Landscape inf luences and management countermeasures of ginseng planting near Changbai mountain nature reserve

Mingfang Tang · Sha Qiu,2 · Lijie Liu · Tao Li,2 ·Shanlin Li · Tianshu Yu

Abstract Ginseng planting in Northeast China brings economic benef its but aff ects forest landscape integrity and native ecological processes. In order to quantify the impacts of ginseng planting on the forest landscape, Fusong County in Jilin Province was selected as a study area. The number and distribution of ginseng f ields over diff erent time was quantif ied based on remote sensing and ground surveys. Grid analysis and multiple regression analysis were used to study the impacts of ginseng planting on the landscape. The results showed that altitudes and slopes of ginseng f ields increased and became increasingly scattered and smaller closer to the Changbai Mountain Nature Reserve. Ginseng f ields and abandoned f ields increased total patches and total area of the local forested landscape, and shared edge lengths between ginseng f ields and forests, resulting in continuous fragmentation of the landscape. Although the total area of existing and abandoned ginseng f ields accounts for a small fraction of the total landscape, their negative impacts on ecosystem conservation is signif icant. The local government needs to rationally plan ginseng planting, scientif cially implement the restoration of abandoned ginseng lands, and enhance awareness of ginseng farmers to environmental stewardship. Our study has important signif icance for maintaining the healthy and stable development of the local ginseng industry and for improving the quality of regional ecological environment.

Keywords Ginseng planting · Agroforestry system ·Land use · Landscape fragmentation · Forest degradation ·Changbai Mountain Nature Reserve

Introduction

The inf luence of human activities on landscape patterns,processes and dynamics is an important aspect of landscape ecology (Forman and Godron 1986; Fu et al. 2001; Linderman et al. 2006). Before anthropogenic activities, the main driving force of landscape change was the succession process by natural factors. However, the increasing scope and intensity of human activities has become the main driving force of forest landscape change, and has become a research focus of forest landscape ecology (Li et al. 2006; Lira et al.2012; Shif ley et al. 2017). According to their scope, frequency, and intensity, human activities, related to the landscape, may be divided into two categories: land utilization and forest management (Li et al. 2006). Land utilization includes agricultural production and urbanization expansion,which change forested landscapes into long-term, often permanent non-forest landscapes. Forest management is a series of activities with harvesting and regeneration as the main objectives and sustainable utilization as the goal.

Ginseng (Panax ginsengC.A. Mey) is a herb belonging to the Araliaceae family and is both wild and cultivated, the latter termed garden ginseng. Wild ginseng grows naturally in forests and distributed in East Asia such as China and Korea and North America such as America and Canada.Garden ginseng is ginseng planted by clear-cutting forests to make ginseng f ields. China is a major ginseng producing country. In the Changbai Mountains, the uniquely suitable natural environment provides favorable conditions for planting ginseng planting and has become the main producing area of ginseng, and accounts for about 85% of the country’s production (Sheng et al. 2012). The ginseng industry plays an important role in economic development in the Changbai Mountain area and ref lects traditional ginseng culture with regional characteristics (Qiao 2019).

Ginseng planting has become the main source of ginseng resources, with large growing areas, high yields, and a short growth cycle (Ding et al. 2012). The planting cycle from sowing to harvesting is six years. Two or three years after sowing, the plants are transplanted to new ginseng f ields due to the changes of soil properties in the seedbeds. The ginseng is harvested after three more years of growing. There are numerous research studies on remote sensing interpretation and spatial distribution mapping of ginseng f ields, ginseng planting technology, and the eff ects of ginseng planting on soil physical and chemical properties (Ding et al. 2012,2014). Because ginseng has negative eff ects on soil (Han et al. 1998; Shan 2009; Xu et al. 2014), such as changing pH, reducing fertility, often leading to soil toxicity, ginseng f ields are often abandoned after two or three years. Although nearby forests may provide abundant tree seed sources, and a relatively stable secondary forest may be formed after 20-30 years of fallow, ginseng planting still results in the fragmentation of the forest landscape and the degradation of the forest ecosystem. To some extent, the inf luence of ginseng planting on forest landscapes is ref lected in changing the structure and pattern of land use and in increasing the diffi culty of forest management. However, there have been limited eff orts to studying the impact of ginseng planting on forest landscapes over a long time.

Fusong County in Jilin Province, adjacent to the Changbai Mountain Nature Reserve, is the well-known location of ginseng, and brings in considerable economic benef its,but also aff ects the regional forest landscapes and ecological environment. Nature reserves can generally survive the impact of urbanization and eff ectively reduce the fragmentation of forests within and even around the reserve (Janeczko et al. 2019). However, due to the eff ect of continuous deforestation, ginseng planting, and urban expansion, the forested landscapes surrounding the Changbai Mountain Nature Reserve has been fragmented, causing the Reserve to become an “isolated island” (Tang et al. 2011; Deng et al.2016; Ouyang et al. 2016; Dai et al. 2019). In recent decades, ginseng planting, as a traditional agroforestry system(Santoro et al. 2020), has changed from disorderly to formalized cultivation. However, the long-term disorderly or random cultivation of ginseng of the past has a lasting impact on the forest landscape and the environment. Therefore, the eff ect of ginseng planting on forest landscapes and habitat quality around the Reserve should be determined. This study combined remote sensing imaging with ground surveys to study the number and distribution of ginseng f ields in different periods, to analyze their dynamics and inf luencing factors, and to quantitatively evaluate the impact of ginseng planting on regional forest landscapes. This can provide information for the regional sustainable development of ginseng industry and the protection of the Changbai Mountain Nature Reserve.

Materials and methods

Study area

Fusong County is in the southeast of Jilin Province,127°01′-128°06′ E and 41°42′-42°49′ N, adjacent to the Changbai Mountain Nature Reserve (Fig. 1). It is a county under the jurisdiction of Baishan City, a key forest county and a famous ginseng producing area. The topography inclines gradually from the southeast to northwest, and is part of the Changbai Mountain landform area (Shan 2009).The area is a humid, cold temperate zone, with annual average temperatures of 1.9-4.3 °C due to the diff erences of terrain (Yang et al. 2010). There are abundant plant species in the territory, typical of a temperate zone. The vegetation is broad-leaved Korean pine forest. However, due to human disturbances such as logging, natural secondary forest and plantations account for a large proportion of the forest cover.The study area, within the red line is 124 km from north to south, 87 km wide from east to west, and the total area is 5173 km 2 (Fig. 1). The area in the lower right corner is the Changbai Mountain Nature Reserve and is not included in this study. The Changbai Mountain Nature Reserve was established in 1960 and is one of the earliest nature reserves in China. It was upgraded to a National Nature Reserve in 1986 and is classif ied as a forest ecosystem in the natural ecosystem category. It is important to maintain forest ecosystem balance and biodiversity security in Northeast China.In 2005, the Changbai Mountain Protection Development Zone Management Committee was established to undertake"unif ied planning, unif ied protection, unif ied development and unif ied management" of Changbai Mountain. Since the establishment of the Reserve 60 years ago, protection within the reserve has been good. However, activities such as tree cutting, urbanization and tourism facilities, fragmentation of forests around the Reserve has been obvious, leading to isolation of the Reserve (Deng et al. 2016).

Fig. 1 Location of the study area

Materials

The remote sensing images used in this study are Landsat TM-5 data with 30 m resolution ( http:// earth explo rer. usgs.gov/; http:// www. gsclo ud. cn). According to the development of the ginseng industry in Fusong County (1978-1989, rapid development stage; 1990-1996, adjustment stage; since 1996, steady development stage) and the Reserve, and considering the availability and quality of data, remote sensing images in f ive periods of 1989, 1995, 2002, 2010 and 2014 were selected for analysis. The DEM data were derived from GDEMV2 ( http:// www. gsclo ud. cn/ ), and the resolution is 30 m × 30 m.

Methods

Remote sensing interpretation

In this study, due to the small area of unused land and grasslands in Fusong County, according to the main land use characteristics, the area was divided into f ive types: forest,croplands, ginseng f ields, wetlands (including waters), and construction lands.

The main purpose of ground feature recognition is to distinguish the existing ginseng f ields from other land use types. How to identify and segment the patches accurately is the main problem. In general, ginseng f ields were reclaimed into a regular rectangle. During almost the entire growing period, blue, purple, or white plastic f ilm was spread of 1-2 m above the ground for shading, rain proof ing, and for adjusting the temperature. These features can help to distinguish ginseng f ields from surrounding lands. This was also used in another study (Chen et al. 2005). Based on the spectral feature, the rule set of automatic extraction of existing ginseng f ields could be determined.

In the remote sensing images, 80 ginseng f ields were randomly selected, and checked by f ield observations, visiting ginseng farmers, and by consulting relevant management departments. The results showed that only f ive patches were misclassif ied, and the accuracy of ginseng patches classif ication was 93.7%.

Landscape pattern indexes

A landscape pattern index is a quantitative index that indicates the structure and distribution of landscape patterns and other features, including patch level, patch class level and landscape level (Mcgarigal and Marks 1995). The patch class level ref lects the structural characteristics of diff erent patch types in the landscape, while the landscape level ref lects the overall structural characteristics of the landscape(Mcgarigal and Marks 1995). Ginseng planting changes the structure and distribution of the forest landscape, leading to its fragmentation. Therefore, considering the high correlation and redundancy between indexes, nine representative class and landscape level indexes were selected to describe the landscape pattern of the study area from the aspects of area, aggregation and diversity (Chen et al. 2002) and to quantify the impact of ginseng planting on the landscape.The nine landscape pattern indexes included: Class Area(CA), Percentage of Landscape (PLAND), Number of Patches (NP), Patch Density (PD), Edge Length (EL), Splitting Index (SPLIT), Landscape Division Index (DIVISION),Connectance Index (CONNECT), Contagion (CONTAG),while DIVISION, CONNECT and CONTAG were used for landscape level analysis. The landscape pattern of Fusong County from 1989 to 2014 was analyzed using the software Fragstats v4 (Mcgarigal and Marks 1995). The meaning and calculation method of the landscape pattern indexes are shown in Table S1.

Grid analysis

The Fishnet tool in ArcGIS software was used to divide the study area into 78 10 km × 10 km grids and the land use type in each grid was obtained through Extract by Mask tool.Among the 78 grids, ones with less than 80% of the grid area within the study were excluded. Thirty-six grids were selected to study the eff ect of the NP and CA of diff erent patch types on the overall landscape pattern using multiple regression analysis.

Results

Spatial pattern of land use

From the perspective of land use composition, forests were the main land use in Fusong County, accounting for more than 86% of the total area (Table 1). Due to regional development, the expansion of construction lands and the decrease of croplands were the main land use changes. The area of construction lands has increased by 459.1%, the area of croplands by 30.9%. Moreover, the area of ginseng f ield also was drastically reduced.

The land use pattern of Fusong County from 1989 to 2014 showed regional characteristics (Fig. 2). Due to the high southeast and low northwest topography, forests were concentrated in the eastern part of the county, and croplands in the central and western parts where the terrain was relatively f lat. Construction lands were distributed at lower elevation and showed aggregation distribution. Construction lands,croplands and forests showed a nested pattern. The area surrounding construction lands were mainly croplands, while the periphery of croplands was forests, showing the spatial gradient change from an artif icial environment to a natural one. Ginseng f ields with relatively smaller patch area andtotal area was mainly distributed in the interior of the forests,showing a discrete distribution.

Table 1 Land use composition of Fusong County 1989-2014

Fig. 2 Land use patterns of Fusong County 1989-2014

Patch characteristics of ginseng f ields in diff erent years

The number and total area of ginseng f ield patches f luctuated over 1989-2014, but the number trended upwards,while the total area showed a downward trend, resulting in a decrease in average area of f ield patches (Table 2). At the same time, the average altitude and slope trended upwards.As a result, the center of ginseng f ield patches gradually shifted to the southeast and closer to the Changbai Mountain Nature Reserve. As a result of the changes of patch number and area, PD of ginseng f ields increased, PLAND decreased,and SPLIT increased signif icantly, indicating that the dispersed distribution of ginseng f ields.

The recovery process of each abandoned ginseng f ield was complicated, even resulting in an increase in the number of new patches. The quantitative dynamics of new patches formed within the abandoned ginseng f ields during the recovery period are shown in Table 3. For example, the 868 ginseng f ield patches in 1989 became 1845, 1895, 1823and 1851 new patches after abandonment in 1995, 2002,2010 and 2014, respectively, and most of these patches were in forests at the boundary of the original ginseng f ields.Because these forest patches could merge with the surrounding forest, there was little impact on landscape fragmentation. However, the numbers of non-forest patches within the 868 abandoned ginseng f ields were 781, 817, 713 and 803 in 1995, 2002, 2010 and 2014, respectively, and indicates that the restoration of early abandoned ginseng f ields was poor. Ginseng f ields in 1995 and 2002 also became the largest number of new patches in the initial stage of abandonment, but the number of total patches and non-forest patches formed as a result of abandoned ginseng f ields decreased over time. Since the 1990s, with greater ecological protection consciousness, former ginseng f ields had a better restoration during fallow stage.

Table 2 Patch characteristics of ginseng f ields in diff erent years

Table 3 Patch dynamics of abandoned ginseng f ields

Inf luence of ginseng f ields on forest landscape

Due to the small proportion of ginseng f ields to the regional area, using the land use data of 1989 as an example, multiple regression analysis combined with grid analysis quantif ied the impacts of patch number and patch area of ginseng f ields and other land uses on the regional landscape. Ten independent variables, including patch number and area of f ive land use types were used for multiple regression. The dependent variables were DIVISION, CONNECT, and CONTAG,ref lecting the degree of separation of patch distribution, the degree of connection between patches, and the degree of agglomeration or spread of diff erent types of patches in the landscape, respectively.

Multiple regression analysis of 36 grids showed that CA and NP of croplands, the NP of forests and CA of wetlands were correlated with CONTAG (Table 4). The CA of forest and NP of croplands and wetlands correlated with DIVISION. While the dependent variable was CONNECT, there were no clear statistical results. Therefore, the main land uses aff ecting the landscape pattern were croplands, forests,and wetlands. The total area of ginseng f ields were kept at a low proportion and had no eff ect on the index DIVISION,CONNECT, and CONTAG.

The results show that croplands were the main factor aff ecting the regional landscape. Because ginseng f ields were mainly located in the forest, to eliminate the inf luence of croplands, 20 grids with a proportion of cropland less than 10% were selected to repeat the same multiple regression analysis. The results show that the CA of croplands and construction lands, the NP of forests and the CA of ginseng f ields were correlated with CONTAG (Table 4). The CA of croplands, wetlands, ginseng f ields, construction lands, and the NP of forests were correlated with DIVISION. While the dependent variable was CONNECT, there was no clear statistical result. This implies that croplands were the main factor aff ecting the landscape pattern in areas with lowdisturbance intensity, and that other land uses also had a signif icant impact on the landscape. In this situation, the total area of ginseng f ields aff ected DIVISION and CONTAG.

Table 4 Results of multiple regression analysis

Table 5 Edge length dynamics from 1989 to 2014

To further explain the inf luence of ginseng f ields on forest landscapes, the edge lengths of forests (ELF), as well as shared edge lengths between ginseng f ields and forests(ELGF), were counted (Table 5). The results show that,although ginseng f ields were proportionally small, they contributed an average 8.5% of the forest edge.

Discussion

Eff ects of ginseng f ields on regional forest landscapes

Fusong County, which is adjacent to the Nature Reserve, is a principal forestry county and well-known for ginseng in the Changbai Mountain area. Its urbanization, land use changes,timber cutting, and management practices strongly aff ected the regional landscapes and had a detrimental impact on the Reserve. Research showed that establishing a nature reserve could reduce the impact of urbanization and eff ectively reduce the fragmentation of forests within and around the protection area (Tang et al. 2011). This study showed that the area of forest increased from 1989 to 2014. However,surrounding areas of Changbai Mountain Nature Reserve were once the main production base for commodity timber and the forested landscapes around the Reserve were fragmented after years of timber harvesting and rapid urbanization. Due to the establishment of the Changbai Mountain Protection and Development Zone in 2005, construction lands expanded. Although continuous reforestation has led to increased forest area, timber harvesting and other disturbances have resulted in landscape fragmentation and habitat degradation, thus turning the Changbai Mountain Nature Reserve into an “isolated island” (Deng et al. 2016; Ouyang et al. 2016; Dai et al. 2019). Our results also show that the increased area of cropland was the major factor aff ecting regional landscapes.

Although commercial timber production and urbanization were the main reasons causing forest landscape fragmentation, ginseng planting as a high intensity anthropogenic activity also had inf luence. The number of patches in an area had a positive correlation with landscape fragmentation, and in some studies (Huang et al. 2020; Li and Deng 2020), the number of patches was used as an index of landscape fragmentation. Our study showed that ginseng f ields had a high patch number, accounting for 24% of the total patches, and had a signif icant impact on landscape fragmentation. However, the proportion of patch area and total area of ginseng f ields were low, so the eff ects of ginseng f ields on division, agglomeration, and connection degree of regional landscapes were not obvious. Therefore, 20 grids were selected with the proportion of cropland area less than 10% to further study the inf luence of ginseng f ields on partial landscapes under lesser human disturbances. The results show that ginseng planting inf luenced the degree of landscape division and agglomeration. The existing ginseng f ields and abandoned f ields increased the forest edge length and exterior forest, where factors such as light, humidity,and temperature might exceed the tolerance limits of the original species, leading to increased mortality (Williams-Linera 1990). Nutrient and carbon levels in the exterior forest will be signif icantly lower than that of the interior forest, resulting in poor vegetation growth, leading to changes in the structure and composition of the original forest and aff ecting ecosystem services (Lin and Cao 2009; Cardelús et al. 2020).

The recovery of abandoned ginseng f ields was complicated, and diffi cult to restore to the original forest landscape.In diff erent periods, various measures were given to their restoration, leading to diff erent eff ects. If the abandoned ginseng f ield was not eff ectively restored, the number of non-forest patches would remain high for a long period, and the deleterious impact of abandoned ginseng f ields in diff erent periods would accumulate. In addition, ginseng planting aff ected soil characteristics such as reducing soil pH and fertility and even leading to toxicity, which caused diffi culties for vegetation renewal and aff ected the growth of surrounding forests (Shan 2009; Xu et al. 2014). The gradual decrease of f lat lands suitable for ginseng planting in low-altitude areas forced the increasingly scattered ginseng f ields closer to the Reserve, which may further aggravate the fragmentation of the landscape and aff ect the ecological environment of the Changbai Mountain Nature Reserve. Therefore, relevant government departments should implement restoration measures to restore abandoned ginseng f ields to reduce the impact of ginseng planting on forest landscapes.

Countermeasures for the development of ginseng planting

Research has shown that the degree of fragmentation of the forest landscape caused by logging activities was higher than that caused by natural disturbances such as f ire, and the specif ic impacts included changes of forest stand pattern, forest patch density and patch size, as well as core habitat loss(Gustafson and Crow 1994; Wang et al. 2014). After the prohibition of logging natural forests in China, ginseng planting became one of the main reasons for landscape fragmentation around Changbai Mountain. This brought a series of negative impacts on habitat quality and biodiversity (Halkos and Managi 2017; Buff a et al. 2018), while research on the area surrounding Changbai Mountain also showed that habitat quality was signif icantly aff ected by landscape fragmentation and decreased connectivity (Dai et al. 2019).

In addition to the impacts on the landscape, ginseng planting was also prone to soil loss, which occurred in three stages of the process of planting ginseng (f ield preparation,growth production, and the fallow stage). Ginseng planting provided important support for local industrial development and incomes but it also had a negative impact on forest resources. Therefore, it is necessary to comprehensively analyze the benef its and impacts of ginseng planting (Baral et al. 2019; Nguyen et al. 2019) from a green economy perspective or by considering social economic and index decisions for reference.

From the perspective of ecosystem services (Wang et al.2017), deforestation for ginseng planting was inappropriate. However, considering the demand for ginseng and its cultural value as a traditional product, ginseng planting will endure for a long time. Therefore, some countermeasures should be considered:

(1) In order to reduce the impact of ginseng planting on forest landscapes, the administrative department should scientif ically plan and control the number, area and distribution of ginseng f ields at an appropriate level.

(2) Restoration of abandoned ginseng fields to forests should be strengthened to reduce their fragmentation.In addition to forest tending measures, soil restoration and erosion control should be implemented.

(3) Future ginseng plantings on deforested lands should be replaced by other cultivation modes. Large-scale ginseng planting should give priority to using croplands for cultivation, and high-quality ginseng could be planted under forest canopies.

(4) Increasing stakeholder awareness can lead to suitable management solutions (P?tru-Stupariu et al. 2020).Therefore, it is necessary to strengthen the public prof ile of ginseng farmers, which might encourage them to take measures to limit the impact of existing ginseng plantings, to strengthen the restoration of abandoned ginseng f ields and to change the nature of ginseng planting in the future.

Conclusions

The average altitude and slope of ginseng f ields from 1989 to 2014 showed an upward trend, while the average area decreased sharply. This was mainly due to the continuous decrease in forest resources suitable for planting ginseng in low altitude areas. As a result, the increasingly scattered ginseng f ields became closer to the Reserve. The increase area of construction lands and forest area corresponded to the decrease of croplands, because the decrease in croplands had buff ered the contrast between urbanization and aff orestation. Because croplands had a considerable proportion of patch numbers and total area, the impact of croplands and its change was the major driving force of regional landscape changes. Ginseng f ields and abandoned ginseng f ields increased total patches and total area of the local forest landscapes, and shared edge lengths between ginseng f ields and forests, resulting in continuous fragmentation of the local landscape. Although the total area of existing and abandoned ginseng f ields accounts for a small fraction of the total landscape, their negative impact on regional ecosystem conservation cannot be underestimated. In order to help maintain the healthy and stable development of the local ginseng industry and improve the quality of regional ecological environment, Ginseng planting should be planned rationally,and it is necessary to restore abandoned lands scientif ically and enhance the environmental protection awareness of ginseng farmers.

This research examined the impact of ginseng planting on the forest landscape through spatial analysis and logistic regression analysis, but did not look at the eff ect of ginseng planting on the physical and chemical properties of the soil and how further aff ect the forest landscape pattern. Therefore, future research needs to determine soil physical and chemical properties to comprehensively record the impact of ginseng planting on the forest landscape so as to guide ginseng planting more scientif ically and accurately.

Conf licts of interest

The authors declare that there is no conf lict of interest.

Author contributions Conceptualization, TY and MT; investigation, SQ, LL, SL and TY; formal analysis, TL; writing, MT, TY and SQ. All authors have read and agreed to the published version of the manuscript.