Data from: Trait-based species selection for restoration: A case study from tropical landscapes dominated by bracken
Data files
Nov 05, 2025 version files 50.47 KB
-
Gallegos_abundance.csv
2.29 KB
-
Gallegos_dataset.csv
42.62 KB
-
README.md
5.56 KB
Abstract
Deforestation leads to non-forested areas that often need active ecological restoration to promote forest succession. Here, we developed a trait-based approach to guide the design of restoration projects and tested our approach in tropical areas deforested by fires and dominated by bracken. Deforestation caused by human-induced fires creates complex conservation problems in tropical areas. After fire, montane tropical deforested areas are often dominated by bracken fern (Pteridium spp.), which prevents the establishment of many species of trees. This means that effective ecological restoration strategies are needed for forest recovery. We explored whether a trait-based approach could be used to guide the selection of tree species in active restoration projects in deforested areas. We first tested whether traits could be used to predict which species can overcome bracken-associated filters. We then tested the relationship between the trait-based predictions for each species and their abundance in a reference old-growth forest. To identify key traits, we conducted a seed addition experiment in bracken-dominated areas that had developed 5-15 years after a fire, using 23 tree species common in the Bolivian montane forests, and monitored seedling establishment, survival, and growth for 36 months. We then related seedling performance to the functional traits of adults and seedlings. We found that six functional traits improved tree seedling performance in bracken-dominated areas, and that only one-third of the species evaluated exceeded a 25% probability of recruitment. Most of the favored traits correspond to conservative strategies. The species dispersed by birds and with large seeds were the most likely to overcome the filters created by bracken. Based on the trait predictions, we found that the species predicted to overcome bracken are not common in old-growth forests. Our results can be used by stakeholders to select the species best suited for active restoration projects in bracken-dominated areas throughout the tropical region. Furthermore, our trait-based approach, which considers seedling performance in disturbed areas, can serve as a guide for species selection in restoration programs in other disturbed systems.
Dataset DOI: 10.5061/dryad.8931zcs4q
Description of the data and file structure
The dataset contains all data required to replicate analyses in Gallegos et al. 2025. We established a seed addition experiment at eight sites in the vicinity of Chulumani village, Sud Yungas province, La Paz, Bolivia. The experiment was established in burned areas from the tropical montane forest dominated by the bracken fern Pteridium esculentum, at 100 m from the forest edge. In each site, we planted 23 tree species in different quantities according to their seed size. Seeds were collected from at least ten individuals per species. We monitored seedling establishment, survival, and growth after 6, 12, 24, and 36 months. We measured the following functional traits in five adult individuals of each species: maximum tree height (HM, in m), leaf area (LA, in cm2), specific leaf area (SLA, in cm2/g), leaf dry matter content (LDMC, in mg/g), stem specific density (SSDc, in mg/mm3, corrected for secondary branches multiplying by 1.411), seed size (i.e., seed length in mm), seed mass (in g) and dispersal vector (birds, bats or wind). The same traits were obtained for the species from the community from a database taken in the same study area (Missouri Botanical Garden – Bolivia Program, 2024). We related the functional traits to seedling performance of the planted species using GLMMs, and selected the best models for seedling establishment, survival, and relative growth rate (RGR), respectively. With these models, we predicted seedling performance of the species from the community by calculating their active restoration value (ARV). Finally, we related the ARV of each species to its abundance in old-growth forests from the study area (Lippok et al., 2014).
Time of data collection:
Seed addition: November 2019
Last monitoring: December 2022
Geographic location of data collection:
The study was carried out on the eastern slope of the Andes, in the vicinity of Chulumani village, Sud Yungas, La Paz, Bolivia (16°24’ S, 67°31’ W). The study area is located in the tropical montane humid forest between 1,850 and 2,450 m a.s.l.
Files and variables
File: Gallegos_dataset.csv
Description: Seedling performance of the planted species and functional traits of the planted species and of the species from the community.
Variables
- SITE: (= PLOT): Acronym of eight study sites or plots where we established the seed addition experiment.
- DATABASE: planted (species that were planted); community (species from the community, non-planted).
- SPECIES: Acronym from the species.
- SPECIES_NAME: Full species name.
- PLANTEDseeds: Number of sown seeds per site.
- ESTABLtot: Number of seedlings established within the first 12 months out of the number of seeds sown per site.
- NONESTABL: Number of seeds that did not establish within the first 12 months out of the number of seeds sown per site.
- SURFtot: number of seedlings that survived until the end of the experiment (36 months) relative to the number of established seedlings.
- NONSURF: number of seedlings that did not survive until the end of the experiment (36 months) relative to the number of established seedlings.
- RGR: Relative growth rate (RGR) using the formula (ln(ht+1) - ln(ht))/t, where h is the individual height in cm at time t and t+1, and t is the time elapsed in days between measurements.
- Dispersal: Dispersal vector (birds, bats, wind).
- HM: maximum tree height (m).
- LA: leaf area (cm2).
- SLA: specific leaf area (cm2/g).
- LDMC: leaf dry matter content (mg/g).
- BarkT: bark thickness (mm).
- SSDc: stem specific density (mg/mm3, corrected for secondary branches multiplying by 1.411).
- SS: seed length (mm).
- SM: seed mass (g).
File: Gallegos_abundance.csv
Description: Active restoration value of each species from the community and its abundance in old-growth forests from the study area.
Variables
- SPECIES_NAME: Full species name.
- ARV: Active restoration value. We predicted the establishment, survival, and RGR for each species in the community from our models, using the coefficients from the best models (Table S3 in Supplementary Information). We then multiplied the predicted values to obtain the 'active restoration value' (i.e., recruitment probability) for each species.
- ARVP: Active restoration value in percentage. We assigned 100% of the active restoration value to the first species and calculated a relative active restoration value (%) for all other species.
- ABUNDANCE: Abundance of each species in old-growth forests from our study area, based on the number of individuals per species found in 12 plots of 400 m2 each (extracted from Lippok et al., 2014).
References
- Lippok, D., Beck, S. G., Renison, D., Hensen, I., Apaza, A. E., & Schleuning, M. (2014b). Topography and edge effects are more important than elevation as drivers of vegetation patterns in a neotropical montane forest. Journal of Vegetation Science, 25(3), 724–733. https://doi.org/10.1111/jvs.12132
- Missouri Botanical Garden – Bolivia Program. (2024). Chulumani Plots and Checklist Project. Tropicos.Org. Missouri Botanical Garden. http://legacy.tropicos.org/PlotSearch.aspx?projectid=93