Reforestation of agricultural land is an attractive topic for sustainable development. Its contributions, such as preventing surplus of agricultural products, mitigating farmland abandonment and restoring ecosystems, have been discussed worldwide (
Journal of Geographical Sciences, Volume. 30, Issue 9, 1419(2020)
Exploring the spatio-temporal impacts of farmland reforestation on ecological connectivity using circuit theory: A case study in the agro-pastoral ecotone of North China
Farmland reforestation can contribute substantially to ecological restoration. Previous studies have extensively examined the ecological effects of farmland reforestation, but few of them have investigated the spatiotemporal responses of broad-scale landscape connectivity to reforestation. By using a typical agro-pastoral ecotone in northern China as a case study, we addressed this issue based on an innovative integration of circuit theory approach and counterfactual analysis. The forest connectivity through multiple dispersal pathways was measured using the circuit theory approach, and its spatiotemporal changes after reforestation were evaluated by counterfactual analysis. The results showed that from 2000-2015, the reforested farmland occupied 2095 km2, and 12.5% was on steeply sloped land. Farmland reforestation caused a greater increase in ecological connectivity by adding new ecological corridors and stepping stones in scattered forest areas rather than in areas with dense forest distributions. The newly added corridors and stepping stones were fragmented, short and narrow and thus deserve powerful protection. Future reforestation to improve landscape connectivity should highlight pinch point protection and obstacle removal as well as the tradeoff between farmland loss and farmer survival. Our findings are expected to inform the optimization of the Grain for Green policy from the perspective of broad-scale biodiversity conservation.
1 Introduction
Reforestation of agricultural land is an attractive topic for sustainable development. Its contributions, such as preventing surplus of agricultural products, mitigating farmland abandonment and restoring ecosystems, have been discussed worldwide (
To date, numerous effects of farmland reforestation have been investigated in practice (e.g.,
Ecological connectivity is a key concern in ecosystem conservation. It describes habitat connections and the dispersal probability of species in the landscape (
Ecological connectivity can be measured from structural and functional perspectives, and functional connectivity is more popular because it can represent both physical structures and species behaviors in the landscape. Currently, three categories of methods have been developed to measure ecological connectivity: tracking investigation, landscape index analysis and mathematical simulation. The tracking investigation approach emphasizes field surveys of the migration flows of species. The method is direct and accurate but suffers from a heavy workload and a long duration of implementation (
In China, farmland reforestation efforts are also known as the “Grain for Green Project”, which stems from the pilot action of vegetation restoration to mitigate soil erosion and soil desertification (
For the above reasons, by using an agro-pastoral ecotone in southeastern Inner Mongolia as a case study, this article aims to (1) identify the spatiotemporal changes in ecological connectivity after farmland reforestation; (2) assess the potential improvement in ecological connectivity through reforestation; and (3) optimize reforestation policies from the perspective of biodiversity conservation. To do so, we propose an innovative framework based on the integration of circuit theory and counterfactual scenario analysis. This study is expected to provide an optimization strategy for farmland reforestation to achieve the goal of regional ecological security.
2 Materials and methods
The proposed framework for evaluating the influence of farmland reforestation on broad-scale ecological connectivity consists of three stages (
Figure 1.
2.1 Study area and data
Chifeng and Tongliao cities of southeastern Inner Mongolia, a typical agro-pastoral ecotone in northern China, were selected as the study area (
Figure 2.
A series of datasets were used in this research. (1) The empirical land use maps of 2000 and 2015, which indicate the land use status before and after farmland reforestation, were obtained from the Resources and Environment Data Center of the Chinese Academy of Sciences (
2.2 Analysis of ecological connectivity
2.2.1 Identifying ecological sources
Ecological sources are defined as habitats that support species survival and outward dispersal (
where
Ecological threat factors
Ecological threat factors
|
The required habitat area depends on the characteristics of species, and a large area of ecological sources can better benefit species radiation (
Sensitivity of land use types to threat factors
Sensitivity of land use types to threat factors
|
2.2.2 Analysis of key conservation patches, ecological corridors, pinch points and obstacles
Circuit theory links the dispersal process of species in ecology to the random walk of electrons in physics based on random walk theory and graph theory (
Based on Ohm’s law, the dispersal probability of species was inversely proportional to the effective resistance of the landscape. In circuit theory, the effective resistance can incorporate multiple pathways between two adjacent nodes (e.g., nodes 1 and 2), and it decreases as the number of branch resistors increases in a parallel circuit. This feature enables us to identify alternative dispersal paths between the landscape patches rather than the unique path with the minimum distance/cost, thus increasing the probability of ecological connectivity.
According to the study of McRae
Ecological corridors act as green channels that facilitate species dispersal in the landscape and are of great significance for biodiversity conservation. Accordingly, ecological corridors feature lower resistance values, and a larger current passes through these areas in circuit theory. Given an arbitrary pair of forest patches as two ecological sources, e.g., S1 and S2, if the 1A current flows through all effective landscape elements (resistors) between the two sources, then the current density can be calculated based on Ohm’s law. For all pairs of patches, the landscape elements with high cumulative current density can be selected as ecological corridors. The alternative connections between nodes 1 and S1 indicate the strength of circuit theory in identifying multiple dispersal pathways (
Figure 3.
Furthermore, the importance of habitat patches in maintaining the overall connectivity can also be evaluated based on the current density (
Pinch points and obstacles indicate the patches in the ecological network that influence the connectivity of ecological corridors if removed or converted. Pinch points refer to the narrow sections of ecological corridors, and their removal may result in the complete disconnection of ecological sources. For example, node 1 can be regarded as a pinch point because its removal would disconnect source patches S1 and S2 (
Circuitscape 4.0 and Linkage Mapper 2.0.0 (
2.3 Evaluating the effects of farmland reforestation on ecological connectivity
To distinguish the change in ecological connectivity induced by farmland reforestation, two scenarios were defined in the counterfactual analysis: actual-2015 and anti-2015. The actual-2015 scenario exhibited the empirical spatial pattern of land use after reforestation in 2015, and anti-2015 scenario simulated a counterfactual spatial pattern of land use that did not experience the reforestation process during the period 2000-2015. Specifically, the land use patches where reforestation was supposed to be applied would remain as cultivated land, while other land use conversions in the study area occurred as usual in the counterfactual scenario. In the actual and counterfactual scenarios, the ecological connectivity in terms of ecological corridors, key conservation areas, pinch points and obstacles were evaluated and compared using spatial overlay analysis.
3 Results and discussion
3.1 Land use change induced by reforestation
From 2000 to 2015, the area of reforested farmland was 2095 km2, accounting for 15.6% of the net increase in forestland (9263 km2). Of the total reforestated area, 49.8% came from farmland on a slope of 5°-15°, and 8.7% came from farmland on a slope greater than 15°. The reforested farmland accounted for 22.5% of the total farmland loss over 15 years, or 12.5% of the total sloping farmland in 2000. Fortunately, the total amount of farmland did not decrease because of the reforestation during the period 2000-2015 but instead increased by 2545 km2.
The reforested areas were mainly scattered in the eastern and southeastern counties, which suffered from smaller forest distributions, fragile ecosystems and serious soil desertification (
Figure 4.
3.2 Spatiotemporal changes in ecological connectivity
3.2.1 Ecological sources and key conservation areas
Significant expansion of the ecological sources was observed during the period 2000-2015, with an increase rate of 112.6%. In the spatial pattern (Figures 5a and 5b), the majority of the expansion occurred in the western fringe of the study area, which further improved the stability and penetration (i.e., high current density) of the source patches. For the eastern and central regions, a number of small-sized forest patches with high currents were newly added as ecological sources, which could provide additional alternative habitats for forest species.
The importance of ecological sources was classified into three levels: the first had current density values in the top 10%, the second had current density values in the top 10%-20% and the third had current density values in the top 20%-30% (Figures 5c and 5d). From 2000 to 2015, the number of the first- and third-level patches remained unchanged, while the proportion of the second-level patches increased by 8.7%. Meanwhile, the total area of the first-, second-, and third-level patches increased by 75.3%, 88.7%, and 22.6%, respectively. The important changes in the ecological sources in the north and the west mainly relied on spatial expansion of the first-level patches and their combination with low-level patches. The first-level patches in the north were mainly large-scale ecological sources that supported species survival, while those in the west may function as stepping stones for the north-south connection of ecological sources. In addition, key conservation areas should also consider the patches that are irreplaceable for landscape connectivity, for example, the newly developed ecological sources of the third-level importance in the east.
Figure 5.
Figure 6.
3.2.2 Ecological corridors, pinch points and obstacles
From 2000 to 2015, the ecological corridors became longer and wider in the west due to the expansion and consolidation of the ecological sources (Figures 6a and 6b). In the south, extensive new corridors simultaneously emerged to promote the east-west connectivity of ecological sources. In the east, however, only a small number of short and narrow corridors grew out of nothing near the existing ecological sources, which might not effectively mitigate the isolation of forest habitats. A total of 598 obstacle points for ecological connectivity were identified, 43.6% of which were smaller than 10 ha in area (
In 2000, the pinch points were mostly distributed in the northwest, and no pinch point existed in the east. In 2015, the pinch points increased significantly with the development of ecological corridors (
Figure 7.
3.3 Effects of farmland reforestation on ecological connectivity
The quantitative effects of reforestation on ecological sources, corridors and pinch points in different counties are presented in
Changes in ecological connectivity after reforestation in different counties (km2)
Changes in ecological connectivity after reforestation in different counties (km2)
|
Figure 8.
The comparison of the anti-2015 and actual-2015 scenarios showed that the reforestation caused the ecological sources to increase in area by 445.00 km2 but caused a decrease in the patch number and fragmentation. The enlargement and consolidation of ecological sources induced by the reforestation was dominant in the northwestern regions where forestland was densely distributed, while a handful of small, isolated forest patches were generated in the south and middle as stepping stones or corridors for the east-west dispersal of species. In the east, a large area of farmland was reforested to form corridors but exerted only a slight influence on the ecological connectivity due to the extremely sparse distribution of ecological sources and the fragmented, short and narrow characteristics of the new corridors.
Furthermore, the counterfactual analysis suggested the significant effects of farmland reforestation on the importance of forest habitats. Due to the reforestation policy, the importance of 8.83% of the forest patches (Type II) was upgraded from the third- to the second-level and that of 2.18% of the forest patches (Type IV) that were small in size (12 km2 in maximum) grew at turning points or endpoints of the newly added corridors in the central and eastern regions. In addition, farmland reforestation caused an important decline in and loss of forest patches surrounded by sandy grassland and barren areas (Types III and V, accounting for 8.33% in total). It is also noteworthy that reforestation did not affect the importance of 80.65% of the patches (Type I) that were densely distributed and large in size in the west (
Figure 9.
3.4 Discussion
From 2000 to 2015, the reforestation of agricultural land caused remarkable changes in forestland and farmland. Reforestation increased the area of forestland and effectively mitigated the forest loss induced by the development of the forestry economy. Meanwhile, reforestation also decreased the sloping farmland that suffered from soil erosion and/or ecological degradation but did not weaken the effectiveness of farmland protection. A series of land use policies, e.g., the reclamation of abandoned land for cultivation, has been proven feasible to supplement farmland loss after reforestation and to ensure the net growth in cultivated areas. However, it should also be noted that a strictly positive correlation, as suggested by the existing studies, did not exist between farmland reforestation and steep slopes in this case (
In this study, farmland reforestation substantially improved the overall landscape connectivity at the broad scale in the form of ecological corridors and/or stepping stones. Excessively sparse/dense distributions of forest patches reduce the effect of reforestation due to the positive correlation of landscape connectivity with the number of ecological sources but the negative correlation with the proximity of ecological sources (
From the ecological conservation perspective, the future potential implementation of the Grain for Green Project should highlight the protection of pinch points and the removal of obstacle points. The pinch points should be strengthened to protect the connection of ecological sources from human disturbances (
Circuit theory can integrate structural and functional connectivity to represent the dispersal probabilities of species and identify all potential corridors of species dispersal rather than the unique minimum cumulative cost paths. The integration of circuit theory and counterfactual analysis was feasible to evaluate how reforestation affected ecological connectivity and provided guidance for ecological restoration based on multiple scenarios. However, this study only evaluated the changes in ecological connectivity by comparing two point-in-time snapshots of landscapes while neglecting the temporal interactions of the forest patches (
4 Conclusions
Farmland reforestation is an essential ecological restoration measure for the development of ecological civilization in China. Based on circuit theory and the counterfactual scenario analysis approach, this article examined how the reforestation process affected the ecological connectivity of forest habitats in terms of ecological sources, corridors and obstacle points during the period 2000-2015 in the agro-pastoral ecotone of southeastern Inner Mongolia. In contrast to previous studies, this study identified the spatiotemporal changes in landscape connectivity at a broad scale after reforestation and provided solid support for optimization of the policy of returning farmland to forestland from the perspective of biodiversity conservation.
The results showed that the conversion of farmland to forest in the study area was 2095 km2, which was notable and heterogeneous in spatial distribution. This conversion mitigated forest loss but did not cause a significant decrease in cultivated land due to the contemporaneous policy of land reclamation. The reforestation efforts improved ecological connectivity in different ways. In the northern and western regions, which had high habitat quality and high forest cover, the reforestation policy led to an increase in the ecological sources through the expansion and consolidation of forest patches. In the southern and central regions, which had scattered forest distributions, the policy generated a large number of east-west corridors or stepping stones between the ecological sources and upgraded their importance levels. In comparison, the contribution of the Grain for Green Project to the promotion of landscape connectivity was more significant in the southern and central regions than in the northern and western. However, the ecological connectivity in the southern and central deserves better protection because the newly developed corridors are fragmented, short and narrow. Our results suggested that the future potential of farmland reforestation depends on the protection of pinch points and the removal of obstacle points, as well as the balance of farmland loss and farmer survival.
[2] Incorporating connectivity into reserve selection procedures. Biological Conservation, 103, 77-83(2002).
[3] et alDesigning a coherent ecological network for large mammals in northwestern Europe. Conservation Biology, 17, 549-557(2003).
[4] Evaluating China’s Slope Land Conversion Program as sustainable management in Tianquan and Wuqi Counties. Journal of Environmental Management, 98, 1916-1922(2011).
[7] et alUrban landscape pattern design from the viewpoint of networks: A case study of Changzhou city in Southeast China. Ecological Complexity, 8, 51-59(2011).
[11] et alExpert-based and correlative models to map habitat quality: Which gives better support to conservation planning?. Global Ecology and Conservation, 16, 13-26(2018).
[12] et alSpatio-temporal pattern and rationality of land reclamation and cropland abandonment in mid-eastern Inner Mongolia of China in 1990-2005. Environmental Monitoring and Assessment, 179, 137-153(2011).
[14] Improving landscape connectivity in forest districts: A two-stage process for prioritizing agricultural patches for reforestation. Forest Ecology and Management, 261, 154-161(2011).
[15] et alA counterfactual scenario simulation approach for assessing the impact of farmland preservation policies on urban sprawl and food security in a major grain-producing area of China. Applied Geography, 37, 127-138(2013).
[17] Responses of 20 native tree species to reforestation strategies for abandoned farmland in Panama. Ecological Applications, 12, 1626-1641(2002).
[18] et alOver twenty years farmland reforestation decreases fungal diversity of soils, but stimulates the return of ectomycorrhizal fungal communities. Plant and Soil, 427, 231-244(2018).
[19] et alLandscape connectivity for wildlife: Development and validation of multispecies linkage maps. Methods in Ecology and Evolution, 5, 626-633(2014).
[22] et alStructure optimization of circuit theory-based green infrastructure in Nanjing, China. Acta Ecologica Sinica, 28, 4363-4372(2018).
[23] et alBacterial community structure in atrazine treated reforested farmland in Wuying China. Applied Soil Ecology, 98, 39-46(2016).
[24] Spatio-temporal characteristics of soil protection efforts of the Grain for Green Project in northern Shaanxi Province. Journal of Geographical Sciences, 30, 401-422(2019).
[28] Spatio-temporal connectivity: Assessing the amount of reachable habitat in dynamic landscapes. Methods in Ecology and Evolution, 8, 1253-1264(2017).
[32] et alWhere to restore ecological connectivity? Detecting barriers and quantifying restoration benefits. PLoS ONE,, 7, e52604(2012).
[33] Circuit theory to estimate natal dispersal routes and functional landscape connectivity for an endangered small mammal. Landscape Ecology, 32, 1163-1179(2017).
[41] A note on creating robust resistance surfaces for computing functional landscape connectivity. Ecology and Society, 10, 610-611(2005).
[49] et alExpert knowledge based valuation method of ecosystem services in China. Journal of Environmental Management, 23, 911-919(2018).
[51] et alDeveloping ecological networks based on MSPA and the least-cost path method: A case study in Bazhong western new district. Acta Ecologica Sinica, 35, 6425-6434(2015).
[54] et alDeveloping and optimizing ecological networks in urban agglomeration of Hunan Province, China. Acta Ecologica Sinica, 31, 2863-2874(2011).
Get Citation
Copy Citation Text
Xiaojing LIU, Dianfeng LIU, Hongzhuo ZHAO, Jianhua HE, Yaolin LIU. Exploring the spatio-temporal impacts of farmland reforestation on ecological connectivity using circuit theory: A case study in the agro-pastoral ecotone of North China[J]. Journal of Geographical Sciences, 2020, 30(9): 1419
Category: Research Articles
Received: Mar. 15, 2020
Accepted: Jun. 18, 2020
Published Online: Apr. 21, 2021
The Author Email: LIU Dianfeng (liudianfeng@whu.edu.cn)