Introduction

Biodiversity loss has become a globally critical problem (IPCC 2022). The value of agricultural lands for conserving biodiversity has been recognized formally in well-designed agri-environment schemes (e.g., Kleijn et al. 2006; Davey et al. 2012). However, land-use intensification in agri-environments is also a key driver of biodiversity changes. For example, intensified agriculture is characterized by high fertilizer input and higher production output, and keeping the heterogeneous landscape mixed with forest may facilitate the expansion of generalist birds and demote ecological complexity (Olden and Rooney 2006; Santana et al. 2017). As a result, an increase in species richness might be driven by nonforest-originated or non-native species in an agri-environment (Dallimer et al. 2012; de Lima et al. 2013; Soares et al. 2020). In Japan, the traditional agricultural landscape called “satoyama” successfully maintains high biodiversity in mountain villages and has been widely promoted by ecologists (Ohara and Ushimaru 2015). These forest-agriculture mosaic landscapes, with small cultivated areas and low isolation from surrounding forests, were considered a sustainable basis for maintaining biodiversity (Benton et al. 2003; Katoh et al. 2009). However, although the species richness of bird assemblage may increase with the degree of forest conversion to agricultural land, the richness change is mainly attributed to bird species common in disturbed areas (Koh et al. 2018).

Beta diversity is a diversity metric used to quantify the variation in species composition between different communities or “the degree of community differentiation or variation in community composition related to environmental gradients or patterns” (Shmida and Wilson 1985; Rosenzweig 1995; Si et al. 2017). Nowadays, beta diversity receives increasing attention in regional-scale conservation, and it has been considered a critical factor in examining the success of biological conservation plans (Socolar et al. 2016). Therefore, examining the changes in beta diversity at a regional scale can help us better understand the impacts of land-use change on biodiversity in mountain villages situated in forest-agriculture mosaics. The organism groups for studying beta diversity partitioning mainly focus on terrestrial animals and plants (Si et al. 2017). Among them, birds are a primary research target in vertebrate groups. Furthermore, it was found that species of distinct functional groups will respond differently to various disturbance intensities and habitat preferences (Neuschulz et al. 2011; Morante-Filho et al. 2015). Therefore, it is expected to show the species composition change, separately exploring the response of bird functional groups with different habitat preferences, such as forest birds and nonforest birds, to forest conversion in the forest-agriculture mosaic landscape.

Baselga (2010) proposed partitioning Sørensen beta diversity (βsor) into turnover beta diversity (βsim) and nestedness-resultant beta diversity (βsne). Moreover, recent research on the role of land-use in the beta diversity of species has shifted from qualitative comparisons (ANOVA-like inference approaches; Newbold et al. 2013) to quantitative studies of land-use gradients (Sreekar et al. 2020; Soares et al. 2020; Wayman et al. 2021), primarily due to the development of improved data analytical methods. Therefore, the study used beta diversities with different components to explore how the forest cover gradient changes bird species composition in the forest-agriculture landscape.

An increase in the nestedness of bird species is a primary indicator of homogenization (García-Navas et al. 2020). Few studies have examined the potential homogenization effect (increased nestedness species) caused by variations in forest cover resulting from intensive agriculture in mountain villages. This lack of information has resulted in the oversight and misinterpretation of the influence of environmental changes on bird diversity. This study examined the impact of forest cover changes on the beta diversity of forest and nonforest birds. In the forest-agriculture landscape in coastal mountain villages, we were particularly interested in determining if: (1) forest birds exhibited a high turnover component of beta diversity, whereas nonforest birds exhibited nestedness-resultant distribution characteristics, and (2) conversion of forests into agriculture drove species composition toward homogenization in mountain villages.

Materials and methods

Study area and forest cover levels

The study area was located along the coastal mountains of northern Taiwan (East Sea-Datun Volcano Group; 1120 m above sea level) amid three foothill villages located 5–8 km from the coast, namely Xinghua in Sanzhi (SZ), Lianghu in Jinshan (GS), and Shuangxing in Wanli (WL), each with an average slope of 24% (Fig. 1). Using aerial images from a drone, we selected sample sites with a radius of 100 m with different levels of forest cover: high (> 70%, H), medium (25–70%, M), and low (< 25%, L). Nine observation points corresponding to these three levels of forest cover were set up in each village, resulting in 27 observation points across all three villages. Table S1 presents the forest cover ratio within a radius of 100 m around each of the 27 observation points. There are two main types of trees: the “Pyrenaria-Machilus subtropical winter monsoon forest” and “Phoebe-Machilus submontane evergreen broad-leaved forest” (Li et al. 2013). The forest type ranged from primary vegetation to mature secondary vegetation (following the definition of Hudson et al. (2014): forests in the tropics are usually at least 30 years old), and the study area is located in a well-protected national park (established in 1985) near the outer boundary, the forest area has remained undisturbed for a long time.

Fig. 1
figure 1

Landscape and land-use pattern of the study sites. a Three villages are located at the foot of Datun Mountain (1120 m) in northern Taiwan coast. b–d Sample sites and forest cover patterns (the grey shaded areas) in the three villages on the same scale. The hollow black circles represent a 100 m radius around each site. The contour lines represent elevation (m) above sea level

Full size image

Bird survey

Surveys were conducted bimonthly between March 2019 and January 2021, resulting in 12 surveys for each of the 9 sample sites in the three mountain villages. All surveys were conducted on no rainy days and within 3 h of sunrise to record the species and number of birds seen or heard at each sample site. Each sample site's survey time was 10 min before the subsequent site was examined. Birds were detected by the fixed-radius (50 m) point count method (Morante-Filho et al. 2015), and the plot size was 0.785 ha. Each survey required approximately 150 min, including the transit time. To minimize observer bias, all surveys were conducted by the same individual within the same time frame. To reduce systematic bias, the sequence of sample sites was randomly altered each time.

Beta diversity

Sørensen beta diversity index partitioned through the beta.pair function (García-Navas et al. 2020; Si et al. 2015; Sreekar et al. 2020; Wayman et al. 2021) in the R (ver. 4.2.0) package “betapart” (ver. 1.5.6) (Baselga and Orme 2012), to compute pairwise beta diversity (e.g., βsor, βsim, βsne). According to Baselga (2010), Sørensen beta diversity between pairs of sites can be decomposed into turnover (species replacement between sites) and nestedness-resultant (incomplete species replacement between sites) components. The following the equations used to calculate βsor, βsim, and βsne:

$$\beta_{{{\text{sor}}}} = \frac{{\text{b + c}}}{{\text{2a + b + c}}} = \beta_{{{\text{sim}}}} + \beta_{{{\text{sne}}}}$$
(1)

(Baselga 2010)

$$\beta_{{{\text{sim}}}} = \frac{{\text{b}}}{{{\text{a}} + {\text{b}}}}$$
(2)

(Baselga 2010)

$$\beta_{{{\text{sne}}}} = \beta_{{{\text{sor}}}} - \beta_{{{\text{sim}}}} = \frac{{{\text{c}} - {\text{b}}}}{{2{\text{a}} + {\text{b}} + {\text{c}}}} \times \frac{{\text{a}}}{{{\text{a}} + {\text{b}}}}$$
(3)

(Baselga 2010).

where a was the number of shared species between two sites, and b and c were the numbers of unique species in each community, with b being less than c. Wilcoxon signed-rank test was used to determine the pair difference in beta diversity between forest and nonforest bird species within identical sites through the wilcox.test function of the “stats” package of R (ver.4.2.0). A multiple-sample Kruskal–Wallis test was conducted to confirm whether bird beta diversity differed across forest cover levels.

Effect of forest cover difference on beta diversity

Beta diversity is affected by environmental gradients, which can be quantified by community composition (beta diversity) vary with environmental conditions, the gradient such as the area ratio difference of forest cover (∆x) between pairs of sites (Baselga et al. 2015; Si et al. 2015; Gómez-Rodríguez and Baselga 2018; Sreekar et al. 2020; Wayman et al. 2021; Martínez-Santalla et al. 2022). In this study, this calculation was performed using the beta.pair and decay.model functions of the “betapart” package (v.1.5.6) of R, with the Pseudo R2 of a nonlinear model and p-value used to indicate the degree of fit (Martínez-Santalla et al. 2022). The formula of exponential regression: y = 1 − a × exp(− b × x). Here, a is presented as intercept, and b means slope. The p-value of the model was estimated from a permutation test, given as a symbol: 0 < ,‘***’  < 0.001 < ,‘**’ < 0.01 < ,‘*’ < 0.05 < ,‘. ’ 0.1 < ,‘ ’.

Results

Across all 27 sample sites (Table S1), a total of 87 species, including 36 families and 65 genera, were recorded (Table S2). Bird functional groups were used to categorize bird species into forest (F) and nonforest (non-F) bird communities. Table 1 presents the number of species of these two bird groups across areas with high, medium, and low forest cover. 37 forest bird species were identified, with 32, 30, and 22 species found in areas with high, medium, and low forest cover, respectively. Similarly, 50 nonforest bird species were identified, with 15, 25, and 48 species found in areas with high, medium, and low forest cover, respectively. These results indicated these areas attracted species not originally found in forests; thus, low forest cover altered the composition of birds in low-altitude mountain habitats.

Table 1 Species richness among habitats and bird functional groups
Full size table

Beta diversity indices of forest and nonforest birds

Beta diversity of different bird functional groups is depicted in Fig. 2. A beanplot form is used to represent each paired site,s βsor, βsim, and βsne values and the turnover contribution ratio (βsim/βsor) and nestedness-resultant (βsne/βsor) contribution ratio. According to the results of the Wilcoxon signed-rank test of the beta diversity of paired sites (labels “a” and “b” in Fig. 2), the forest and nonforest bird functional groups exhibited statistically significant differences in βsim. The median of the nonforest bird species (βsim = 0.33, Fig. 2b) was significantly higher than that of the forest bird species (βsim = 0.21). In addition, the forest bird species exhibited significantly lower nestedness-resultant diversity (βsne = 0.10, Fig. 2c) between sites than the nonforest bird species (βsne = 0.25). In terms of the ratio of βsne to βsor (βsne/βsor), the forest bird species exhibited a lower contribution to Sørensen diversity (βsne/βsor = 0.30, Fig. 2e) in the 27 sites than the nonforest bird species (βsne/βsor = 0.41), indicating that a subset of shared species (nestedness species), to some extent, enhanced species diversity for the nonforest bird species.

Fig. 2
figure 2

Beta diversity among forest (F) and nonforest (non-F) species. a Sørensen beta diversity (βsor), b turnover component of Sørensen beta diversity (βsim), c nestedness-resultant component of Sørensen beta diversity (βsne), d turnover contribution ratio (βsim/βsor), and e nestedness-resultant contribution ratio (βsne/βsor). The bar in the beanplot represents the median of paired beta diversity. The letters “a” and “b” beside the median values indicate groups defined by a paired Wilcoxon signed-rank test

Full size image

Beta diversity indices of forest and nonforest birds across different proportions of forest cover

Sørensen diversity (βsor), turnover ratio (βsim/βsor), and nestedness-resultant ratio (βsne/βsor) of different forest cover levels in a beanplot form through a Kruskal–Wallis test, with grouping labels a, b, and c (Fig. 3). For all bird species, the βsor medians of forest cover with high (H), medium (M), and low (L) forest cover were 0.33, 0.37, and 0.38, respectively (Table 2). In Fig. 3a, the Kruskal–Wallis test groupings were a, ab, and bc, indicating that the βsor values in forest cover H and M belonged to the same group (p > 0.05). By contrast, βsor values in the H and L belonged to different groups (p < 0.05). By extension, the βsor value of forest birds belonged to the same group, with no apparent difference between the three forest cover levels (with median values ranging from 0.27 to 0.30). Nevertheless, the composition of bird species differed (as discussed later). For nonforest birds within the same forest cover level, the value of βsor significantly decreased (0.75, 0.55, and 0.39) with decreasing forest cover levels from H to L.

Fig. 3
figure 3

Partitioning of beta diversity components across habitat classes for all birds, forest birds, and non-forest birds. (a-c) Sørensen beta diversity (βsor), (d-f) turnover contribution ratio (βsim/βsor), and (g-i) nestedness-resultant contribution ratio (βsne/βsor) of different habitat classes (high, medium, and low) for the entire birds and the two bird groups. The three subplots from left to right represent all, forest and nonforest species. The bar in the beanplot represents the median of paired beta diversity. The letters “a,” “b,” and “c” beside the median values indicate groups defined by a Kruskal–Wallis test. The presence of the same character indicates the absence of a significant difference between habitat classes

Full size image
Table 2 Summary of the beta diversity, turnover, and nestedness-resultant components of breeding birds in the 27 study sites
Full size table

In terms of the turnover ratio or turnover contribution ratio (βsim/βsor, Fig. 3d–f), the turnover ratio of forest birds in sites with medium forest cover (0.86, Table 2) was significantly higher than that in areas with high and low forest cover (0.62 and 0.70, respectively). On the contrary, the nestedness-resultant contribution ratio (βsne/βsor, Fig. 3g–i) of nonforest birds in sites with low forest cover was significantly higher (0.28) than that in sites with high and medium forest cover (0.17 and 0.16, respectively). These results indicated that low forest cover significantly increased the proportion of shared (common) nonforest birds between sites. In summary, for nonforest birds in low forest cover (< 25%), the βsor diversity decreased, the number of different bird species (βsim) decreased, and the number of shared (common) bird species (βsne) increased (Fig. 3c, f, i).

In contrast to Fig. 3, which depicted differences in beta diversity inter forest cover levels, Fig. 4 illustrated changes in paired beta diversity across all sites due to variations in forest cover. Figure 4a–c showed the βsor diversity, and Fig. 4d–f depicted the turnover ratio (βsim/βsor). The x-axis represented the absolute difference in forest cover (%) between paired sites, and the y-axis represented the beta diversity of paired sites. A total of 351 paired data points were observed [27 × (27 − 1)/2]. Although the overall βsor diversity increased with the difference in the forest area (b = 0.0060, Pseudo R2 = 0.48, p < 0.01), the βsor of forest birds (b = 0.0024, Ppseudo R2 = 0.19, p < 0.01) and nonforest birds (b = 0.0103, Ppseudo R2 = 0.35, p < 0.01) varied to different extents. Thus, more absolute differences in the forest area between sites (either an increase or a decrease) brought higher βsor of nonforest birds than forest birds. Furthermore, the mainly increased βsor was contributed by nestedness-resultant (shared) birds of nonforest birds (Fig. 4i). In other words, nonforest birds had fewer attributes of occasional occurrences (Fig. 2d). It implied that nonforest birds had a stronger tendency in sites with few species to be a subset of the rich species in sites with species originating. For example, across the 351 paired sites, only 6% of the forest bird species had a βsim value of 0, whereas 14% of the nonforest bird species was a subset of species from another site (b = 0, βsim = 0, in Eq. 2). Hence, the species βsor diversity between the two sites was entirely contributed by βsne.

Fig. 4
figure 4

Effect of forest cover difference on beta diversity for all birds, forest birds, and nonforest birds. The three subplots from top to bottom represent a–c paired Sørensen beta diversity, d–f turnover contribution ratio (βsim/βsor), and g–i nestedness-resultant contribution ratio (βsne/βsor). The x-axis represents the absolute forest cover (%) difference between pair sites. The line represents the exponential fitted curve. The value of “b” indicates the slope coefficient of exponential regression

Full size image

Discussion

This study was conducted in the low-altitude mountainous regions of northern Taiwan; the number of bird species observed (Table 1) revealed a decrease in forest bird species richness and an increase in nonforest bird species and net total species richness with increased human-mediated deforestation. Although the proportion of forests has decreased in mountainous villages, farming has created various habitat types due to different farming methods (Fu et al. 2022). In our study, agriculture (cropland, paddy, orchard) attracted nonforest species, including widespread species and species in other habitat types (i.e., waterbird), replacing lost forest species and increasing overall species richness. This attracted species with different habitat requirements beyond maintenance and increased species richness, which differed from the previous results (Supp and Ernest 2014; Morante-Filho et al. 2015).

However, there were specific forest-dwelling birds disappeared as forest cover declined. Environmental filtering was expected to influence patterns of bird assemblages in this case. The deforestation in a few sites suffered by this landscape resulted in lower forest cover. Thus, this promotes the selection of birds with high vagility, larger population sizes, and low habitat specificity (Tarifa et al. 2021; Si et al. 2016; Morante-Filho et al. 2016). Supporting this idea, the less forested landscape showed a higher percentage of nonforest birds, such as species from the Columbidae and Sturnidae families. These findings indicated that human activities pose threats to certain forest-dwelling bird species. For instance, a forest bird species known as the Common Emerald Dove (Chalcophaps indica) was detected in four sites with high forest cover, two sites with medium forest cover, and found in no site with low forest cover (Table S2), highlighting the phenomenon of restrictive living environments for the forest species. Previous research has thus recommended introducing species-centric conservation, emphasizing that preserving 30% of the total land area as conservation zones was essential for protecting waterbirds (Wauchope et al. 2022), 50% for forest species mixed waterbird-forest birds communities (Fu et al. 2022) and 70% for forest bird species (Benton et al. 2003). Rapid decreases in vulnerable species were expected in the absence of such measures. Without a species-centric management approach, it may be erroneously concluded that increased human activities (and concurrent forest loss) lead to a net increase in overall species richness and Sørensen beta diversity, potentially underestimating conservation efforts needed for uniquely adapted species (Table 1).

The results of this study indicated that nonforest birds exhibited pronounced nestedness. In contrast, forest birds did not (Fig. 2), showing differences in the ecological niches and dispersal abilities of the two bird functional groups. Some generalists were attracted by nonforest areas (Newbold et al. 2013; Santana et al. 2017), indicating that these habitats might offer abundant food sources, attracting bird species with high reproductive rates and strong colonization abilities (high βsne proportion). These species may contribute to the homogenization of bird species in forest-agriculture villages, thus serving as a primary source of nestedness components. For example, among nonforest birds, some species from the Hirundinidae and Motacillidae families with superior dispersal abilities raise the nestedness-resultant component of beta diversity of nonforest birds.

By contrast, our forest bird species exhibited low beta diversity (βsor), including low turnover (βsim) and low nestedness-resultant (βsne) beta diversity (Fig. 2a–c). Forest birds predominantly inhabit areas with higher forest cover, indicating their strong forest preference. Most of them do not utilize a variety of ecological niches (García-Navas et al. 2020; Sreekar et al. 2020). For instance, in this study, Taiwan Patridge (Arborophila crudigularis) and Varied tit (Sittiparus castaneoventris) were primarily detected in areas with more than 85% forest cover. Their preference for forests might contribute to a low nestedness-resultant beta diversity. The mountain village, in this case, has the characteristics of high forest connectivity and low to moderate development area (Koh et al. 2018), and the forest vegetation is mainly secondary broad-leaved forest composed of low-altitude Machilus and Ficus plants (Fu et al. 2022). In this typical type of mountain village on the northern Taiwan coast, although the forest has experienced varying degrees of development, the composition of forest bird species has lower turnover beta diversity than nonforest bird species. The lower replacement in the forest species composition is speculated to be related to the forest characteristics of Taiwan,s low-altitude mountain mentioned above. Agriculture changes the number and composition of bird species with increasing species richness (García-Navas et al. 2022). Our results indicated that the bird community composition changed with variations in forest cover in the forest-agriculture mosaic mountain villages. Each of the two bird groups responded differently to the environmental change. From the differences between forest and nonforest birds, we discovered that the changes observed in the bird community structure in mountain villages were primarily driven by differences within the nonforest birds (Fig. 4c). Incorporating bird functional groups in a specific region highlights the merit of forest-agriculture systems, which feature moderate species turnover and the coexistence of both forest and agricultural species (García-Navas et al. 2022; Si et al. 2016). Our findings suggest that disregarding different bird functional groups may result in the underestimation of the homogenization of species composition. Therefore, discussing the beta diversity of species turnover and species nestedness separately for different bird functional groups is essential when addressing the effect of habitat degradation on biodiversity.

The overall richness of nonforest bird species increased when the nonforest area increased. However, this increase reduced the diversity of habitat-specific species, thus increasing the proportion of shared (nestedness) species, representing interspecies homogenization (García-Navas et al. 2020). For instance, several sites within low forest cover (type L) that comprised nonforest species exhibited identical species (βsim = 0). Many studies have indicated that intensive agriculture promotes biological homogenization because only a subset of local species can tolerate the conditions (Dormann et al. 2007; Ekroos et al. 2010; Karp et al. 2012; Tarifa et al. 2021). Hence, this process facilitates selecting groups with specific functional attributes (“winning” species; Santana et al. 2017).

Land-use patterns in habitats play a pivotal role in influencing the beta transition of regional birds. Previous research has primarily focused on estimating the effects of agriculture, such as the area ratio, distance to agricultural lands, and paired comparisons with protected areas (Clough et al. 2009; Sreekar et al. 2020; Soares et al. 2020). In this study, different land-use types, including shrubs, dry lands, paddy fields, and suburban areas, were observed in the study region. Thus, excessive variables may compromise model robustness due to the reduced degrees of freedom. Using regression models, Wayman et al. (2021) demonstrated that forest area is a crucial parameter for the beta diversity metrics of birds (βsor, βsim, and βsne) among the different types of land-use. Our research on the effect of interhabitat forest cover differences on diversity indices revealed nonlinear fits of forest and nonforest bird βsor, with Pseudo R2 values of 0.19** and 0.35**, respectively (Fig. 4). In terms of nestedness ratios (βsne/βsor), Pseudo R2 values of 0.02** and 0.35** were observed. These results indicate that the nestedness of forest bird species, in response to forest cover changes, was more stochastic than nonforest bird species.

This study obtained a different angle to consider homogenization. After partitioning the Sørensen beta diversity, the increased nestedness component was our key basis. On this basis, a nestedness ratio (βsne/βsor) versus environment difference matrix (Fig. 4) conducted as a novel metric help us to check the level of the homogenization comparison in species functional groups, which may help cross-region or temporal comparison in the future. Here, the high dependence of nestedness ratio versus forest difference implies a high tendency of homogenization after forest decline. In this study, the nestedness attributes of nonforest bird species were more dependent on forest cover changes, indicating a substantial increase in nestedness species among nonforest birds as cumulative forest decline. Overall, our study highlights forest covers as an efficient proxy monitor of bird community diversity, guiding researchers and policymakers toward considering the consequences of their rural developmental strategies.

In developmental strategies in rural mountain villages, someone would focus on the patterns of land-use change with forest exploitation, leading to an overarching trend of increasing bird beta diversity. It may reflect the likelihood that the mosaic landscape of forests and agriculture in mountain villages drives species composition toward homogenization. Although changes in land-use patterns may provide insights into biodiversity shifts, our findings show that a higher beta diversity does not necessarily guaranty a higher conservation value (Socolar et al. 2016). Instead, based on a misleading mix of bird grouping (i.e., no grouping, all bird), higher Sørensen beta diversity hides that the embedded mosaic landscapes with forests and agriculture of mountain village may intensify species homogenization, which is regarded as a critical problem in species loss. Taking bird communities as an example, we recommend implementing targeted conservation measures to enhance the habitat-specific species biodiversity of low-altitude mountain villages in Taiwan. This approach focuses on vulnerable species, such as forest birds, and their habitats while considering sustainable socioeconomic development.

Conclusions

In the low-altitude mountain villages of northern Taiwan, the diversity of bird communities is predominantly characterized by species turnover. This phenomenon reflects the mosaic landscape of the ecological environment of this region, where diverse habitats, including forests and agricultural lands, are intertwined, creating a biodiversity-rich region. A rapid increase in nonforest bird richness with a decrease in forest bird species increased net total species richness after human-mediated forest cover declined. Nonforest birds exhibited pronounced nestedness that included several generalist birds with high tolerance to habitat variations, serving as a primary source of nestedness birds. Therefore, as forest decline, nestedness-resultant components in Sørensen beta diversity increase, raising nonforest bird species, homogenization in forest-agriculture villages (Fig. 4). By contrast, forest birds exhibited low turnover-resultant components in Sørensen beta diversity, which is speculated to be related to mountain village forest characteristics such as high forest connectivity, low to medium development area, and similar forest vegetation. This study quantifies the bird communities change along forest cover differences on a regional scale. The expansion of agriculture results in a filtering towards more similar nonforest bird communities in low forest cover landscapes. These findings highlight the distinct effects of environmental changes on these two bird functional groups. Thus, conservation plans for the forest-agriculture mosaic landscape should separately address forest and nonforest birds to avoid the risk of underestimating species homogenization due to anthropogenic land-use but the rise in species richness. Here, forest cover is an effective proxy indicator of bird community diversity. The study findings effectively illustrate the varying composition of forest and nonforest bird species due to habitat changes in low-altitude mountain village landscapes composed of mosaic forests and agricultural lands, thus providing valuable guidance for policymakers. Understanding these dynamics may aid in formulating more effective and targeted conservation strategies to preserve biodiversity in mountain village areas while balancing the developmental needs of the cultivation land.