Bertille Mohring, François Brischoux, Frédéric Angelier. 2021: Vineyards, but not cities, are associated with lower presence of a generalist bird, the Common Blackbird (Turdus merula), in Western France. Avian Research, 12(1): 3. DOI: 10.1186/s40657-020-00239-0
Citation:
Bertille Mohring, François Brischoux, Frédéric Angelier. 2021: Vineyards, but not cities, are associated with lower presence of a generalist bird, the Common Blackbird (Turdus merula), in Western France. Avian Research, 12(1): 3. DOI: 10.1186/s40657-020-00239-0
Bertille Mohring, François Brischoux, Frédéric Angelier. 2021: Vineyards, but not cities, are associated with lower presence of a generalist bird, the Common Blackbird (Turdus merula), in Western France. Avian Research, 12(1): 3. DOI: 10.1186/s40657-020-00239-0
Citation:
Bertille Mohring, François Brischoux, Frédéric Angelier. 2021: Vineyards, but not cities, are associated with lower presence of a generalist bird, the Common Blackbird (Turdus merula), in Western France. Avian Research, 12(1): 3. DOI: 10.1186/s40657-020-00239-0
Land-use change is one of the main drivers of the global erosion of biodiversity. In that context, it is crucial to understand how landscape characteristics drive the presence of rare endangered species. Nevertheless, it is also important to study common species in multiple habitats, because they represent a large proportion of biodiversity and are essential to maintain ecological functions. Interestingly, some habitats, as farmlands with permanent crops (e.g. vineyards), have been overlooked in the literature.
Methods
In this study, we investigated the distribution of a widespread and common bird species, the Common Blackbird (Turdus merula), within and between the three main habitats of our study area (rural Western France). We specifically focused on (1) woodlands, (2) farmlands with a high vineyard coverage, and (3) moderately urbanized areas. Specifically, we aimed to assess the beneficial and detrimental effects of these habitats and their fine-scale composition on the presence of a common bird species, relying on a survey by point counts (nearly 100 locations). We studied the effects of habitats and gradients of fine-scale habitat composition on blackbird presence using logistic regression analyses.
Results
Blackbirds were present in all studied habitats. However, their presence varied between habitats, being lower in vineyards than in woodlands and cities. In woodlands and cities, fine-scale analyses did not reveal any component driving the species' presence. However, we found that shrub and tree vegetation cover had a significant positive effect on blackbird presence in vineyards.
Conclusions
Our results are in agreement with the definition of a generalist species. Interestingly, species distribution varied between habitats. The high presence of blackbirds in urban areas suggests that medium-sized cities, despite their artificialization, do not constrain the settlement of this former forest specialist and that green spaces may allow blackbirds to thrive in medium-sized cities. On the contrary, we found an impoverished presence of blackbirds in vineyards and a positive effect of vegetation on their presence in these landscapes. This suggests that permanent crops, and more generally farmlands, may impose important constraints to common species. Future studies should examine how to enhance biodiversity through agricultural management policies, especially in vineyards.
Agricultural intensification is a principal driver of rapid global biodiversity loss, dramatically changed landscape structures and converting diverse, traditional landscapes into simplified agricultural ecosystems. This shift has significant implications for the functionality and stability of ecosystems (Green et al., 2005; Tscharntke et al., 2005). Agricultural land covers more than 47 million km2 over the world, which represents the single greatest land use (Foley et al., 2011). Hence, it is necessary to understand how to fully maximize biodiversity underlying the context of agriculture development. A key step in this process is understanding how various characteristics of agricultural systems influence biodiversity.
The extent to which an agricultural system negatively impacts biodiversity depends on how it alters the landscape (Wilson et al., 2017). This influence is related to the landscape structure of the agricultural system such as the varieties of crops grown, the patch size of croplands, the intensity of techniques used to grow crops, and the configuration of natural land covers (Wilson et al., 2017). In contrast to natural ecosystem, agricultural systems can vary widely, ranging from monocultures with reduced biodiversity or systems designed to support wildlife by preserving significant portions of semi-natural environments (Donald et al., 2001; Tscharntke et al., 2012; Liu et al., 2013).
Yet numerous studies have demonstrated the positive effects of semi-natural habitats (e.g. farmland hedges, buffer strips, field margins, scrubland, waterbody, and traditional orchards) on biodiversity (Kremen and Miles, 2012; Kremen, 2020). They support various ecological communities due to increasing landscape connectivity and heterogeneity or providing special resources for species in the fragmented agricultural landscape (García-Feced et al., 2015; Wilson et al., 2017; Marcacci et al., 2020). Supporting and restoring these habitats could effectively mitigate the negative impacts of agricultural systems on biodiversity. However, few studies have successfully disentangled the variation in the supporting effect of agriculture landscape structure on biodiversity under different agricultural intensities. The primary reason is that the loss of natural habitat and the agricultural intensification often occur simultaneously (Chiron et al., 2014). For example, landscapes with higher agricultural intensity always have fewer semi-natural habitats on a local scale.
Increasing the environmental heterogeneity of agricultural systems is recognized as one of the main conservation tools to mitigate biodiversity loss in agricultural landscapes (Benton et al., 2003; Smith et al., 2010). Traditionally, the assessment of environmental heterogeneity in agricultural landscapes has been based on the diversity of non-crop vegetation types and other non-crop elements (Lee et al., 2024). In recent years, some studies have regarded crop diversity as one of the indicators of agricultural ecosystem heterogeneity and have demonstrated its positive role in supporting biodiversity (Lee and Goodale, 2018; Lee et al., 2024), though the specific effects depend on factors such as the scale of study and the target taxa, among other factors (e.g., Henderson et al., 2009; Lindsay et al., 2013; Fahrig et al., 2015).
The framework offered by the central and eastern Jilin Province of China with its strong gradient of agricultural intensity is an ideal model system to address this topic. In the eastern direction from the central part of Jilin Province, there is a transformation from large expanses of continuous farmland to small agricultural patches that are enveloped by natural habitats (Fig. 1), while a similar proportion of semi-natural habitats was retained across varying degrees of land use intensity at the local scale due to the land management strategy. In recent years, the Chinese government has continuously increased its emphasis on ecological health, cropland is gradually shrinking all over the country due to the project called "Grain for Green" (Xu et al., 2006). Identifying the impacts of agriculture landscape structures including semi-natural habitats of agricultural systems on biodiversity can facilitate planning in the process.
Figure
1.
The study area is situated within the central and eastern Jilin Province of China. The bottom right inset is located in Jilin Province in Northeast China. (A): high-intensity agriculture; (B): middle-intensity agriculture; (C) low-intensity agriculture. Red dots show the distribution of bird sampling points. (D, E, F) show part of farmland with different semi-natural habitat configurations, and each sampling site (red cross) is in the center of imagine.
Birds are among the most studied and ecologically best-known vertebrates, having experienced decline in distribution worldwide over decades due to agricultural change (Gregory et al., 2005; Gregory et al., 2007). The effects of differences in agricultural landscape patterns on avian diversity and the metacommunity dynamics of their communities remain a subject of debate (Vos et al., 2001). Agriculture systems can transform the natural vegetation structure, change the availability of food resources, affect the nesting and reproductive success of birds, and ultimately affect population dynamics (Butler et al., 2007; Elsen et al., 2018; Douglas et al., 2023). Previous studies found that different groups of birds (e.g. ground and shrub breeders) had inconsistent responses to agricultural landscape structure (Beecher et al., 2002; Yahya et al., 2022), emphasizing the importance of investigating various bird groups to better understand the influence of agricultural landscapes on bird communities.
The aim of this study was to disentangle the relative effects of semi-natural habitat on bird species richness and abundance in different agricultural intensity landscapes (low-middle-high) in central and eastern Jinlin Province, China, to get insights for data-driven management strategies. To better compare the variance of the support role of agriculture landscape configurations under different agricultural intensities, we grouped birds according to seasonal status and feeding guilds. Here we try to find long-term sustainable ways to ensure food security without compromising the ecological roles and benefits, while preserving rich biodiversity within an agricultural ecosystem.
2.
Methods
2.1
Study area
Our study was conducted along the direction from Changchun City to Changbai Mountain of eastern Jilin Province (41°58′15″–43°47′37″ N, 125°20′46″–127°23′15″ E; Fig. 1). In this direction, there was a transition from continuous farmland to scattered farmland fragments surrounded by natural habitats, with agricultural practices shifting from machine-based to hand and draft animals-based methods. We selected three regions along this direction to represent low-, middle-, and high-intensity agricultural landscapes (Fig. 1A, B and C). Considering the focus of our study, the three agricultural intensity regions primarily reflect variations in landscape land-use patterns, although agricultural practices (e.g., crop density, fertilizer) are also recognized as relevant components of agricultural intensity (Reif and Hanzelka, 2020). These regions were characterized by small farmland patches surrounded by forests, a mosaic of forest and farmland patches, and large continuous farmland, respectively (Table 1). The temporal pattern of agricultural intensity across the study area was consistent, with cultivation occurring from April to October each year. From November to April, farming activities ceased due to cold weather. Agricultural activities primarily concentrated on the cultivation of maize, paddy, sunflower, and celery cabbage, with a minor share dedicated to legumes and root vegetables. In addition to crops, the agricultural lands included scattered trees and forest patches, grassy patches, wetlands, shrublands, and irrigation canals. The region falls within the temperate continental monsoon climate zone. Elevations ranged from 240 m to about 960 m a.s.l. The annual precipitation and average temperature are approximately 680 mm and 3.8 ℃, respectively.
Table
1.
Explanatory variables in low, middle, and high-intensity agriculture.
The study design consisted of 322 sampling sites (103 in high-intensity agriculture, 118 in middle-intensity agriculture, and 101 in low-intensity agriculture) distributed between three study sites with landscape agricultural intensity from high to low (Fig. 1). The average distance between study sites was 65 km with a minimum of 1 km. We selected the study sites to disentangle the relative effects of the landscape agricultural intensity (high vs. middle vs. low) on the role of semi-natural structures in supporting avian communities. Notably, despite the potential correlation between landscape agricultural intensity and semi-natural vegetation cover (where farms with intensive management often lack natural structures), no significant difference in semi-natural vegetation cover was observed among the three study regions (Table 1). This allowed for a more focused assessment of the relative contributions of landscape agricultural intensity and semi-natural structures to avian community support.
2.3
Bird surveys
We conducted bird repeat-visit point counts in May and June 2023 (Hutto et al., 1986), during the peak of the bird breeding season. Two experienced observers conducted two visits to each sampling site, recording all birds within a 10-min period and a 50-m radius, excluding birds that were flying above the canopy. We utilized the highest value obtained from two counts to represent the relative abundance of each bird species (Toms et al., 2006). We used Sony ICD-PX470 microphones to record unidentified vocalizations which allowed us to conduct subsequent identification using online reference material. All surveys were conducted within the first 4 h after sunrise under clear days without strong wind and rain, ensuring optimal conditions for bird detection. We combined the bird data from both visits into a single dataset, retaining the highest recorded count for abundance from either of the two visits. Subsequently, we attributed functional traits (i.e. seasonal status and feeding guilds) to each bird species according to Zheng (2011), Zhao et al. (2018), and our observations in agricultural lands of Jilin Province of China. Seasonal status was classified as resident or migratory and feeding guilds were assigned to insectivorous (mainly eating insects, especially in breeding season such as chickadees), omnivores (a variety of diets such as magpies), granivores (primarily eating seeds with few insects such as turtledoves) and carnivores (mainly eating vertebrates such as raptors) based on Wang et al. (2021). In addition, considering that the bird surveys were conducted during the breeding season, feeding guilds will be primarily categorized based on the birds' feeding behavior during this period, for example, some birds (e.g. Meadow Bunting Emberiza cioides and Marsh Tit Poecile palustris) mainly feed on grains in the non-breeding season, but shift to primarily consuming insects during the breeding season to meet their increased energy demands.
2.4
Habitats influencing bird communities
Environmental characteristic were measured by combining field surveys and unmanned aerial vehicles (UAVs). The UAV, a DJI Mavic 3 Pro, was flown at a fixed altitude of 200 m, with the camera oriented vertically downward, capturing images that covered approximately 5.6 ha each. These images, known for their high accuracy and timeliness compared to traditional satellite imagery (Everaerts, 2008), were analyzed in ArcGIS 10.2 to determine the proportions of semi-natural habitats and crop types. Semi-natural habitats were categorized into four types: woodland, shrub, herb (including grassy strips between the fields) and waterbody (including ponds and rivers). Additionally, in light of the crucial support provided by human settlements (e.g. Marcacci et al., 2020), we have included them in the analysis. Details of environmental data see Appendix Table S1. These semi-natural habitat types were selected because they were biologically meaningful for the bird community in agricultural ecosystems (Wretenberg et al., 2010; Wilson et al., 2017; Garcia et al., 2023). In light of the crucial support that traditional human settlements provide within agricultural ecosystems, particularly in terms of nesting sites and food resources for birds—especially for generalist species—we have included human settlements as one of the semi-natural habitat types (e.g., Guyot et al., 2017; Marcacci et al., 2020; Rosin et al., 2021). Habitat complexity is considered a key determinant in influencing bird diversity and habitat selection (Fischer et al., 2011; Sam et al., 2019). To assess habitat complexity and crop diversity, we calculated the number of semi-natural habitat types and crop types present in each sampling site. Details of environmental variables are shown in Table 1 and Appendix Table S1.
2.5
Statistical analyses
All analyses were conducted in the R 3.6.1 environment (R Core Team, 2019). Before the analysis, all variables were standardized (mean = 0, standard deviation = 1) to improve model fitting and facilitate variable selection. The correlations between variables were examined to avoid multicollinearity by the variance inflation factor (VIF) analyses (VIF < 4; Hijmans and Van Etten, 2012; Appendix Table S2). The same statistical approach was used for the six groups: “all species”, “residents”, “insectivorous”, “granivorous”, “omnivorous”, and “carnivorous”.
2.5.1
Birds in different landscape agricultural intensities
We used ‘vegan’ package to conduct nonmetric multidimensional scaling (NMDS) analysis on the presence-absence data in each sampling site by using Jaccard distance to visualize the differences in species composition of birds among the low-, middle-, and high-intensity agriculture (Oksanen et al., 2020). To test for differences in community composition among different landscape agricultural intensities, we performed a pairwise permutation MANOVAs test using abundance data to compare bird composition in different landscape agricultural intensities in the ‘RVAideMemoire’ package (Herve, 2021). We also performed Tukey pairwise comparisons to test for differences in species richness and abundance of diverse bird groups within different landscape agricultural intensities using the ‘multcomp’ package (Hothorn et al., 2008).
2.5.2
Factors of semi-natural habitats influencing bird communities
We explored the relationships between explanatory variables, including environmental predictors (woodland cover, shrub cover, herb cover, human settlement cover, waterbody cover, and landscape agricultural intensity) and interaction between predictors and agricultural intensity, and response variables such as species richness and abundance, using generalized linear models (GLMs) with Poisson distribution in the ‘lme4'package (Bates et al., 2015). We used the Akaike’s Information Criterion score corrected for a small sample size (AICc; Burnham and Anderson, 2002) to identify factors that influencing bird communities. We selected the most or equally plausible models based on ranked ΔAICc scores and Akaike’s weights (ωi) using the ‘dredge’ function in the ‘MuMln’ package (Burnham and Anderson, 2002). Finally, we merged the candidate models with ΔAICc of ≤2 into a single, averaged model to calculate the model-averaged coefficients and the relative importance of each predictor. This approach mitigates the bias inherent and uncertainty in model selection (Lukacs et al., 2009). Model fitting was checked by analyzing residual plots for normality and heteroscedasticity in the ‘DHARMa’ package (Hartig, 2021). We used c-hat value to evaluate the overdispersion of best models, where values > 1 indicate overdispersion (Burnham and Anderson, 2002), and none of them was overdispersed. We used diagnostic plots to exam the final model of each analysis to check for heteroscedasticity, non-normal errors and outliers. The spatial autocorrelation of all dependent variables was tested using global Moran’s I in the ‘ape’ package (Gittleman and Kot, 1990; Appendix Table S3).
2.5.3
Habitat complexity versus crop diversity
Additionally, we used GLMs to investigate the effects of habitat complexity and crop diversity on bird communities, where habitat complexity and crop diversity as well as their interaction were set as fix effects, and species richness and abundance of residents, insectivores, omnivores, granivores and overall bird species as response variables respectively.
3.
Result
3.1
Birds in different landscape agricultural intensities
We recorded a total of 3781 individuals of 72 species were recorded across the 322 sampling sites. To minimize potential biases in the survey data, we excluded species that were observed at three or fewer sampling sites. Ultimately, 3742 records of 49 bird species remained in the bird relative abundance matrix, of which 34 species are migrants, 15 species are residents; 31 species are insectivores, 9 species are omnivores, 6 species are carnivores and 3 species are granivores (Appendix Table S4). Due to the low abundance, carnivores were excluded from further analysis.
The 10 most abundant species were Eurasian Tree Sparrow (Passer montanus) (746 individuals), Red-rumped Swallow (Cecropis daurica) (320), Oriental Turtle-dove (Streptopelia orientalis) (270), Eurasian Magpie (Pica pica) (237), Azure-winged Magpie (Cyanopica cyanus) (233), Common Pheasant (Phasianus colchicus) (205), Common Cuckoo (Cuculus canorus) (168), Barn Swallow (Hirundo rustica) (159), Great Tit (Parus major) (126), Grey-backed Thrush (Turdus hortulorum) (108), Marsh Tit (Poecile palustris) (105). Together, these 10 species accounted for 72% of the total individuals recorded.
The NMDS ordination plots suggested strong community composition differences between different agricultural intensity landscapes (Fig. 2). The pairwise permutation MANOVA comparison also revealed significant differences in species composition between different agricultural intensity landscapes (all pairwise comparisons had P < 0.001). The middle-intensity agriculture supported the greatest overall species richness and abundance (7.08 ± 2.28; 33.20 ± 55.99) compared to the low- (4.82 ± 1.84; 19.31 ± 33.15) and high-intensity (5.62 ± 1.91; 23.85 ± 47.91) agriculture. The high-intensity agriculture supported the most species-rich of omnivores (1.98 ± 0.85) and granivores (1.59 ± 0.77). Low-intensity agriculture is only better than high-intensity agriculture in supporting insectivores’ species richness and abundance (3.42 ± 1.52 vs. 1.85 ± 1.30; 5.65 ± 3.80 vs. 2.90 ± 2.56), but lower than middle-intensity agriculture, for all other biodiversity index, it is the lowest. For more details see Fig. 3.
Figure
2.
Nonmetric multidimensional scaling (NMDS) ordination based on Jaccard distance using bird presences data in low-, middle- and high-intensity agriculture (stress = 0.17).
Figure
3.
Box plot showing the richness and abundance of residents, insectivores, omnivores, granivores, and overall bird species in low-, middle-, and high-intensity agriculture. Significance levels from post-hoc Tukey tests adjusted for multiple comparisons as follows: *P < 0.05; **P < 0.01.
3.2
Factors of semi-natural habitats influencing bird communities
3.2.1
All species and resident species
We found a significant positive effect of increased woodland cover in middle-intensity agriculture on overall bird species richness (β = 0.15, P < 0.01) and abundance (β = 0.34, P < 0.01) despite woodland cover being negatively correlated with overall bird abundance in the entire study area (β = −0.11, P < 0.01). Increased human settlement cover in middle-intensity agriculture promotes the abundance of both all species (β = 0.11, P < 0.01) and residents (β = 0.22, P < 0.01), while human settlement cover was negatively correlated with residents’ abundance in other landscape agricultural intensities. The waterbody cover significantly impacted overall species abundance (β = 0.05, P < 0.01), independently of the landscape agricultural intensity. Species richness and abundance of residents decreased with the decline of landscape agricultural intensity (β = −0.88, P < 0.01; β = −0.16, P = 0.04). For more details see Table 2 and Fig. 4.
Table
2.
Model-averaged parameter estimates for the effects of semi-natural habitats on different bird groups within low-, middle- and high-intensity agriculture, only significant explanatory variables (P < 0.05) were shown in this table. FarmL and FarmM represent low- and middle-intensity agriculture respectively.
Figure
4.
The relative effects of predictors in low-, middle-, and high-intensity agriculture. Low-, middle-, and high-intensity agriculture is represented with blue, green, and red lines respectively, 95% confidence interval are represented by varying colored belts.
The landscape agricultural intensity significantly affected insectivores’ species richness (β = 0.62, P < 0.01; β = 0.79, P < 0.01) and abundance (β = 0.66, P < 0.01; β = 0.90, P < 0.01). Shrub cover was the predictor which best explained the abundance of insectivores. An interaction between shrub cover and the landscape agricultural intensity suggested that insectivores will benefit from increasing shrub proportion only in high-intensity agriculture (β = 0.14, P < 0.01). High-intensity agriculture was found to play an important role in maintaining omnivores and granivores. Abundance of omnivores was negatively affected by herb cover in all landscape agricultural intensity (β = 0.15, P < 0.01), and positively affected by human settlement cover only in high-intensity agriculture (β = 0.27, P < 0.01). In addition, we did not find the influence of semi-natural habitats on granivores in this study. For more details see Table 2 and Fig. 4. All competitive models from the model selection procedure were showed in Appendix Table S5.
3.2.3
Habitat complexity vs. crop diversity
Crop diversity, habitat complexity, and interactions between them all showed positive effects on bird species. Specifically, crop diversity influenced overall (β = 0.13, P < 0.01), residents (β = 0.09, P < 0.01), omnivores (β = 0.09, P < 0.01), granivores abundance (β = 0.13, P < 0.01), and overall bird species richness (β = 0.05, P = 0.04). Habitat complexity had an effect on overall bird species richness (β = 0.12, P < 0.01) and abundance (β = 0.22, P < 0.01), and resident species richness (β = 0.07, P = 0.03). Interactions with habitat and crop diversity affected the species richness of residents (β = 0.07, P = 0.04) and the abundance of residents (β = 0.05, P = 0.02), insectivores (β = 0.07, P < 0.01), and granivores (β = 0.09, P = 0.02). More details see Table 3.
Table
3.
Effects of habitat complexity, crop diversity, and interactions between them on bird communities, only significant explanatory variables (P < 0.05) were shown in this table. FarmL and FarmM represent low- and middle-intensity agriculture respectively.
Most studies examining the relationship between agricultural intensity and bird diversity have focused on Western countries (Guerrero-Casado et al., 2023). The positive effects of semi-natural habitats on biodiversity have been extensively validated (e.g. Marcacci et al., 2022; Olimpi et al., 2022; Garcia et al., 2023). In this study, we investigated the relative effects of the semi-natural habitats within different landscape agricultural intensities (low-middle-high) on avian community characteristics in the central and eastern of Jinlin Province of China. By disentangling these two independent factors, we demonstrated that the effects of agriculture on bird communities were predominantly influenced by the amount of semi-natural habitats, with varying effects depending on landscape agricultural intensity. We found that species richness and abundance of overall bird species were highest in middle-intensity agriculture. The effects of semi-natural habitats on bird communities were stronger within middle-intensity agriculture than low and high intensity, which emphasized the importance of considering the effects of environmental context when exploring the benefits of wildlife-friendly habitats. High-intensity agriculture tended to attracts more omnivores, granivores, and residents, many of which are habitat generalists. In contrast, low-intensity agriculture played an important role in supporting typical insectivores, likely due to the surrounding natural vegetation, emphasizing the necessity of investigating different avian guilds. In addition, both increased crop diversity and habitat complexity have positive effects on bird communities.
4.1
Birds in different landscape agricultural intensities
While previous studies have demonstrated the significant role of low-intensity agriculture in maintaining bird diversity (Doxa et al., 2010; Karp et al., 2012; Elsen et al., 2017), our study found that only insectivore species richness was higher in low-intensity agriculture compared to other landscape agricultural intensities. The surrounding contiguous natural habitats in low-intensity agriculture could provide specific resources (e.g. tree holes) for part of insectivores such as Great Tit (Parus major) and Marsh Tit (Poecile palustris) (Edwards et al., 2010). In contrast, high-intensity agriculture supports higher abundances of residents, omnivores, and granivores, potentially attracted by seeds and plant debris from spring ploughing. However, further agricultural expansion may lead to species homogenization (Karp et al., 2012). The reduced species richness and abundance of insectivores in high-intensity agriculture may be associated with the decreased food availability due to agrochemicals use (e.g., Zielonka et al., 2024). The middle-intensity agriculture supported the highest overall bird abundance. A possible reason is that middle-intensity agriculture simultaneously allowed most insectivores that relied on forest resources and species prefer open, flat habitats to coexist. Birds depending on forest or forest patches as nesting habitats can exploit farmlands for resources, such as White-cheeked Starling (Spodiopsar cineraceus) which usually nests in the forest but often forage in farmlands. Similarly, birds that prefer open habitat and relied on corps as important food resources could benefit from field border vegetation (Conover et al., 2009), suggesting natural vegetation and farmland mosaics are crucial for maintaining higher bird diversity in farmlands.
4.2
Factors of semi-natural habitats influencing bird communities
4.2.1
All species and resident species
In our study, bird composition in different intensity agriculture was largely driven by the amount of semi-natural habitat. While numerous studies have demonstrated the positive impact of woodland cover on bird diversity in agricultural landscapes (Assandri et al., 2016; Guyot et al., 2017; Zingg et al., 2018), our findings revealed a nuanced relationship. Specifically, woodland cover exhibited a negative correlation with overall bird abundance, despite positively affecting bird populations in middle-intensity agriculture. Although the spatial configuration of semi-natural structures was not measured in this study, in most cases, increased woodland cover was positively correlated with increased vegetation spatial heterogeneity (Le Brocque et al., 2009; Lindenmayer et al., 2023), which provided more available niches for insectivores. In high-intensity agriculture, the lack of surrounding natural habitat results in the fewest insectivorous species dependent on stratified vegetation, localized increases in forest cover appear to be insufficient for birds that require continuous natural habitat. While increased forest cover in low-intensity agriculture results in fewer birds which prefer open, flat habitats. The broadly acknowledged viewpoint holds that urbanization leads to an increase in abundance but not species richness of animals (Shochat et al., 2010; Saari et al., 2016). In line with this, we found that human settlement cover has a positive effect on overall bird abundance, especially in low- and middle-intensity agriculture, species richness of all groups was not affected by human settlement cover. One possible reason is that some species with generalist niche requirements such as swallows or pigeons, could benefit from scattered human buildings due to the availability of nesting sites, food resources, and lower predation rates (McKinney and Lockwood, 1999; Shochat et al., 2010). Enhancing the supporting role of agricultural landscapes especially in high-intensity agriculture to bird diversity through increased human settlements should be treated with caution to prevent generalists increase in numbers due to their competitive superior abilities, and thus decrease community evenness (Shochat et al., 2010). Furthermore, the contribution of human buildings to biodiversity may vary with various factors such as scale, surrounding habitat properties, area, and density of human buildings (Morelli et al., 2021; Petit et al., 2023). The water surface in agricultural landscapes such as ponds, ditches, and natural river could provide a water storage capacity of the farmland to avoid the negative effects of inundation of farmland caused by heavy rainfall (Bradbury and Kirby, 2006). Meanwhile, waterbody can also provide important resource supplements in the breeding season for some wetland-dependent bird species such as wagtails and ducks (Sebastián-González et al., 2010), especially in earlier reproductive periods with the spring thaw.
4.2.2
Feeding guilds
In line with previous studies, insectivores are the most common guild in agricultural landscapes (Redlich et al., 2018; Narayana et al., 2019). Shrub and woodland cover are the most important predictors of insectivore abundance and their effects were dependent on landscape agricultural intensity. Woodland and shrub patches can provide food resources, shelter, and nesting substrates for bird communities, especially in highly intensive agricultural landscapes (Frid and Dill, 2002). These patches can enhance landscape connectivity and supplement the lack of surrounding natural vegetation, thereby benefiting insectivorous birds (Stanton et al., 2021). By protecting or planting semi-natural vegetation such as woodland and shrubs within high-intensity agriculture could not only promote bird diversity but also potentially increase the provision of pest control services provided by avian species. In contrast to woody elements, semi-natural herbs vary significantly in structural heterogeneity and quality (Tschumi et al., 2020). Our results indicate that herb cover has a negative effect on omnivores. We speculate that the simple structure and low height of herb vegetation within agricultural landscape may be responsible for this outcome. Most omnivores recorded in this study, such as the Carrion Crow (Corvus corone) mainly nest on shrubs or branches. Increased herb cover not only fail to provide adequate refuge but also reduce food availability (Andersen and Steidl, 2020). While an increase in herbaceous vegetation may benefit classic grassland birds (Korejs et al., 2024), this is dependent on the surrounding natural landscape. Previous studies have proposed that granivores are more influenced by the availability of reliable foraging opportunities across broader patch matrix scales rather than by local environmental characteristics, compared to other foraging guilds (Chamberlain et al., 2010; Hempson et al., 2015). Consistent with this, several granivores identified in this study appear to be driven by landscape agricultural intensity (see section 4.1) rather than by the composition of semi-natural habitats, at least during the breeding season.
4.2.3
Habitat complexity vs. crop diversity
Crop diversity and habitat complexity within agricultural landscapes are recognized as significant predictors for birds elsewhere (Josefsson et al., 2017; Wilson et al., 2017). Our findings indicate that crop diversity plays a predominant role in explaining the abundance of most bird groups, except for insectivores, suggesting that they are more sensitive to the landscape agricultural intensification of large monocultures (Marcacci et al., 2020). Agricultural practice with mixed crops could provide varied food resources that support generalists, and even specialist birds may utilize them based on the proportion of available crop resources (Katuwal et al., 2022). The positive effects of interactions between habitat complexity and crop diversity may be explained by an extension of the “area–heterogeneity trade off hypothesis” by Khan et al. (2023). Increased crop diversity within a given area is usually achieved at the expense of area per crop. The positive effects of diverse crops are more pronounced when abundant semi-natural structures (habitat complexity) in agricultural landscapes can provide additional resources for bird communities especially for insectivores. Hence, when considering increasing crop diversity as a method to promote biodiversity in cropped areas of landscapes, semi-natural structures should be considered at the same time to maximize conservation efforts and improve conservation efficiency.
4.3
Conclusions and conservation implications
As the growing human population and the demand for food intensifies, there is a pressing need to design agroecosystems that harmoniously support both people and nature in the context of the global agriculturalization process (Marcacci et al., 2022). Our study showed, for the first time in China, the effect of landscape-scale agricultural intensity on the potential benefit of semi-natural habitats on bird community and evaluated the supporting role of habitat complexity in agro-ecosystem. The negative impacts of agricultural intensification on most of the bird communities can be mitigated by preserving or promoting semi-natural structure Given the varying impacts of semi-natural habitats across different levels of landscape agricultural intensity, we propose several conservation strategies to enhance the efficiency of agricultural landscape management. Specifically, priority should be given to preserving or/and planting these semi-natural habitats in middle-intensity agriculture due to the stronger effect sizes on bird diversity. In high-intensity agricultural areas, practices that encourage shrub regrowth should be implemented. Increasing shrub cover can promote avian diversity, especially those offered by insectivorous birds that prey on insect pests by enhancing the vertical heterogeneity of vegetation. Agricultural practice with mixed crops could provide different food choices for variety bird groups, especially in areas with high semi-natural habitat complexity. The Chinese government steadfastly adheres to the policies of the “cultivated land red line” and “ecological conservation redline” to ensure the attribute of existing farmland and natural habitats must not be altered easily. On this basis, our findings could provide new insight for the development of effective conservation and management measures within varying landscape agricultural intensities. From a socio-economic perspective, the extent to which this agricultural landscape management approach is sustainable still requires further investigation.
We declare that we have no financial and personal relationships with other people or organizations that can inappropriately influence our work, there is no professional or other personal interest of any nature or kind in any product, service and/or company that could be construed as influencing the position presented in, or the review of, the manuscript entitled “Disentangling the relative effects of semi-natural habitats within different agricultural intensities on avian community characteristics”.
Abs M, Bergen F. A long term survey of the avifauna in an urban park. In: Marzluff JM, Shulenberger E, Endlicher W, Alberti M, Bradley G, Ryan C, et al., editors. Urban ecology: an international perspective on the interaction between humans and nature. Boston: Springer; 2008. p. 373–6.
Assandri G, Bogliani G, Pedrini P, Brambilla M. Insectivorous birds as 'non-traditional' flagship species in vineyards: applying a neglected conservation paradigm to agricultural systems. Ecol Indic. 2017a;80: 275–85.
Assandri G, Giacomazzo M, Brambilla M, Griggio M, Pedrini P. Nest density, nest-site selection, and breeding success of birds in vineyards: management implications for conservation in a highly intensive farming system. Biol Conserv. 2017b;205: 23–33.
Baillie J, Hilton-Taylor C, Stuart SN. IUCN red list of threatened species: a global species assessment. Gland: IUCN–The World Conservation Union; 2004.
Barton K. MuMIn: Multi-model inference. R package version 1.43.17. 2020. .
Beaugeard E, Brischoux F, Angelier F. Green infrastructures and ecological corridors shape avian biodiversity in a small French city. Urban Ecosyst. 2020. .
Betts MG, Diamond AW, Forbes GJ, Villard M-A, Gunn JS. The importance of spatial autocorrelation, extent and resolution in predicting forest bird occurrence. Ecol Model. 2006;191: 197–224.
Bivand RS, Wong DWS. Comparing implementations of global and local indicators of spatial association. TEST. 2018;27: 716–48.
Brédif H, Simon L. Ordinary biodiversity, local stakeholders and forest management as a driver for regional sustainable development. J Forest. 2014;04: 249–58.
Brouat C, Chevallier H, Meusnier S, Noblecourt T, Rasplus J-Y. Specialization and habitat: spatial and environmental effects on abundance and genetic diversity of forest generalist and specialist Carabus species. Mol Ecol. 2004;13: 1815–26.
Buckley NJ. The new atlas of breeding birds in Britain and Ireland: 1988–1991. In: Gibbons DW, Reid JB, Chapman RA, editors. The Auk. London: T & AD Poyser; 1995. p. 812–3.
Burnham KP, Anderson DR. Model selection and multi-model inference: a practical information-theoretic approach. 2nd ed. New York: Springer; 2002.
Butler RW. Population regulation of wading Ciconiiform birds. Colon Waterbird. 1994;17: 189–99.
Carrara E, Arroyo-Rodríguez V, Vega-Rivera JH, Schondube JE, de Freitas SM, Fahrig L. Impact of landscape composition and configuration on forest specialist and generalist bird species in the fragmented Lacandona rainforest, Mexico. Biol Conserv. 2015;184: 117–26.
Clavel J, Julliard R, Devictor V. Worldwide decline of specialist species: toward a global functional homogenization? Front Ecol Environ. 2011;9: 222–8.
Clergeau P, Croci S, Jokimäki J, Kaisanlahti-Jokimäki M-L, Dinetti M. Avifauna homogenisation by urbanisation: analysis at different European latitudes. Biol Conserv. 2006;127: 336–44.
Cockle KL, Martin K, Drever MC. Supply of tree-holes limits nest density of cavity-nesting birds in primary and logged subtropical Atlantic forest. Biol Conserv. 2010;143: 2851–7.
Dabelsteen T. Variation in the response of freeliving Blackbirds Turdus merula to playback of song: I. Effect of continuous stimulation and predictability of the response. Z Tierpsychol. 1982;58: 311–28.
Dabelsteen T. Variation in the response of freeliving Blackbirds Turdus merula to playback of song: II. Effect of time of day, reproductive status and number of experiments. Z Tierpsychol. 1984;65: 215–27.
Dabelsteen T. Interactive playback: a finely tuned response. In: McGregor PK, editor. Playback and studies of animal communication. Boston: Springer; 1992. p. 97–109.
Dabelsteen T, Pedersen SB. Song and information about aggressive responses of blackbirds, Turdus merula: evidence from interactive playback experiments with territory owners. Anim Behav. 1990;40: 1158–68.
Davis BNK. Soil animals as vectors of organochlorine insecticides for ground-feeding birds. J Appl Ecol. 1966;3: 133–9.
Devictor V, Julliard R, Jiguet F. Distribution of specialist and generalist species along spatial gradients of habitat disturbance and fragmentation. Oikos. 2008;117: 507–14.
Diehl P, Helb H-W. Radiotelemetric monitoring of heart-rate responses to song playback in blackbirds (Turdus merula). Behav Ecol Sociobiol. 1986;18: 213–9.
Donald PF, Green RE, Heath MF. Agricultural intensification and the collapse of Europe's farmland bird populations. Proc Biol Sci. 2001;268: 25–9.
Ellis EC, Ramankutty N. Putting people in the map: anthropogenic biomes of the world. Front Ecol Environ. 2008;6: 439–47.
Eötvös CB, Magura T, Lövei GL. A meta-analysis indicates reduced predation pressure with increasing urbanization. Landsc Urban Plan. 2018;180: 54–9.
Evans KL, Hatchwell BJ, Parnell M, Gaston KJ. A conceptual framework for the colonisation of urban areas: the blackbird Turdus merula as a case study. Biol Rev. 2010;85: 643–67.
Fernández-Juricic E, Jimenez MD, Lucas E. Bird tolerance to human disturbance in urban parks of Madrid (Spain): management implications. In: Marzluff JM, Bowman R, Donnelly R, editors. Avian ecology and conservation in an urbanizing world. Boston: Springer; 2001. p. 259–73.
Foley JA, Defries R, Asner GP, Barford C, Bonan G, Carpenter SR, et al. Global consequences of land use. Science. 2005;309: 570–4.
Fritsch C, Coeurdassier M, Faivre B, Baurand P-E, Giraudoux P, van den Brink NW, et al. Influence of landscape composition and diversity on contaminant flux in terrestrial food webs: a case study of trace metal transfer to European blackbirds Turdus merula. Sci Total Environ. 2012;432: 275–87.
Futuyma DJ, Moreno G. The evolution of ecological specialization. Annu Rev Ecol Syst. 1988;19: 207–33.
Gaston KJ, Fuller RA. Commonness, population depletion and conservation biology. Trends Ecol Evol. 2008;23: 14–9.
Guittet M, Sibe V, Gaudin J-C. Les vignobles: de nouveaux réservoirs de biodiversité. Faune sauvage. 2011: 9.
Godet L. La « nature ordinaire » dans le monde occidental. L'Espace géographique. 2010;39: 295–308.
Gregory RD, Noble DG, Custance J. The state of play of farmland birds: population trends and conservation status of lowland farmland birds in the United Kingdom. Ibis. 2004;146: 1–13.
Hanowski JM, Niemi GJ, Blake JG. Statistical perspectives and experimental design when counting birds on line transects. Condor. 1990;92: 326–35.
Hatchwell BJ, Chamberlain DE, Perrins CM. The demography of Blackbirds Turdus merula in rural habitats: is farmland a sub-optimal habitat? J Appl Ecol. 1996;33: 1114.
Hinsley SA, Bellamy PE, Newton I, Sparks TH. Habitat and landscape factors influencing the presence of individual breeding bird species in woodland fragments. J Avian Biol. 1995;26: 94–104.
Hiron M, Berg Å, Eggers S, Josefsson J, Pärt T. Bird diversity relates to agri-environment schemes at local and landscape level in intensive farmland. Agr Ecosyst Environ. 2013;176: 9–16.
Hobday AJ, Chambers LE, Arnould JPY. Prioritizing climate change adaptation options for iconic marine species. Biodivers Conserv. 2015;24: 3449–68.
Hole DG, Whittingham MJ, Bradbury RB, Anderson GQA, Lee PLM, Wilson JD, et al. Widespread local house-sparrow extinctions — Agricultural intensification is blamed for the plummeting populations of these birds. Nature. 2002;418: 931–2.
Jankowiak Ł, Pietruszewska H, Wysocki D. Weather conditions and breeding season length in blackbird (Turdus merula). Folia Zool. 2014;63: 245–50.
Jokimäki J, Suhonen J. Distribution and habitat selection of wintering birds in urban environments. Landsc Urban Plan. 1998;39: 253–63.
Julliard R, Jiguet F. Un suivi intégré des populations d'oiseaux communs en France. Alauda. 2002;70: 137–47.
Julliard R, Clavel J, Devictor V, Jiguet F, Couvet D. Spatial segregation of specialists and generalists in bird communities. Ecol Lett. 2006;9: 1237–44.
Kassen R. The experimental evolution of specialists, generalists, and the maintenance of diversity: experimental evolution in variable environments. J Evol Biol. 2002;15: 173–90.
Kubel JE, Yahner RH. Detection probability of Golden-winged Warblers during point counts with and without playback recordings. J Field Ornithol. 2007;78: 195–205.
Larsen AE, Noack F. Identifying the landscape drivers of agricultural insecticide use leveraging evidence from 100, 000 fields. Proc Natl Acad Sci USA. 2017;114: 5473–8.
Lê S, Josse J, Husson F. FactoMineR: an R package for multivariate analysis. J Stat Softw. 2008;25: 1–18.
Lee DC, Marsden SJ. Adjusting count period strategies to improve the accuracy of forest bird abundance estimates from point transect distance sampling surveys: count period strategies for distance sampling surveys. Ibis. 2008;150: 315–25.
Lennon JJ, Beale CM, Reid CL, Kent M, Pakeman RJ. Are richness patterns of common and rare species equally well explained by environmental variables? Ecography. 2011;34: 529–39.
Luck GW, Smallbone LT. Species diversity and urbanisation: patterns, drivers and implications. In: Gaston KJ, editor. Urban Ecology. Cambridge: Cambridge University Press; 2010. p. 88–119.
Luniak M. Synurbization — adaptation of animal wildlife to urban development. In: Shaw WW, Harris LK, Vandruff L, editors. Proceedings 4th international urban wildlife symposium. Tucson, Arizona: University of Arizona; 2004. p. 50–5.
Luniak M, Kozlowki P, Nowicki W. Magpie Pica pica in Warsaw — abundance, distribution and changes in its population. Acta Orn. 1997;32: 77–86.
Mac Nally RC. The relationship between habitat breadth, habitat position, and abundance in forest and woodland birds along a continental gradient. Oikos. 1989;54: 44–54.
Melles S, Glenn SM, Martin K. Urban bird diversity and landscape complexity: species-environment associations along a multiscale habitat gradient. Conserv Ecol. 2003;7: 5.
Mennechez G, Clergeau P. Effect of urbanisation on habitat generalists: starlings not so flexible? Acta Oecol. 2006;30: 182–91.
Najmanová L, Adamík P. Effect of climatic change on the duration of the breeding season in three European thrushes. Bird Study. 2009;56: 349–56.
Newbold T, Hudson LN, Arnell AP, Contu S, De Palma A, Ferrier S, et al. Has land use pushed terrestrial biodiversity beyond the planetary boundary? A global assessment. Science. 2016;353: 288–91.
Newton I. Population limitation in birds: the last 100 years. Br Birds. 2007;100: 518–39.
Olden JD. Biotic homogenization: a new research agenda for conservation biogeography. J Biogeogr. 2006;33: 2027–39.
Paquet M, Arlt D, Knape J, Low M, Forslund P, Pärt T. Quantifying the links between land use and population growth rate in a declining farmland bird. Ecol Evol. 2019;9: 868–79.
Pithon JA, Beaujouan V, Daniel H, Pain G, Vallet J. Are vineyards important habitats for birds at local or landscape scales? Basic Appl Ecol. 2016;17: 240–51.
R Core Team. R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing; 2019.
Ralph CJ, Droege S, Sauer JR. Managing and monitoring birds using point counts: standards and applications. In: Ralph CJ, Droege S, Sauer JR, editors. Monitoring bird populations by point counts. Albany: U.S. Department of Agriculture, Forest Service, Pacific Southwest Research Station; 1995. p. 161–8.
Reif J, Voříšek P, Šťastnyˇ K, Bejček V, Petr J. Population increase of forest birds in the Czech Republic between 1982 and 2003. Bird Study. 2007;54: 248–55.
Rotenberry JT, Wiens JA. A synthetic approach to principal component analysis of bird/habitat relationships. In: Capen DE, editor. The use of multivariate statistics in studies of wildlife habitat. Fort-Collins: Rocky Mountain Forest and Range Experiment Station; 1981. p. 197–208.
Sandström UG, Angelstam P, Mikusiński G. Ecological diversity of birds in relation to the structure of urban green space. Landsc Urban Plan. 2006;77: 39–53.
Shultz SB, Bradbury RL, Evans KD, Gregory RM, Blackburn T. Brain size and resource specialization predict long-term population trends in British birds. P Roy Soc B-Biol Sci. 2005;272: 2305–11.
Sierro A, Arlettaz R. L'avifaune du vignoble en Valais central: évaluation de la diversité à l'aide de transects. Nos Oiseaux. 2003;50: 89–100.
Siriwardena GM, Baillie SR, Buckland ST, Fewster RM, Marchant JH, Wilson JD. Trends in the abundance of farmland birds: a quantitative comparison of smoothed Common Birds Census indices. J Appl Ecol. 1998;35: 24–43.
Snow DW. Territory in the Blackbird Turdus Merula. Ibis. 1956;98: 438–47.
Sol D, González-Lagos C, Moreira D, Maspons J, Lapiedra O. Urbanisation tolerance and the loss of avian diversity. Ecol Lett. 2014;17: 942–50.
Stanton RL, Morrissey CA, Clark RG. Analysis of trends and agricultural drivers of farmland bird declines in North America: a review. Agr Ecosyst Environ. 2018;254: 244–54.
Steel ZL, Steel AE, Williams JN, Viers JH, Marquet PA, Barbosa O. Patterns of bird diversity and habitat use in mixed vineyard-matorral landscapes of Central Chile. Ecol Indic. 2017;73: 345–57.
Wretenberg J, Lindström Å, Svensson S, Thierfelder T, Pärt T. Population trends of farmland birds in Sweden and England: similar trends but different patterns of agricultural intensification. J Appl Ecol. 2006;43: 1110–20.
Table
2.
Model-averaged parameter estimates for the effects of semi-natural habitats on different bird groups within low-, middle- and high-intensity agriculture, only significant explanatory variables (P < 0.05) were shown in this table. FarmL and FarmM represent low- and middle-intensity agriculture respectively.
Table
3.
Effects of habitat complexity, crop diversity, and interactions between them on bird communities, only significant explanatory variables (P < 0.05) were shown in this table. FarmL and FarmM represent low- and middle-intensity agriculture respectively.
DownLoad:
Email Alerts
RSS Feeds