Skip to main content

Data from: Shifts in species interactions and farming contexts mediate net effects of birds in agroecosystems

Cite this dataset

Olimpi, Elissa; Garcia, Karina; Karp, Daniel; Gonthier, David (2020). Data from: Shifts in species interactions and farming contexts mediate net effects of birds in agroecosystems [Dataset]. Dryad.


Some birds are viewed as pests and vectors of foodborne pathogens in farmlands, yet birds also benefit growers by consuming pests. While many growers seek to prevent birds from accessing their farms, few studies have attempted to quantify the net effects of bird services and disservices, let alone how net effects shift across farm management strategies. We quantified the net effect of birds on crop production across 20 California strawberry (Fragaria x ananassa) farms that varied in local management practices and landscape context. We surveyed farms for berry damage and bird droppings (as potential sources of pathogens), and implemented a large‐scale exclusion experiment to quantify the impact of birds on production. We found that birds had only a slightly negative overall impact on strawberry production, reducing economic value by 3.6%. Direct bird damage and intraguild predation contributed equally to this net effect, underscoring the importance of indirect trophic interactions that may be less apparent to growers. In simple landscapes (e.g., low proportions of surrounding seminatural habitat), birds provided pest control in the interiors of farm fields, and costs from bird damage to crops peaked at field edges. In complex landscapes (e.g., high proportions of seminatural habitat), birds were more likely to disrupt pest control by feeding as intraguild predators. Nonetheless, seminatural habitat dampened bird services and disservices, and our models predicted that removing habitat around farm fields would increase costs from bird damage to crops by up to 76%. Fecal contamination of crops was extremely rare (0.01%). However, both fecal contamination and bird damage did increase on farms with higher densities of fencing and wires, where birds often perch. Our results demonstrate that maintaining seminatural habitat around farms may enhance bird diversity and mitigate bird damage without increasing food‐safety risks. We also show that the net effects of birds depend on farming context and vary in complex ways in relation to locations within a farm, local farm attributes, and the surrounding landscape. This context‐specific variation must be considered in order to optimize the management of wild birds in agroecosystems.


Bird exclosure experiment and berry damage and fecal transect surveys on California strawberry farms. This dataset contains information on 1) berry damage and invertebrate sampling from a bird exclosure experiment (N=15), and 2) berry damage and bird fecal contamination from transect surveys (N=20) on organic strawberry farms in the California Central Coast in 2018. We excluded birds from strawberry plants with mesh-net cages and compared berry damage and insect communities between exclosure (no birds) and control (bird accessible) plots. We erected three paired exclosure and control plots on each farm for one month each during the spring (April-May, when strawberry yields were low) and summer (June-July, during peak strawberry production). Each week, we harvested berries from treatment plots and scored all ripe berries for damage. At the beginning and end of each round of the exclosure experiment, we vacuum-sampled invertebrates from eight plants within each plot. We surveyed berry damage and bird fecal contamination along three parallel, 20m transects in strawberry rows on each farm. Berry damage surveys were limited to 20 plants per transect, sampling every 5 plants. We located exclosures and transects at the farm edge with the most seminatural habitat, as far from a farm edge as a possible, and halfway in between.

Usage notes

Each csv file has an accompanying txt file defining the column headings.

  1. All ripe berries were harvested from exclosure and control treatment plots at the beginning of each round of the exclosure experiment. Initial berry harvests during exclosure set-up are not included in this dataset.
  2. One paired exclosure treatment plot was removed from all exclosure analyses:
    1. Farm12, exclosure pair 1, early season – shading from vegetation made treatment plots not comparable
  3. Two additional paired exclosure treatment plots were removed from all exclosure berry damage analyses:
    1. Farm17, exclosure pair 3, late season- berries from control plots harvested by farmworkers
    2. Farm18, exclosure pair 1, late season- berries from control plots harvested by farmworkers


United States Department of Agriculture, Award: 2017-67019-26293