4 Discussion
Assessing landscape connectivity could be a complex task because species have different biological features and migration capabilities, and the ways they perceive the landscape are also differentiated (Petsas et al., 2020 ). We used circuit theory-based approach to model the impacts of land use changes on landscape connectivity for North China leopard, and to determine the priority restoration areas. The results achieved in our work showed that land use changes and human activities have most important impacts on landscape connectivity, especially the changes of the migration corridors and key barrier areas. In order to protect North China leopard, we should not only protect the habitats where the species lived, but also improve landscape connectivity at a broad spatial scale.
Circuitscape borrows algorithms from electronic circuit theory to predict patterns of movement, gene flow, and genetic differentiation among plant and animal populations in heterogeneous landscapes. Circuit theory complements least-cost path approach because it considers effects of all possible pathways across landscape simultaneously, and could identify multiple alternative paths among all patches,so Electric circuit theory is becoming very popular for connectivity analysis (McRae et al., 2008 ). In this study, we aggregated information from expert opinions and various factors that potentially affect the migration of North China leopard, in an effort to provide the migration corridors and key barrier areas. This could be used as a systematic approach for conservation planning of North China leopard.
When applying the circuit theory-based connectivity model to analyze the functional connectivity, it is necessary to determine the landscape resistance surface, by the means of using sets of resistance values. The movement patterns of species is affected by several factors, including landscape properties and individual variability, so it is difficult to accurately obtain the landscape resistance surface when species migrate between habitat patches(Sahraoui et al., 2017 ; Petsas et al., 2020 ). Especially for the North China leopard, in addition to the infrared trigger camera research in several nature reserves, there is almost no large-scale radio continuous tracking research. In this study, to somehow overcome limitations, we evaluated the main factors affecting the migration of North China leopard by aggregating information from literature collation and expert opinion. Expert knowledge could not replace the information obtained from actual movement and species behavioral responses, but we do recognize that their contributions could be valuable, especially in the absence of species characteristics data at broad spatial scales. .
We determined 18 nature reserves (Fig. 2, 1. Taikuanhe, 2. Lishan, 3. Sushui River source, 4. Renzushan, 5. Huoshan, 6. Wulushan, 7. Lingkongshan, 8. Chaoshan, 9. Mengxinnao, 10. Sixiannao, 11. Tieqiaoshan, 12. Bafuling, 13. Pangquangou, 14. Heichashan, 15. Luyashan, 16. Xiaowutaishan, 17. Taihangshan, and 18. Manghe River Nature Reserve), the main activity regions of North China leopard, according to the investigation and consultation with relevant personnel. Finally, taking 18 nature reserves as core areas, we analyzed the landscape connectivity. It is considered that it is more in line with the actual situation to take these nature reserves where North China leopard is mainly distributed as the habitat patches (Cao et al., 2020 ).
Landscape connectivity modeling based on circuit theory needs to define a distance to characterize the movement capacity of species. Each species was characterized by its dispersal distance, which is different according to their physiological and behavioral attributes and difficult to obtain accurate data. In the study, we first extracted information on body weight and home range of North China leopard. Then based on the allometric growth equation, the maximum dispersal capacity were estimated according to body weight and home range, about 110 kilometers. The accuracy of the data needs to be further verified by field experiments. There are still more corridors longer than 110 kilometers, beyond the maximum migration capacity of species.
From 1990 to 2020, land use changes in the study area are characterized by the proportion of forest land increased by 8.24%, the grassland decreased by 10.97% and the built-up areas increased by 2.80%. Although forest land dominates the landscape type of habitats and activities of North China leopard, with the land use changes, especially the significant increase of forest land, the landscape connectivity has not been greatly improved. On the contrary, there is a decreasing trend in some key movement barrier areas.
In barrier area A (Fig. 5A, Fig. S3, Table S2), grassland and cropland are the main land use types. From 1990 to 2020, forest land increased 110.40 km2, with the percentage by 7.02%. But grassland decreased 234.71 km2 (by 14.93 percentage), and the area of cropland increased 108.76 km2 (by 6.92 percentage). As the result of land use changes, the migration path crossing the barrier area A is not detected in 2020. In barrier area B and D (Fig. 5B and 5D, Fig. S3, Table S2), the land use type is characterized by cropland. In 2020, the proportion of cropland was 76.46% and 55.28%, respectively, and the forest land was 6.5% and 13.75%, respectively. The proportion of forest land and grassland is low. From 1990 to 2000, the area of built-up areas has increased significantly. Within the barrier area B and D, the important barriers to migration corridors detected by Barrier Mapper indicate a further increasing trend. In barrier area C (Fig. 5C, Fig. S3, Table S2), the proportion of cropland increased by 23.94%, and the built-up areas increased by 158.64%, but the forest land increased by 259.48%, and the proportion reached to 27.11% in 2020. Within the barrier area C, barriers to migration corridors tend to slow down. Area E (Fig. 5E, Fig. S3, Table S2) is dominated by forest land and grassland. In 2020, the proportion of forest land and grassland was 51.12% and 25.48%, respectively. Comparing with 1990, the proportion changed by 60.04% and -51.55%, respectively. The new migration corridors appeared in 2020, which were more conducive to improve the connectivity between core areas and promote species migration.
The results of our analysis further reveal that conservation efforts of North China leopard should not only be limited to the protected areas, which traditionally tend to be higher altitudes and isolated sites. We also need to consider the landscape connectivity on a larger scale. Otherwise, it may render the conservation network less effective and make the distribution of species more isolated.
[CHART][CHART][CHART][CHART][CHART]
Fig.5 Land use changes within key barrier areas marked with A, B, C, D, and area E where connectivity effectively improved