males dispersed 29.9 ± 3.5 km (mean ± SE) and 12 females dispersed 9.8 ± 1.6 km. Eleven of these males dispersed the equivalent of at least the diameter of 1 adult male home range, whereas only 3 of the females dispersed at least the diameter of 1 adult female home range. The longest dispersals recorded were 67 km for a male and 20 km for a female. Because the social system consists of numerous overlapping home ranges of both sexes, long dispersal distances may not be required to avoid inbreeding or competition with relatives. Simple models suggest that 61% of the ranges of brother and sister pairs would not overlap, but the home range of every daughter would overlap her father's range. The home range of an estimated 19 ± 4 (mean ± SD) adult males, however, would overlap at least a portion of each female's range, thereby reducing the chance of a female mating with her brother or father. Understanding the dispersal behaviour of grizzly bears is essential for developing conservation strategies. Our results suggest that meta-population reserve designs must provide corridors wide enough for male grizzly bears to live in with little risk of being killed.

Sign in / Sign up

Export Citation Format

Share Document