Data from: Rainforest fragmentation decreases the robustness of plant-frugivore interaction networks
Data files
Aug 04, 2025 version files 67.12 KB
-
Metadata_TABLE_S5.xlsx
8.93 KB
-
README.md
1.66 KB
-
TABLE_S5.csv
56.54 KB
Abstract
Tropical rainforests are biodiversity hotspots that provide a variety of ecosystem functions and services. Seed dispersal by fruit-eating birds is an important ecosystem process in the regeneration of tropical rainforests, which is increasingly threatened by widespread deforestation. In particular, the expansion of agricultural land often leads to forest fragmentation, which can have a negative impact on the interactions between plants and frugivores and thus on seed dispersal. However, little is known about how forest fragmentation affects the structure and robustness of plant-frugivore interaction networks. Here, we examined the effects of forest fragmentation on species richness of frugivorous birds interacting with focal tree species, and the structure and robustness of plant-frugivore interaction networks in the tropical lowland forests of northern Costa Rica. Species richness of frugivorous birds at the forest edges increased with fragment size and forest cover in the surrounding landscape, as well as with local fruit availability. Modularity and robustness of plant-frugivore networks increased with enhanced fragment size and forest cover, while network specialization (H2´) increased only with greater forest cover. Additionally, the three common tanager species (Ramphocelus passerinii, Thraupis palmarum, and Thraupis episcopus) were identified as key bird species for network functioning by promoting among-module and within-module connectivity. Conservation measures should therefore not only focus on threatened specialist species, but more on the key species that enhance network structure and consequently increase the robustness of these trophic interaction networks. Ultimately, our study demonstrates that tropical forest fragmentation simplifies network structure, making these interactions more vulnerable to anthropogenic disturbances.
Dataset DOI: 10.5061/dryad.fttdz0957
Description of the data and file structure
Between 5 October and 3 December 2023, we carried out 252 hours of active bird–fruit interaction observations (05:30–09:30 AM) along fixed 300 m forest-edge transects in six old-growth rainforest fragments in northern Costa Rica. On three randomized days per fragment, three observers recorded the following data fields for every interaction (TABLE_S5.csv): Fragment ID; Date; Round; TreeID; Time; Tree species; Bird species; Bird Individuals Interacting; Tree Surrounding; Fruit amount. The Metadata_TABLE_S5.xlsx data file includes the description of all the variables.
To ensure consistency, all observations of bird-fruit interactions were conducted along a fixed 300 m transect at the forest edge of each fragment, based on accessibility and availability of fruit-bearing trees. Interaction frequency was defined as the number of individual birds pecking or swallowing a fruit. If a bird remained in the tree for an extended period and consumed multiple fruits, this was still counted as a single interaction. To be considered a new interaction by the same bird species, the individual had to leave the tree and return after at least one minute. In addition, for each observed fruit-bearing tree, the number of ripe fruits were estimated at the time of observation and additionally the tree surrounding (single standing = no other tree in 5 m radius; edge tree = embedded in stand of numerous trees) was classified.
