학술

Forest proximity and habitat heterogeneity support insectivorous birds and arthropods in California vineyards

PLOS ONE
CC BY
이 매체는 공공·자유 라이선스로 본문을 직접 표시합니다.

Figures
Abstract
Agricultural intensification often simplifies habitat structure, reshaping multi-trophic communities by altering arthropod prey and the insectivorous birds that depend on them throughout the year, particularly during breeding. In vineyards, insectivorous birds may also contribute to pest suppression, creating opportunities to link biodiversity conservation with agroecosystem function. We paired point count bird surveys with concurrent sticky trap arthropod sampling across 20 organic or near-organic vineyards in Napa Valley, California, USA, to test whether the same habitat features predict variation in insectivorous bird guilds and arthropod communities. Specifically, we modeled guild-level bird abundance and arthropod order-level diversity and abundance as a function of habitat heterogeneity at local (50 m) and landscape (1000 m) scales, as well as proximity to forest, riparian, and water habitats. Foliage- and bark-gleaning insectivorous birds were more abundant in locally heterogeneous vineyards close to forest and riparian edges, whereas aerial insectivores were not associated with measured habitat covariates. Several prey-relevant arthropod orders were more abundant closer to forests and in locally heterogeneous sites, while arthropod order-level diversity was not associated with measured habitat covariates. Structural equation modeling demonstrated that forest proximity and landscape-scale heterogeneity indirectly supported foliage-bark gleaners by enhancing large beetle and large fly prey abundances, respectively. Together, our results indicate that increasing habitat heterogeneity within vineyards, coupled with maintaining adjacent natural habitats, can support insectivorous bird assemblages and their prey, with potential benefits for biodiversity and biologically based pest management in vineyard landscapes.
Citation: MacDonald EK, Karp DS, Turner A, Martinico BL, Wilson H, Wilson-Rankin EE, et al. (2026) Forest proximity and habitat heterogeneity support insectivorous birds and arthropods in California vineyards. PLoS One 21(11): e0363054. https://doi.org/10.1371/journal.pone.0363054
Editor: Nickson Erick Otieno, National Museums of Kenya, KENYA
Received: July 2, 2026; Accepted: November 3, 2026; Published: November 20, 2026
Copyright: © 2026 MacDonald et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All data and code files supporting the findings of this study are publicly available from the Zenodo repository (DOI: https://doi.org/10.5281/zenodo.23369690, URL: https://zenodo.org/records/23369690).
Funding: This project was supported by the California State University Agricultural Research Institute (https://ari.calstate.edu/) through grant number 22-99-107, awarded to D.S.K., H.W., E.E.W-R., and M.D.J. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
1 Introduction
The joint dependence of insectivorous birds and arthropods on habitat structure makes them vulnerable to landscape simplification, a hallmark of agricultural conversion and intensification that has reconfigured ecological communities from regional to global scales [1–7]. Despite being negatively affected by intensive agricultural practices, insectivorous birds can contribute ecosystem services in working agricultural landscapes by suppressing herbivorous arthropods and influencing crop-pest dynamics [8–10]. Where crop diversification and habitat heterogeneity are maintained, biodiversity and ecological functions, including functions supported by birds, tend to be more stable over time [11–15]. Vineyard ecosystems exemplify the connection between biodiversity stability and habitat heterogeneity, because management practices (i.e., inter-row cover crops, presence of semi-natural habitat, reduced pesticide use) and landscape context can strongly shape within-farm habitat heterogeneity and the biotic communities it supports [16,17]. In a recent global meta-analysis, within-vineyard diversification and reduced-input vineyard practices increased biodiversity and ecosystem service provisioning, with biological pest control showing the strongest gains in landscapes with more semi-natural habitat [18].
During the avian breeding season, habitats at local and landscape scales are expected to influence where insectivorous birds can nest and forage. Correspondingly, studies in agroecosystems, including vineyards, report positive associations between more complex vegetation structure, habitat heterogeneity, and bird composition or abundance [19–25]. During breeding, many passerines, including granivores and frugivores, shift their diet towards insectivory to optimize provisioning their nestlings [26,27]. This seasonal diet shift is often overlooked in habitat and land use assessments, even though prey availability during breeding can constrain reproductive investment and nestling provisioning, as well as shape guild-specific responses to vegetation structure [28–30].
Diet ecology ties these patterns to trophic pathways, with foliage-gleaning, bark-gleaning, and aerial insectivores exploiting distinct prey communities. Among foliage-gleaning passerines, nestling provisioning frequently centers on large, soft-bodied, and protein-rich prey such as Lepidoptera larvae and Orthoptera [31]. Negative demographic responses can thus occur when prey phenology deviates from avian hatching, consistent with trophic mismatch dynamics [30,32–35]. Bark-gleaners exploit arthropods associated with woody substrates; as such, their foraging should track habitat elements that concentrate bark-dwelling prey [36–38]. Finally, aerial insectivores rely on aerial insects during breeding, primarily Diptera but also Ephemeroptera, Trichoptera, Coleoptera, Hemiptera, Hymenoptera, and Odonata [29,39].
The same habitat gradients that shape bird assemblages often also structure arthropod communities. Local diversification (e.g., ground-cover diversity and vegetation structure) and landscape heterogeneity are generally associated with higher arthropod abundance and diversity, though the degree to which different taxa respond may vary with scale and on-farm management, ultimately dictating biological pest control services [40–45]. In vineyard and other perennial agroecosystems, richer herbaceous cover and structural complexity are linked to greater abundances of arthropods (e.g., through diverse, dense cover crops enhancing overall arthropod richness and abundance; [46,47]). At landscape scales, heterogeneous mosaics and flowering resources can also support diverse Diptera and other arthropod orders with season- and site-specific contributions to prey availability [46–49]. In California’s Mediterranean-climate vineyards, habitat heterogeneity is associated with broad prey spectra documented in diets of insectivorous birds, such as western bluebirds (Sialia mexicana), suggesting that structural complexity can propagate through prey to realized diets [50].
Despite these parallels, insectivorous birds and arthropods are rarely evaluated together to test whether similar habitat features predict both groups within the same landscapes and temporal windows. This gap is notable given broad reductions across insectivorous bird guilds and widespread arthropod declines [51,52]. Indeed, avian insectivores have exhibited some of the steepest abundance declines among bird guilds, with aerial insectivores particularly vulnerable. This underscores the need to understand how on-farm management and landscape context affects them, particularly during breeding [1,5,53]. Increasing habitat heterogeneity in working agricultural landscapes could buffer prey supplies and sustain breeding insectivores, though effects may differ among guilds depending on foraging strata and prey phenology [5,7,54]. Management practices that alter near-vine structure (e.g., cover crop diversification, hedgerows, etc.) may further modulate these pathways during the breeding window [25,55].
Here, we examine how local and landscape habitat features shape passerine insectivore assemblages and arthropod prey communities in California vineyards. Our study advances agricultural multi-trophic ecology by using parallel analyses within a common Bayesian framework to model guild-level bird populations (correcting for imperfect detection) as well as order and size class-resolved arthropod prey communities across the same viticultural landscape. Reflecting well-known preferences for large prey among breeding birds, we resolve arthropod prey by both taxonomic order and body size class (< 3 mm vs. ≥ 3 mm) to evaluate size-specific prey availability. To better understand the joint dependence of insectivorous birds and arthropods on habitat structure, we then use candidate piecewise structural equation models (pSEMs) across co-located sites to test whether landscape habitat effects on birds operate directly or indirectly via size-selective prey mediation.
Our work was guided by three inter-related questions: (1) which local and landscape features best predict insectivorous bird abundance and do responses differ by foraging guild (e.g., aerial insectivores vs. foliage- and bark-gleaners)? (2) which local and landscape habitat features best predict arthropod abundance and diversity during the breeding season? and (3) do birds respond to the same habitat features as potential arthropod prey? We predicted that insectivorous bird abundance—including both foliage-bark gleaning and aerial insectivore guilds— would be higher in more heterogeneous habitats, consistent with prior studies in agroecosystems [19,20,22–24,50]. We further predicted that fine-scale structural heterogeneity and broader landscape heterogeneity would be positively associated with insectivorous bird abundance and arthropod community metrics (abundance and diversity). Because diet ecology closely links predator to prey, we expected that local and landscape habitat heterogeneity supporting high arthropod abundance will also predict avian breeding assemblage structure in vineyard ecosystems.
2 Materials and methods
2.1 Study area
The Napa Valley American Viticultural Area (AVA) in California, USA, hereafter Napa Valley, is located ~100 km north of San Francisco, and exhibits a Mediterranean climate, with warm, dry summers and cool, wet winters. This, coupled with a variety of soil types, creates an optimal environment for cultivating winegrapes. Throughout the years, Napa Valley has transformed from a landscape of forests, oak savannas, grasslands, and riparian habitats, tended and maintained with fire and other practices by Indigenous communities (the Wappo or Onasatis, [56]), to one dominated by managed vineyards [57,58], with smaller pockets of natural habitat within and around farms and urban development.
We conducted surveys on 20 vineyards in Napa Valley (Fig 1). Vineyards were selected to span a range of sizes (12.5 to 72 ha, mean 45.04 ± 7.96 SE ha), as well as a broad gradient of local habitat complexity (e.g., cover crop extent, woody vegetation, etc.) and landscape composition (from valley-floor locations surrounded by near-monoculture, to vineyards on the valley edge with more surrounding natural habitat cover). All vineyards in the study system were organic or “near-organic” (non-certified but applying no chemical insecticides).
Left panel shows an example vineyard illustrating the sampling layout with avian point count stations (green) and sticky trap locations (orange). Right panel shows the distribution of sampled vineyards (yellow points) across Napa Valley overlaid on the seven-class land cover raster. The seven land cover classes included and the methodology used to create the land cover raster are described in [59]. Aerial imagery courtesy of the USDA National Agriculture Imagery Program (NAIP), public domain. Base map data sources: County of Napa GIS, USGS National Map, and California Department of Forestry and Fire Protection [60].
2.2 Point count surveys
A single trained observer (A.T.) conducted 16-min, 50 m radius point counts to quantify bird abundance on each vineyard [61]. To maximize farm coverage while minimizing double-counting individuals, survey points were located at least 150 m from each other. Half of the points were located at least 75 m from a vineyard edge (interior points), and half were located approximately 25 m from a vineyard edge (edge points, 4 vines or two rows in, often with a dirt road along the edge as well). To co-locate bird and arthropod sampling, we placed at least one (and in some cases two) point count locations within each arthropod sampling plot (Fig. 1); specifically, 16 arthropod plots contained two co-located point count surveys and 4 plots contained one survey, yielding 36 total co-located surveys. Since vineyards varied in size, the number of survey points per vineyard also varied (2–12 points; mean 7.4 ± 3.02 SD), for a total of 75 interior and 74 edge survey points respectively. Point counts were repeated 3 times at the same locations, in April, May, and June/July corresponding with the passerine nesting season. All surveys were conducted in 2023, began at sunrise, and continued for 5 hours. We rotated vineyard survey order to avoid time-of-day biases. All birds seen and heard were noted. Birds beyond the edge of the fixed radius were noted but not included in the analysis. Likewise, birds that were flying overhead were noted but not included in analyses, except for birds that were actively using the habitat (i.e., foraging swallows). All procedures were in accordance with the Institutional Animal Care and Use Committee (IACUC) (Protocol 2023W12-A).
2.3 Insectivorous bird classification
To evaluate how land use influences insectivorous birds, we examined birds in the order Passeriformes that are insectivorous during the breeding season. We restricted this classification to Passeriformes to maintain a consistent community subset for guild-based analyses; non-passerine insectivores were comparatively infrequent in detections and were therefore not included in the insectivore community metrics. We used the SAviTraits 1.0 trait database [62], which reports the monthly percentage of invertebrates in species’ diets. From this database, we classified a species as a breeding season insectivore if their diet included ≥ 50% invertebrates for at least two months within our April-July survey window. To assign foraging guilds, we used AVONET [63] and classified each species in the insectivore subset by its primary foraging mode. Aerial sallying and aerial insectivory were coded as aerial insectivores, whereas species listed as foliage or bark gleaners were coded as foliage-bark gleaners (Supplementary Table S1 in S1 File).
2.4 Arthropod sampling
Concurrently, we deployed an array of six 23 x 14 cm standard yellow sticky traps (Seabright Laboratories [64],) across a small (~55 x 155 m) grid within each vineyard to survey arthropod communities, with one edge typically bordering non-crop habitat. Traps were mounted at a height of approximately 1 m within the vineyard canopy, positioned between vine rows; trap locations remained fixed. All traps were collected and replaced every two weeks throughout the 12-week field season; however, temporary vineyard access restrictions prevented deployment of 12 individual traps (of 720 planned trap deployments), yielding 708 realized samples.
Captured arthropods were tallied under a dissecting microscope and classified to the taxonomic order level: Hemiptera, Hymenoptera, Coleoptera, Neuroptera, Diptera, Lepidoptera, Thysanoptera, Psocodea, Dermaptera, Araneae, Orthoptera, Blattodea, Ephemeroptera, Raphidioptera, Acari (mites), Collembola, Pseudoscorpiones, and a pooled “Other” category for unidentifiable taxa. We recorded size difference in arthropods to describe their potential value to insectivores, with larger items generally yielding higher dietary returns. Arthropod size was recorded as “small” (< 3 mm) or “large” (≥ 3 mm) for Araneae, Orthoptera, Diptera, Lepidoptera, Coleoptera, Hemiptera, Blattodea, and Ephemeroptera. All other orders were analyzed without size classes. Specifically, groups that were predominantly < 3 mm on our traps (Acari, Collembola, Thysanoptera, Psocodea, Pseudoscorpiones) were not partitioned further, and rarer orders with mixed sizes (Hymenoptera, Neuroptera, Dermaptera, Raphidioptera, Blattodea, Ephemeroptera) were retained as single categories.
2.5 Local and landscape habitat characterization
Using a fine-resolution (~4 m) land cover map developed from multispectral imagery and Lidar data [59], we calculated the proportional cover of seven categories — water, urban, forest, vineyard, riparian, oak savanna, and grassland — within circular buffers of 50 m (local) and 1000 m (landscape) around every point count station and arthropod sampling location. We selected the 50 m scale to match the point count radius, representing the immediate area utilized by birds. For arthropods, this 50 m radius captures the highly localized vegetation structure, including non-crop field margins immediately surrounding the within-vineyard traps, that may provide microclimate, shelter, and resource availability for less mobile taxa. The 1000 m scale was chosen because it encompasses the biologically relevant foraging ranges of mobile avian predators, and reflects the broader landscape matrix that influences the immigration and dispersal of aerially navigating arthropods into the vineyard [65,66].
We validated the land cover raster by visually comparing 50 randomly generated points per land cover class to 2023 Google Earth imagery and recording matches and mismatches prior to analysis. Overall, 4% of points did not match. Given this low disagreement rate, we considered classification accuracy sufficient for our landscape metrics and retained the raster for subsequent analyses. We computed habitat heterogeneity as the Shannon diversity index of the seven land cover proportions at each scale (50 and 1000 m). Because heterogeneity was strongly correlated with the proportion of semi-natural cover (forest, grassland, oak savanna, and riparian), we retained heterogeneity because it better captured compositional diversity. While extensive literature has focused on the total area of semi-natural cover [e.g., 18,24,44], we specifically analyzed landscape heterogeneity to evaluate how the compositional diversity and spatial configuration of surrounding habitats shape community responses, while capturing both the presence and structural diversity of semi-natural features across the landscape. At each point count and arthropod sampling location, we also measured the distance to the nearest patch of each land cover type.
2.6 Nest boxes
Several passerines in our community are secondary cavity nesters, and nest box installation can be used by passerines in agricultural systems [21,67,68]. Therefore, we recorded the number of nest boxes within each 50 m point count radius to account for potential attraction effects, as several sites contained pre-existing songbird nest boxes.
2.7 Cover crop diversification index
Seasonally dynamic cover crop data were collected during each visit to the point counts (3 visits in April, May, June/July) and arthropod sample locations (6 visits at 2-week intervals, April-July). At the local site level (measured directly at and immediately surrounding each point count survey and sticky trap location within vineyard inter-rows), we quantified 4 cover crop management variables: phenology, vegetation height, cover crop composition, and spatial extent, subsequently combining these into a single diversification index. Cover crop phenology was classified as tilled, mowed, or unmowed/untilled (scored as 0, 0.5, and 1.0, respectively) to reflect increasing habitat structure. Standing vegetation height (measured in cm) was binned into low (< 20 cm), medium (20–50 cm), and high (> 50 cm) classes (scored as 0, 0.5, and 1.0). Taller vegetation height classes confer greater structural complexity and support higher arthropod abundance, thereby enhancing prey availability for insectivorous birds [69]. Cover crop composition was recorded as tilled, mowed, planted mostly with grasses, feral mixed species, or planted with mixed species (grasses, flowers, and forbs; scored as 0, 0.25, 0.5, 0.75, and 1.0, respectively). Finally, cover crop extent was categorized as none, between-rows only, or full (under vines and between vine rows; scored as 0, 0.5, and 1.0 respectively). For each visit, the four scores were averaged to yield a composite cover crop diversification index ranging from 0 (uniform, low-diversity cover) to 1 (high-diversity, structurally complex cover) for each trap and point count station, at each visit. Cover crop diversification values ranged from 0 to 1 for both bird point count locations (mean ± 1 SD = 0.63 ± 0.22) and arthropod trap locations (mean ± 1 SD = 0.60 ± 0.27). Because cover crop traits such as height and phenological state co-varied across sampling visits due to seasonal growth and management practices, combining these features into a single composite index provided an integrated metric of realized inter-row vegetation complexity.
2.8 Statistical analyses
2.8.1 Avian insectivores.
All analyses were conducted using R (v4.5.2; [70]). To estimate abundance while accounting for imperfect detection, we modeled point count data (three visits to each point count station) with separate multi-species Bayesian N-mixture models with partial pooling for aerial insectivores and foliage-bark gleaning insectivores [71]. We modeled latent abundance using a negative binomial distribution and modeled imperfect detection by treating observed counts as binomial outcomes conditional on latent abundance, with detection probability modeled on the logit scale. Because monthly revisits relax the closure assumption, we interpret λ (and the derived Ni) as guild-level superpopulation abundance [72,73]. Models included a farm random intercept (to account for repeated measures and spatial dependence). We modeled expected abundance (λ; and derived Ni) as a function of a focused set of abundance covariates: number of songbird nest boxes within 50 m, local (50 m) and landscape (1000 m) heterogeneity, percent water within 1000 m, the cover crop diversification index, and distances to water, forest, and riparian habitat. We allowed for potential non-linear responses to landscape heterogeneity by including both linear and quadratic terms for 1000 m heterogeneity and interpreted these jointly. We included the quadratic term for landscape heterogeneity because we expected guild abundance could peak at intermediate heterogeneity, consistent with the intermediate landscape complexity hypothesis [66], whereas the other habitat covariates were expected to have directional relationships. Detection was modeled with a farm random intercept and several visit-level covariates, including wind, noise, start time, weather, temperature, visit number, and Julian date (entered quadratically). Continuous predictors were centered and scaled. Because distance-to-habitat covariates were right-skewed, we log-transformed distances, using log(distance + 1) to accommodate sites with zero distances (i.e., several sampling locations located on the forest edge). Only species detected at ≥ 5 sites were included to ensure stable estimation (N = 31 species included and 14 excluded). Species-level abundance and detection parameters were drawn from community-level (hyper)distributions so that rarer species could borrow strength from common ones.
We implemented the Markov chain Monte Carlo (MCMC) chains for all models, which were fit in JAGS via R2jags [74] with 3 chains, 50,000 iterations per chain, burn-in of 20,000, and thinning of 5. For the Bayesian analysis, uninformative normal priors were used for covariate coefficients, and half-t or uniform distributions were used for variance parameters. Multi-collinearity among continuous habitat covariates was assessed and found to be well within acceptable limits (VIF < 3). For all models, we assessed model fit using posterior predictive checks (PPCs). Convergence was evaluated using the Gelman-Rubin statistic; we considered chains to have converged by inspecting the R̂ statistics (R̂ ≤ 1.05), trace plots, and effective sample sizes. We report posterior means/medians and 95% Bayesian credible intervals.
2.8.2 Arthropods.
Because foraging selectivity by breeding insectivores often depends on prey size, we analyzed each focal order in three ways when data permitted: the whole order, the subset of “small” individuals (< 3 mm adult body length), and the subset of “large” individuals (≥ 3 mm). We also modeled Shannon diversity (log-transformed for modeling) at the order level for each trap. For focal prey orders, we modeled Diptera, Hemiptera, Coleoptera, Araneae (entire order and small), and Hymenoptera. We excluded from analyses Lepidoptera, large Araneae, and other taxa that were too infrequent (high zero frequency and/or low total counts) to support order- or size-specific models. Although abundant, Thysanoptera and Psocodea were excluded from order-specific models because their small body size makes them unlikely to represent energetically important prey for breeding passerines and because very small arthropods are often underrepresented in passerine nestling diets relative to larger, higher-quality prey such as lepidopteran larvae [75,76]. That said, both orders were retained to quantify Shannon diversity and are reported for descriptive purposes. Arthropod community diversity was assessed at each sticky trap with the Shannon diversity index calculated at the order level, integrating both abundance and evenness among arthropod orders.
Count responses were modeled using Generalized Linear Mixed Models (GLMMs) with a negative binomial distribution (log link) to accommodate overdispersion, while log-transformed Shannon diversity was modeled using a Gaussian distribution. All models used a Bayesian framework and contained local habitat predictors used in the bird models (local [50 m] and landscape [1000 m] heterogeneity, including linear and quadratic terms for 1000 m heterogeneity, cover crop diversification index, and distances to water, forest, and riparian habitat, alongside Julian sampling date, entered quadratically). A random intercept of vineyard site was included to account for repeated measures across sampling dates and spatial clustering of sticky traps within farms. Continuous predictors were centered and scaled; right-skewed distance-to-habitat covariates were log-transformed using log (distance +1). Prior to all model fitting, we assessed multicollinearity among continuous habitat covariates by calculating variance inflation factors (VIF) and inspecting pairwise Pearson correlations. Multi-collinearity among continuous habitat covariates was assessed and found to be within acceptable limits (VIF < 3).
We implemented the MCMC chains for all arthropod models in JAGS via R2jags with 3 chains, 40,000 iterations per chain, a burn-in of 10,000, and a thinning of 10. Uninformative normal priors were used for fixed-effect regression coefficients, and uniform distributions were used for variance parameters. For all models, we assessed model fit using posterior predictive checks (PPCs). Convergence was evaluated using the Gelman-Rubin statistic. We considered chains to have converged by inspecting the R̂ statistics (R̂ ≤ 1.05), trace plots, and effective sample sizes. We report posterior means/medians and 95% Bayesian credible intervals.
2.8.3 Piecewise structural equation modeling.
To explicitly test whether natural habitat proximity influences insectivorous birds directly or indirectly through arthropod prey availability, we additionally constructed a piecewise structural equation model (pSEM; piecewiseSEM package [77]), using the subset of point counts directly co-located within the sticky trap sampling grids (n = 36 point count surveys). Because aerial insectivores did not have a sufficient sample size at this localized plot scale, the pSEM focused solely on foliage-bark gleaning insectivores. For each prey group, we specified a component linear model for arthropod abundance as a function of local (50 m) and landscape (1000 m) heterogeneity, cover crop diversification index, and distances to water, forest, and riparian habitat. We specified a second component model for foliage-bark gleaner abundance containing the same set of habitat predictors plus arthropod prey abundance. All continuous variables were standardized prior to model fitting. Model fit and d-separation were evaluated using Fisher’s C statistic. Standardized path coefficients, standard errors, p-values, and R2 variance explained were extracted for all direct and indirect pathways.
3 Results
Across all avian point count surveys (n = 447), we detected a total of 5,735 individual birds of 89 species. Of these detections, 3,989 were breeding season insectivores detected during 5 or more point counts, including 975 aerial insectivores and 3,014 foliage-bark gleaners, totaling 32 species (Supplementary Table S1 in S1 File). Across all sticky traps (n = 708), we recorded 160,356 individual arthropods spanning at least 17 orders (Supplementary Table S2 in S1 File). Order composition was dominated by Thysanoptera (thrips), which accounted for 65.1% of individuals, followed by Diptera (16.0%), Hemiptera (7.2%), Hymenoptera (5.0%), and Coleoptera (4.5%); the remaining orders together comprised the remaining ~2.2%.
3.1 Avian insectivores
Foliage-bark gleaners showed clear habitat responses: expected abundance was lower at points farther from riparian habitat and forest, and higher in more locally heterogeneous vineyards within 50 m (Supplementary Table S3 in S1 File, Fig 2). Specifically, comparing points near vs. far from riparian habitat (~15 m vs. ~ 467 m), expected abundance was ~ 56% lower (median; 95% CrI: ~ 33% to ~71% lower). Comparing points near vs. far from forest (~8 m vs. ~ 320 m), expected abundance was also ~56% lower (median; 95% CrI: ~ 30% to ~74% lower). Expected abundance increased with local heterogeneity; a 1 SD increase in local habitat heterogeneity at 50 m was associated with an estimated ~53% increase (median; 95% CrI: ~ 22% to ~93%; Supplementary Table S3 in S1 File). In contrast, we did not observe clear responses to landscape heterogeneity (1000 m), percent water within 1000 m, cover crop diversity, distance to water, or nest box abundance (95% CrIs overlapped 1; Supplementary Table S3 in S1 File). Furthermore, aerial insectivores did not show clear responses to any measured covariates (Supplementary Table S3 in S1 File).
Each panel shows the community-level posterior prediction from the multi-species N-mixture model as a function of (A) distance to forest (m), (B) distance to riparian habitat (m), and (C) local heterogeneity within 50 m (Shannon index). The solid line represents the posterior median predicted abundance; the shaded band represents the 95% credible interval. Points represent observed site-level posterior median abundances.
3.2 Arthropods
Arthropod abundance and diversity exhibited taxon-specific responses to landscape and local habitat structure, alongside seasonal effects (Fig 3; Supplementary Table S4 in S1 File). At the order level, Coleoptera abundance declined with increasing distance to forest (~26.5% decrease per 1 SD increase in forest distance; 95% CrI: ~ 36.8% to ~14.5% lower; Supplementary Table S4 in S1 File), as did Hymenoptera (~19.3% decrease per 1 SD; 95% CrI: ~ 27.6% to ~10.3% lower; Supplementary Table S4 in S1 File) and Hemiptera (~16.3% decrease per 1 SD; 95% CrI: ~ 25.8% to ~5.4% lower; Supplementary Table S4 in S1 File). In contrast, order level Diptera abundance responded primarily to local vegetation structure, increasing with local heterogeneity at 50 m (~11.0% increase per 1 SD; 95% CrI: ~ 2.3% to ~20.5%; Supplementary Table S4 in S1 File). Order level Shannon diversity and Araneae abundance exhibited strong seasonal dynamics (declining early before rising later in the season for Shannon diversity and increasing steadily for Araneae; Supplementary Table S4 in S1 File) but showed no statistically supported spatial habitat associations. Full Bayesian GLMM posterior parameter estimates and 95% Credible Intervals are presented in Supplementary Table S4 in S1 File.
Points are posterior medians and bars represent 95% Bayesian credible intervals (CrIs). Responses are predicted abundance for taxon groups (Hymenoptera, Diptera [large], Diptera [small], Coleoptera [large], Coleoptera [small], Hemiptera [small]) that had significant habitat associations. Covariates shown are cover crop diversity, distance to forest, distance to riparian, distance to water, landscape heterogeneity (1000 m), and local heterogeneity (50 m).
Analyzing prey size classes revealed more habitat associations (Fig 3; Supplementary Table S4 in S1 File). Forest distance was associated with abundance declines for large Coleoptera (~26.6% decrease per 1 SD; 95% CrI: ~ 39.6% to ~10.7% lower), small Coleoptera (~18.3% decrease per 1 SD; 95% CrI: ~ 30.5% to ~3.7% lower), large Diptera (~17.2% decrease per 1 SD; 95% CrI: ~ 30.4% to ~1.7% lower), and small Hemiptera (~15.9% decrease per 1 SD; 95% CrI: ~ 24.9% to ~5.9% lower; Supplementary Table S4 in S1 File). Local habitat heterogeneity at 50 m supported increased abundances for large Coleoptera (~17.9% increase per 1 SD; 95% CrI: ~ 3.9% to ~34.0%), large Diptera (~12.0% increase per 1 SD; 95% CrI: ~ 1.0% to ~24.5%), and small Diptera (~11.4% increase per 1 SD; 95% CrI: ~ 1.8% to ~21.8%; Fig 3; Supplementary Table S4 in S1 File). In addition, large Diptera was the only taxon to respond significantly to broader landscape context, increasing with landscape heterogeneity at 1000 m (~36.5% increase per 1 SD; 95% CrI: ~ 2.1% to ~82.6%) and declining farther from water (~20.1% decrease per 1 SD; 95% CrI: ~ 34.2% to ~2.5% lower; Supplementary Table S4 in S1 File).
3.3 Piecewise structural equation model
Distance to forest (m) showed a negative trend with large Coleoptera prey abundance (β = −0.546, p = 0.103; Fig 4, Supplementary Table S5 in S1 File), though not statistically significant, consistent with higher beetle prey availability near forest edges. Large Coleoptera prey abundance subsequently drove foliage-bark gleaner abundance (β = 0.758, p = 0.019). The direct path from forest distance to bird abundance was non-significant (β = 0.281, p = 0.425), indicating that forest proximity supports foliage-bark gleaners primarily through shifts in large beetle prey (indirect effect β = −0.414, p = 0.019; Fig 4).
(A) Pathway linking forest-proximity to bird abundance through large Coleoptera abundance (Fisher’s C = 2.044, p = 0.360). (B) Pathway linking landscape-heterogeneity within 1,000 m with bird abundance through large Diptera abundance (Fisher’s C = 2.214, p = 0.331). Solid colored arrows indicate statistically supported pathways (p < 0.05), dashed colored arrows indicate non-significant component paths (p > 0.05), and dashed grey lines indicate non-significant direct residual paths. Values adjacent to arrows are standardized path coefficients (β) with statistical significance (*p < 0.05, ns = non-significant). Box values show variance explained (R2) for prey and bird response equations.
In addition to beetle-mediated effects, the pSEM revealed a second significant indirect pathway driven by large Dipteran prey (Fig 4). Landscape heterogeneity within 1000 m significantly increased large Diptera prey abundance (β = 0.633, p = 0.026), which in turn positively influenced foliage-bark gleaner abundance (β = 0.796, p = 0.044). The direct path from 1000 m landscape heterogeneity to foliage-bark gleaner abundance was non-significant (β = −0.679, p = 0.096), suggesting that the primary effect of landscape heterogeneity may occur indirectly through large Dipteran prey availability (indirect effect β = 0.500, p = 0.025), while other mechanisms such as shelter or alternative food resources may also contribute. Across all candidate prey models, direct path models for small Coleoptera, small Diptera, large and small Hemiptera, Hymenoptera, and total Araneae were non-significant (Supplementary Table S5 in S1 File).
4 Discussion
Proximity to natural forest edges emerged as a dominant, consistent predictor supporting both foliage-bark gleaning bird abundance and arthropod prey communities in vineyard agroecosystems. Across sticky trap sampling, forest proximity supported higher abundances of seven focal arthropod groups. Together, forest proximity and local habitat heterogeneity within 50 m of survey locations (i.e., the mix of vineyard cover and adjacent edge/matrix land covers captured in the 50 m buffer) emerged as consistent predictors of insectivorous birds and arthropods in vineyards. More heterogeneous vineyards supported higher abundances of several key arthropod orders and size classes; these same conditions were associated with a greater abundance of foliage-bark gleaning birds. These patterns are consistent with the general expectation that structurally complex environments should expand niche space, diversify foraging substrates, and stabilize temporal resource availability for consumers [13–15,45,78]. In perennial systems, fine-scale structural and compositional variation around crop edges can also moderate microclimate conditions and increase the availability of foraging strata during avian breeding [19,20,79,80]. By evaluating bird superpopulation abundances and order and size class arthropod communities across co-located sampling stations, our study provides a novel empirical demonstration of prey-mediated trophic coupling in agricultural landscapes.
4.1 Habitat effects on birds
Foliage-bark gleaning birds tracked fine-scale habitat structure most clearly. Higher 50 m heterogeneity and proximity to woody and riparian features may increase access to arthropod prey on leaves, twigs, and bark, while also reducing thermal and desiccation stress during the nestling period [19,20,50,81]. This positive association with local landscape heterogeneity complements extensive research highlighting the importance of surrounding semi-natural habitat in agricultural systems [18,24,44,80]. Rather than reflecting total natural cover alone, landscape heterogeneity captures a functional habitat mosaic that provides a complementary suite of daily resources (e.g., diverse prey pools, microclimate shelter) required by mobile avian insectivores across their breeding cycle. Our results also align with a well-documented breeding season diet shift: many passerines primarily provision nestlings with arthropods during breeding, even when adult diets are more variable outside the breeding season [75,82]. In systems for which prey phenology mismatches nestling demand, reproductive performance can decline, emphasizing the importance of habitat configurations that sustain prey availability throughout the nestling period [28,30].
By contrast, aerial insectivores showed weaker and/or unsupported associations with the same habitat covariates. A likely explanation is that aerial insectivores forage across airspaces that often extend beyond point count radii, and they can track short-lived prey swarms shaped by weather and atmospheric conditions [39,83–85]. Reviews also emphasize that aerial insectivore responses are shaped by interacting drivers (prey dynamics, phenology, airspace configuration) and that species differ in how they use open corridors, built structures, and woodland edges [86–88]. Metrics that quantify open-air foraging area, connectivity of flight corridors, and proximity to wetlands may be more informative predictors for aerial insectivores than compositional heterogeneity alone [86,87].
4.2 Habitat effects on arthropods and links to birds
Proximity to natural forest edges was the single most consistent habitat predictor across the arthropod community, supporting higher abundances across seven focal order and size class groups (Coleoptera overall, large, and small; Hemiptera overall and small; Hymenoptera; and large Diptera; Supplementary Table S4 in S1 File). This strong positive association aligns with ecological research emphasizing that natural forest borders and woody vegetation provide critical overwintering refugia, microclimate buffering, and complementary floral and prey resources for agricultural arthropods [66]. However, Diptera showed pronounced seasonal declines, reflecting the strong influence of growing-season phenology and daily weather (temperature, humidity, wind) on adult flight activity and sticky trap capture rates [39,89]. Coleoptera increased with local (50 m) heterogeneity and declined with distance to forest, consistent with many beetle taxa being supported by woody edges and adjacent non-crop habitats that provide refugia, overwintering sites, and complementary resources [66]. Large Hemiptera aligned more with broader habitat context (e.g., forest vs. riparian proximity), suggesting that different orders (and size classes) may track different structural elements or moisture/vegetation gradients around vineyards.
Notably, large Araneae and Hemiptera did not show the same positive local heterogeneity association observed in several other orders. For spiders, habitat and vegetation-management effects can be stratum-dependent (vine vs. ground cover) and may shift in composition more than vine-level abundance, which can reduce detectability in sticky trap data [90]. For Hemiptera, order level patterns may mask divergent responses among species and families within the order. Species-focused vineyard studies of the western grape leafhopper (Erythroneura elegantula), a key pest, show that population density and biological control vary with surrounding habitat context [91,92]. Accordingly, finer taxonomic resolution and targeted sampling are needed to evaluate whether and how habitat structure shapes taxa of commercial interest.
In contrast to forest proximity and 50 m habitat heterogeneity, the cover crop diversification index showed weak main effects across arthropod taxonomic and bird guilds. Our composite index was designed to measure a unified gradient of groundcover complexity, from uniform, low-diversity ground cover to structurally complex, high-diversity cover, while avoiding collinearity among individual between-vine vegetation traits. While evaluating specific cover crop plant species or targeted management practices separately could potentially reveal finer-scale taxonomic responses, our sampling focused on sticky cards deployed at cordon height and foliage-gleaning birds foraging in tree and vine canopies. Cover crop diversification may primarily benefit ground-dwelling arthropods or soil biota, exerting a less direct influence on canopy-stratum insects and canopy-foraging birds [49,90].
Even so, the overall pattern we observed was that local habitat configuration around vineyard sampling locations as well as proximity to forest habitat were associated with both (i) arthropod abundance for multiple taxa and (ii) stronger responses by foliage-bark gleaners, which are tightly linked to vegetation substrates and short travel distances among profitable foraging patches [19,20]. Crucially, our piecewise structural equation model resolved how these habitat-prey associations translate into avian predator responses. While pSEM has been increasingly applied to model ecosystem services in agroecosystems [93,94], our study uniquely disentangles the multi-trophic pathway linking landscape forest proximity to bird guild abundance. Rather than responding through direct spatial attraction to habitat edges, foliage-bark gleaner abundance was associated with two complementary bottom-up, prey-mediated pathways involving large-bodied arthropods. First, an indirect pathway linked forest proximity to greater foliage-bark gleaner abundance via large Coleoptera, where forest distance was the primary habitat driver despite being a non-significant trend in the localized path model. Second, broader landscape heterogeneity within 1000 m supported foliage-bark gleaners indirectly through greater abundance of large Diptera.
Notably, both supported prey-mediated pathways involved large-bodied arthropods (large beetles and large flies), whereas small-bodied prey groups showed no significant indirect coupling. Prey body size plays a fundamental role in shaping avian foraging energetics, as larger-bodied prey items yield a substantially higher biomass and caloric return per capture event, a critical energetic requirement for breeding passerines provisioning rapidly growing nestlings [31,35,95]. Taken together, our results suggest that within-farm habitat structure, especially near woody forest edges, shapes prey communities in ways that most directly benefit substrate-oriented foragers. This may help explain why foliage-bark gleaners tracked local habitat most clearly, even when broader-scale effects were weaker. It also provides a novel empirical demonstration of prey-mediated trophic coupling in working vineyard landscapes. These dual pathways highlight how different habitat features shape avian food availability, with forest edges supporting large beetles and landscape scale heterogeneity supporting large flies.
4.3 Limitations and future directions
Several limitations define our scope of inference and point to priority next steps. First, while our multi-species N-mixture models evaluated broader landscape-scale bird responses across the full point count surveys (n = 447), evaluating direct predator-prey linkages required restricting structural equation modeling to co-located sampling locations (n = 36). By leveraging this design, our pSEM explicitly confirmed the indirect prey-mediation pathway, demonstrating that large flies and beetles elevate foliage-bark gleaner abundance. However, our analyses were restricted to foliage-bark gleaners due to sample size constraints for aerial insectivores at the localized plot scale. Nevertheless, sticky traps primarily capture flying arthropods at specific canopy strata, which may under-represent some arthropod taxa groups that may also influence bird abundance.
Mechanistic follow-up studies could include manipulative habitat experiments (e.g., edge plantings or woody-structure additions; mowing/tillage timing treatments), higher-resolution microclimate measurements, and/or repeated prey sampling throughout the nestling period. Integrating diet composition analysis (e.g., via DNA metabarcoding) with repeated arthropod sampling could allow direct tests of whether habitat-driven shifts in prey availability translate into differences in prey delivery and breeding season bird use [50,76].
4.4 Conclusions
Collectively, our results suggest that maintaining or enhancing local and landscape habitat heterogeneity, including vegetation along vineyard edges and in adjacent non-crop elements, could help support both arthropod communities and foliage-bark gleaning birds that have the potential to provide biological pest control in these systems. Potential management actions might include retaining or planting shrubby/woody edge structure, protecting and widening riparian buffers, and maintaining diversified groundcover within vineyards where feasible [19,20,96]. Vineyard managers may be concerned that these practices could attract grape-damaging birds; however, prior work in this system found that proximity to wooded uncultivated habitats did not increase grape damage [97]. Reducing insecticide inputs during peak nestling provisioning windows, particularly near edges and riparian zones, could also help sustain prey availability for breeding insectivores and beneficial arthropods, potentially contributing to pest suppression within vineyards [44,98–100]. For aerial insectivores, habitat planning may need to explicitly consider water features, open-air corridor configuration, and cross-property connectivity to better match the spatial scale at which these birds forage [86–88]. In conclusion, our findings demonstrate that natural habitat in vineyard landscapes supports insectivorous birds by enhancing large arthropod prey. Conservation strategies that combine local habitat management with broader landscape diversity can work to sustain both beneficial arthropod communities and contribute to bird-mediated pest control in vineyard ecosystems.
Acknowledgments
We would like to thank the 20 participating vineyards for allowing us access to conduct research on their farms. We especially appreciate the coordinating vineyard and winery managers who provided us with invaluable information related to viticulture that was helpful for developing our study design and sampling methods. We also acknowledge the many contributions of colleagues and students who assisted during fieldwork and laboratory activities for this project.
References
- 1. Díaz S, Settele J, Brondízio ES, Ngo HT, Agard J, Arneth A, et al. Pervasive human-driven decline of life on Earth points to the need for transformative change. Science. 2019;366(6471):eaax3100. pmid:31831642
- 2. Dirzo R, Young HS, Galetti M, Ceballos G, Isaac NJB, Collen B. Defaunation in the Anthropocene. Science. 2014;345(6195):401–6. pmid:25061202
- 3. Foley JA, Defries R, Asner GP, Barford C, Bonan G, Carpenter SR, et al. Global consequences of land use. Science. 2005;309(5734):570–4. pmid:16040698
- 4. Newbold T, Hudson LN, Hill SLL, Contu S, Lysenko I, Senior RA, et al. Global effects of land use on local terrestrial biodiversity. Nature. 2015;520(7545):45–50. pmid:25832402
- 5. Rosenberg KV, Dokter AM, Blancher PJ, Sauer JR, Smith AC, Smith PA, et al. Decline of the North American avifauna. Science. 2019;366(6461):120–4. pmid:31604313
- 6. Steffen W, Broadgate W, Deutsch L, Gaffney O, Ludwig C. The trajectory of the Anthropocene: The Great Acceleration. Anthr Rev. 2015;2(1):81–98.
- 7. Wagner DL. Insect Declines in the Anthropocene. Annu Rev Entomol. 2020;65:457–80. pmid:31610138
- 8. Lindell C, Eaton RA, Howard PH, Roels SM, Shave ME. Enhancing agricultural landscapes to increase crop pest reduction by vertebrates. Agric Ecosyst Environ. 2018;257:1–11.
- 9. Lourenço R, Pereira PF, Oliveira A, Ribeiro-Silva J, Figueiredo D, Rabaça JE, et al. Effect of vineyard characteristics on the functional diversity of insectivorous birds as indicator of potential biocontrol services. Ecol Indic. 2021;122:107251.
- 10. Monteagudo N, Benayas JMR, Andivia E, Rebollo S. Avian regulation of crop and forest pests, a meta-analysis. Pest Manag Sci. 2023;79(7):2380–9. pmid:36810937
- 11. Cardinale BJ, Duffy JE, Gonzalez A, Hooper DU, Perrings C, Venail P, et al. Biodiversity loss and its impact on humanity. Nature. 2012;486(7401):59–67. pmid:22678280
- 12. Foley JA, Ramankutty N, Brauman KA, Cassidy ES, Gerber JS, Johnston M, et al. Solutions for a cultivated planet. Nature. 2011;478(7369):337–42. pmid:21993620
- 13. Isbell F, Cowles J, Dee LE, Loreau M, Reich PB, Gonzalez A, et al. Quantifying effects of biodiversity on ecosystem functioning across times and places. Ecol Lett. 2018;21(6):763–78. pmid:29493062
- 14. Loreau M, de Mazancourt C. Biodiversity and ecosystem stability: a synthesis of underlying mechanisms. Ecol Lett. 2013;16 Suppl 1:106–15. pmid:23346947
- 15. Tilman D, Isbell F, Cowles JM. Biodiversity and Ecosystem Functioning. Annu Rev Ecol Evol Syst. 2014;45(1):471–93.
- 16. Beaumelle L, Auriol A, Grasset M, Pavy A, Thiéry D, Rusch A. Benefits of increased cover crop diversity for predators and biological pest control depend on the landscape context. Ecol Sol Evidence. 2021;2(3).
- 17. Puig-Montserrat X, Stefanescu C, Torre I, Palet J, Fàbregas E, Dantart J, et al. Effects of organic and conventional crop management on vineyard biodiversity. Agric Ecosyst Environ. 2017;243:19–26.
- 18. Winter S, Paredes D, Beaumelle L, Alcalá Herrera R, Chen Y, Hoffmann C, et al. Extensive vineyard management and semi-natural habitats increase biodiversity and ecosystem services: insights from a global meta-analysis. J Environ Manage. 2025;395:128029. pmid:41259979
- 19. 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. 2017;80:275–85.
- 20. Barbaro L, Assandri G, Brambilla M, Castagneyrol B, Froidevaux J, Giffard B, et al. Organic management and landscape heterogeneity combine to sustain multifunctional bird communities in European vineyards. J Appl Ecol. 2021;58(6):1261–71.
- 21. Jedlicka JA, Greenberg R, Letourneau DK. Avian conservation practices strengthen ecosystem services in California vineyards. PLoS One. 2011;6(11):e27347. pmid:22096555
- 22. Shackelford G, Steward PR, Benton TG, Kunin WE, Potts SG, Biesmeijer JC, et al. Comparison of pollinators and natural enemies: a meta-analysis of landscape and local effects on abundance and richness in crops. Biol Rev Camb Philos Soc. 2013;88(4):1002–21. pmid:23578337
- 23. Tscharntke T, Sekercioglu CH, Dietsch TV, Sodhi NS, Hoehn P, Tylianakis JM. Landscape constraints on functional diversity of birds and insects in tropical agroecosystems. Ecology. 2008;89(4):944–51. pmid:18481519
- 24. Tscharntke T, Klein AM, Kruess A, Steffan‐Dewenter I, Thies C. Landscape perspectives on agricultural intensification and biodiversity – ecosystem service management. Ecol Lett. 2005;8(8):857–74.
- 25. Winter S, Bauer T, Strauss P, Kratschmer S, Paredes D, Popescu D, et al. Effects of vegetation management intensity on biodiversity and ecosystem services in vineyards: A meta-analysis. J Appl Ecol. 2018;55(5):2484–95. pmid:30147143
- 26. De Graaf RM, Tilghman NG, Anderson SH. Foraging guilds of North American birds. Environ Manag. 1985;9(6):493–536.
- 27. Martin TE. Food as a limit on breeding birds: a life-history perspective. Annu Rev Ecol Syst. 1987;18(1):453–87.
- 28. Both C, Bouwhuis S, Lessells CM, Visser ME. Climate change and population declines in a long-distance migratory bird. Nature. 2006;441(7089):81–3. pmid:16672969
- 29. Paquette SR, Garant D, Pelletier F, Bélisle M. Seasonal patterns in tree swallow prey (Diptera) abundance are affected by agricultural intensification. Ecol Appl. 2013;23(1):122–33. pmid:23495641
- 30. Visser ME, Holleman LJM, Gienapp P. Shifts in caterpillar biomass phenology due to climate change and its impact on the breeding biology of an insectivorous bird. Oecologia. 2006;147(1):164–72. pmid:16328547
- 31. Robinson SK, Holmes RT. Foraging Behavior of Forest Birds: The Relationships Among Search Tactics, Diet, and Habitat Structure. Ecology. 1982;63(6):1918.
- 32. Greenberg R. Insectivorous Migratory Birds in Tropical Ecosystems: The Breeding Currency Hypothesis. J Avian Biol. 1995;26(3):260.
- 33. Holmes RT, Schultz JC. Food availability for forest birds: effects of prey distribution and abundance on bird foraging. Can J Zool. 1988;66(3):720–8.
- 34. Janes SW. Variation in the species composition and mean body size of an avian foliage-gleaning guild along an elevational gradient: correlation with arthropod body size. Oecologia. 1994;98(3–4):369–78. pmid:28313914
- 35. Naef‐Daenzer L, Naef‐Daenzer B, Nager RG. Prey selection and foraging performance of breeding Great Tits Parus major in relation to food availability. J Avian Biol. 2000;31(2):206–14.
- 36. Adamík P, Kornan M. Foraging ecology of two bark foraging passerine birds in an old-growth temperate forest. Ornis Fennica. 2004;81(1):13–22.
- 37. Flower CE, Long LC, Knight KS, Rebbeck J, Brown JS, Gonzalez-Meler MA, et al. Native bark-foraging birds preferentially forage in infected ash (Fraxinus spp.) and prove effective predators of the invasive emerald ash borer (Agrilus planipennis Fairmaire). For Ecol Manag. 2014;313:300–6.
- 38. Franzreb KE. Foraging ecology of brown creepers in a mixed-coniferous forest. J Field Ornithol. 1985;56(1):9–16.
- 39. McCarty JP, Winkler DW. Foraging Ecology and Diet Selectivity of Tree Swallows Feeding Nestlings. Condor. 1999;101(2):246–54.
- 40. Albrecht M, Kleijn D, Williams NM, Tschumi M, Blaauw BR, Bommarco R, et al. The effectiveness of flower strips and hedgerows on pest control, pollination services and crop yield: a quantitative synthesis. Ecol Lett. 2020;23(10):1488–98. pmid:32808477
- 41. Chaplin-Kramer R, de Valpine P, Mills NJ, Kremen C. Detecting pest control services across spatial and temporal scales. Agric Ecosyst Environ. 2013;181:206–12.
- 42. Chaplin-Kramer R, Kremen C. Pest control experiments show benefits of complexity at landscape and local scales. Ecol Appl. 2012;22(7):1936–48. pmid:23210310
- 43. Dainese M, Martin EA, Aizen MA, Albrecht M, Bartomeus I, Bommarco R, et al. A global synthesis reveals biodiversity-mediated benefits for crop production. Sci Adv. 2019;5(10):eaax0121. pmid:31663019
- 44. Karp DS, Chaplin-Kramer R, Meehan TD, Martin EA, DeClerck F, Grab H, et al. Crop pests and predators exhibit inconsistent responses to surrounding landscape composition. Proc Natl Acad Sci U S A. 2018;115(33):E7863–70. pmid:30072434
- 45. Letourneau DK, Armbrecht I, Rivera BS, Lerma JM, Carmona EJ, Daza MC, et al. Does plant diversity benefit agroecosystems? A synthetic review. Ecol Appl. 2011;21(1):9–21. pmid:21516884
- 46. Geldenhuys M, Gaigher R, Pryke JS, Samways MJ. Diverse herbaceous cover crops promote vineyard arthropod diversity across different management regimes. Agric Ecosyst Environ. 2021;307:107222.
- 47. Sáenz-Romo MG, Veas-Bernal A, Martínez-García H, Campos-Herrera R, Ibáñez-Pascual S, Martínez-Villar E, et al. Ground cover management in a Mediterranean vineyard: Impact on insect abundance and diversity. Agric Ecosyst Environ. 2019;283:106571.
- 48. Litovska I, van der Plas F, Kleijn D. Arthropod abundance is most strongly driven by crop and semi-natural habitat type rather than management in an intensive agricultural landscape in the Netherlands. Agric Ecosyst Environ. 2025;378:109298.
- 49. Zielonka NB, Shutt JD, Butler SJ, Dicks LV. Management practices, and not surrounding habitats, drive bird and arthropod biodiversity within vineyards. Agric Ecosyst Environ. 2024;367:108982.
- 50. Jedlicka JA, Vo A-TE, Almeida RPP. Molecular scatology and high-throughput sequencing reveal predominately herbivorous insects in the diets of adult and nestling Western Bluebirds (Sialia mexicana) in California vineyards. Auk. 2017;134(1):116–27.
- 51. Cooke R, Grames EM, Ollerton J, Antão LH, Bell JR, Bulbert MW, et al. The importance of insect biodiversity for vertebrates. Nat Rev Biodivers. 2026;1–17.
- 52. Spiller KJ, Dettmers R. Evidence for multiple drivers of aerial insectivore declines in North America. Condor. 2019;121(2).
- 53. Nebel S, Mills A, McCracken JD, Taylor PD. Declines of aerial insectivores in North America follow a geographic gradient Présence d’un gradient géographique dans le déclin des insectivores aériens. Avian Conserv Ecol. 2010;5(2):1.
- 54. Hallmann CA, Sorg M, Jongejans E, Siepel H, Hofland N, Schwan H, et al. More than 75 percent decline over 27 years in total flying insect biomass in protected areas. PLoS One. 2017;12(10):e0185809. pmid:29045418
- 55. Pywell RF, Heard MS, Woodcock BA, Hinsley S, Ridding L, Nowakowski M, et al. Wildlife-friendly farming increases crop yield: evidence for ecological intensification. Proc Biol Sci. 2015;282(1816):20151740. pmid:26423846
- 56. Suscol Intertribal Council. Napa Valley First People’s History [Internet]. 2025. Available from: https://suscolcouncil.org/about-us/history/
- 57.
Grossinger R, Askevold R. Napa Valley historical ecology atlas: exploring a hidden landscape of transformation and resilience. Univ of California Press; 2012.
- 58. Wendt CA, Johnson MD. Multi-scale analysis of barn owl nest box selection on Napa Valley vineyards. Agric Ecosyst Environ. 2017;247:75–83.
- 59.
Corro L. A look at land cover classification methods in northern california with the use of high spatial resolution geospatial data [Master’s thesis]. Arcata (CA): California State Polytechnic University-Humboldt; 2021.
- 60.
Carlino J. Effects of habitat selection and individual quality on the reproductive success of barn owls (Tyto furcata) [Master’s thesis]. Arcata (CA): California State Polytechnic University-Humboldt; 2024.
- 61.
Ralph C, Droege S, Sauer J. Managing and monitoring birds using point counts. Gen Tech Rep PSW-GTR-149. Albany (CA): Pacific Southwest Research Station, Forest Service, US Department of Agriculture; 1995.
- 62. Murphy SJ, Bellvé AM, Miyajima RJ, Sebunia NA, Lynch MM, Jetz W. SAviTraits 1.0: Seasonally varying dietary attributes for birds. Glob Ecol Biogeogr. 2023;32(10):1690–8.
- 63. Tobias JA, Sheard C, Pigot AL, Devenish AJM, Yang J, Sayol F, et al. AVONET: morphological, ecological and geographical data for all birds. Ecol Lett. 2022;25(3):581–97. pmid:35199922
- 64. Daugherty MP, Gruber BR, Almeida RPP, Anderson MM, Cooper ML, Rasmussen YD, et al. Testing the Efficacy of Barrier Plantings for Limiting Sharpshooter Spread. Am J Enol Vitic. 2011;63(1):139–43.
- 65. Kross SM. Insect pest control and bird damage as a function of distance from riparian habitat in a California vineyard. Proc Vertebr Pest Conf. 2016;27:27–31.
- 66. Tscharntke T, Tylianakis JM, Rand TA, Didham RK, Fahrig L, Batáry P, et al. Landscape moderation of biodiversity patterns and processes - eight hypotheses. Biol Rev Camb Philos Soc. 2012;87(3):661–85. pmid:22272640
- 67. Ke A, Cedeño E, Karubian J, Carrasco L, Castillo F, Gonzales N, et al. Cavity‐nesting birds are limited by nesting habitat in Neotropical agricultural landscapes. Biotropica. 2023;55(5):1045–57.
- 68. Olmos-Moya N, Díaz-Siefer P, Pozo RA, Fontúrbel FE, Lavandero B, Abades S, et al. The use of cavity-nesting wild birds as agents of biological control in vineyards of Central Chile. Agric Ecosyst Environ. 2022;334:107975.
- 69. Andow D. Vegetational diversity arthropod population. Annu Rev Entomol. 1991;36:561–6.
- 70.
R Core Team. R [Internet]. Vienna, Austria: R Foundation for Statistical Computing; 2025. Available from: https://www.R-project.org/
- 71. Royle JA, Nichols JD, Kéry M. Modelling occurrence and abundance of species when detection is imperfect. Oikos. 2005;110(2):353–9.
- 72. Fogarty FA, Fleishman E. Bias in estimated breeding-bird abundance from closure-assumption violations. Ecol Indic. 2021;131:108170.
- 73.
Kéry M, Royle J. AHM1—Modeling distribution, abundance and species richness using R and BUGS. Volume 1: Prelude and Static Models. 2016.
- 74. Su YS, Yajima M. R2jags: Using R to Run ‘JAGS’. R package version 0.07-1. 2021. Available from: https://CRAN.R-project.org/package=R2jags
- 75. Höhn D, Masello JF, Kümmel MN, Griep S, Goesmann A, Quillfeldt P. Nestling Diet of Two Sympatric Insectivorous Passerines in Different Habitats—A Metabarcoding Study. Birds. 2024;5(1):67–89.
- 76. Zurdo J, Reverter M, Barrero A, Bustillo-de la Rosa D, Gómez-Catasús J, Pérez-Granados C, et al. Prey choice in insectivorous steppe passerines: New insights from DNA metabarcoding. Glob Ecol Conserv. 2023;48:e02738.
- 77. Lefcheck JS. Piecewise SEM: Piecewise structural equation modelling in R for ecology, evolution, and systematics. Methods Ecol Evol. 2016;7(5):573–9.
- 78. Cardinale BJ, Srivastava DS, Duffy JE, Wright JP, Downing AL, Sankaran M, et al. Effects of biodiversity on the functioning of trophic groups and ecosystems. Nature. 2006;443(7114):989–92. pmid:17066035
- 79. Edo M, Entling MH, Rösch V. Agroforestry supports high bird diversity in European farmland. Agron Sustain Dev. 2023;44(1):8.
- 80. Rösch V, Chavez F, Krey L, Möth S, Petrović B, Richart-Cervera S, et al. Semi-natural habitats are key to breeding bird diversity in intensified vineyard landscapes across Europe. Basic Appl Ecol. 2024;81:66–74.
- 81. Muñoz-Sáez A, Heaton EE, Reynolds M, Merenlender AM. Agricultural adapters from the vineyard landscape impact native oak woodland birds. Agric Ecosyst Environ. 2020;300:106960.
- 82. Pinkowski BC. Feeding of nestling and fledgling Eastern Bluebirds. Wilson Bull. 1978;90(1):84–98.
- 83. Mohamed A, Taylor GK, Watkins S, Windsor SP. Opportunistic soaring by birds suggests new opportunities for atmospheric energy harvesting by flying robots. J R Soc Interface. 2022;19(196):20220671. pmid:36415974
- 84. Shi X, Schmid B, Tschanz P, Segelbacher G, Liechti F. Seasonal Trends in Movement Patterns of Birds and Insects Aloft Simultaneously Recorded by Radar. Remote Sensing. 2021;13(9):1839.
- 85. Warrick DR, Hedrick TL, Biewener AA, Crandell KE, Tobalske BW. Foraging at the edge of the world: low-altitude, high-speed manoeuvering in barn swallows. Philos Trans R Soc Lond B Biol Sci. 2016;371(1704):20150391. pmid:27528781
- 86. Diaz Bohorquez AM, Morrissey C, Hobson K. Tracing aquatic resource contributions to Tree and Barn Swallow nestling diets across a cropland and wetland gradient. ACE. 2025;20(1).
- 87. Kardynal KJ, Imlay TL. Ecology and conservation of avian aerial insectivores. Can J Zool. 2022;100(10):615–6.
- 88. Bianchi FJJA, Booij CJH, Tscharntke T. Sustainable pest regulation in agricultural landscapes: a review on landscape composition, biodiversity and natural pest control. Proc Biol Sci. 2006;273(1595):1715–27. pmid:16790403
- 89. Taylor LR. Analysis of the Effect of Temperature on Insects in Flight. J Anim Ecol. 1963;32(1):99–117.
- 90. Costello MJ, Daane KM. Influence of ground cover on spider populations in a table grape vineyard. Ecol Entomol. 1998;23(1):33–40.
- 91. Wilson H, Miles AF, Daane KM, Altieri MA. Landscape Diversity and Crop Vigor Influence Biological Control of the Western Grape Leafhopper (E. elegantula Osborn) in Vineyards. PLoS One. 2015;10(11):e0141752. pmid:26555074
- 92. Wilson H, Miles AF, Daane KM, Altieri MA. Vineyard proximity to riparian habitat influences Western grape leafhopper (Erythroneura elegantula Osborn) populations. Agric Ecosyst Environ. 2015;211:43–50.
- 93. Maas B, Clough Y, Tscharntke T. Bats and birds increase crop yield in tropical agroforestry landscapes. Ecol Lett. 2013;16(12):1480–7. pmid:24131776
- 94. Barbaro L, Rusch A, Muiruri EW, Gravellier B, Thiery D, Castagneyrol B. Avian pest control in vineyards is driven by interactions between bird functional diversity and landscape heterogeneity. J Appl Ecol. 2017;54(2):500–8.
- 95. Ricklefs R. The energetics of reproduction in birds. Avian Energetics. 1974;15:152–297.
- 96. Judt C, Guzmán G, Gómez JA, Cabezas JM, Entrenas JA, Winter S, et al. Diverging Effects of Landscape Factors and Inter-Row Management on the Abundance of Beneficial and Herbivorous Arthropods in Andalusian Vineyards (Spain). Insects. 2019;10(10):320. pmid:31561623
- 97.
Adler K, Martinico BL, MacDonald E, Kross SM, Karp DS, Turner A, et al. Unraveling the influence of uncultivated habitat on avian grape damage in a viticultural landscape. In: Proceedings of the Vertebrate Pest Conference. 2024. 31 p.
- 98. Bouvier J-C, Delattre T, Boivin T, Musseau R, Thomas C, Lavigne C. Great tits nesting in apple orchards preferentially forage in organic but not conventional orchards and in hedgerows. Agric Ecosyst Environ. 2022;337:108074.
- 99. Reiff JM, Kolb S, Entling MH, Herndl T, Möth S, Walzer A, et al. Organic Farming and Cover-Crop Management Reduce Pest Predation in Austrian Vineyards. Insects. 2021;12(3):220. pmid:33806420
- 100. Roubos CR, Rodriguez-Saona C, Isaacs R. Mitigating the effects of insecticides on arthropod biological control at field and landscape scales. Biol Control. 2014;75:28–38.

관련 뉴스