Data from: bringing the forest back: restoration priorities in Colombia
Data files
Feb 22, 2024 version files 2.11 GB
-
Diversity_and_distributions_submission.zip
2.11 GB
-
README.md
5.68 KB
Abstract
Aim: Colombia has committed to ambitious forest restoration targets which include a one million ha Bonn Challenge commitment and 6.47 - 8.31 million ha (rehabilitation and restoration, respectively) under the National Restoration Plan. Determining where and how to implement programs to achieve these targets remains a significant challenge.
Location: Colombia
Methods: We adopt a multi-objective optimisation framework for restoration planning and apply it to Colombia. We explore cost-effective solutions that leverage the potential for assisted natural regeneration benefits while accounting for opportunity and establishment costs of restoration and maximising biodiversity conservation and climate change mitigation benefits. We explore four politically relevant restoration areal targets (one, six, 6.47 and 8.31 million ha) and identify minimum cost, and suites of maximum benefit and cost-effective solutions.
Results: We identify solutions that simultaneously perform well across biodiversity and carbon objectives, despite trade-offs between these objectives. We find that cost-effective solutions can achieve on average 91.1%, 90.8%, 90.5%, and 90.1% of maximum carbon benefit and 100% of the maximum biodiversity benefit while significantly reducing costs. On average, the maximum benefit solutions reduce the cost by 16.9%, 30.2%, 31.1%, and 34.4% when considering the one, six, 6.47 and 8.31 million ha restoration targets respectively.
Main conclusions: Colombia has committed to bold restoration and conservation targets, such as those under the new 2030 Convention on Biological Diversity Global Biodiversity Framework. Strategic forest restoration planning will play an important role in achieving Colombia’s climate mitigation goals. We provide quantitative evidence to inform planning for environmentally and economically sensible restoration policy and practice in the country. Our framework and results can help guide Colombia towards meeting its ambitious forest restoration targets cost-effectively.
README
# WePlan Colombia
Code to run and create the dataframes for the analysis; however, we also provide the input dataframes that we used to run the analysis.
## Description of the data and file structure
The run_col_opt_submission.R is the file to run the optimisation (which calls in the functions.R file). The preprocessing_submission.R file can be used to create all of the necessary dataframes to run in the run_col_opt_submission.R code.
Description of .RData files (which must be placed with the "Input data" folder)
final_species.df.RData - Dataframe with species ID, the species class, and the path to the species file
hat.RData - Species habitat matrix
pu.df.RData - Dataframe with planning unit id numbers and associated values for each variable
pu.xy.RData - The x y coordinates of the planning units
puid_v3.RData - Planning unit id numbers
species.df.RData - Dataframe with species id, taxon group, area of occurrence, and current area of occupancy
species.pu.list.RData - A list of all of the planning units that each species occurs in
The code is set up so that the name of the input rasters/shp files are not that important, as long as the folder names are correct. All of this data feeds into the preprocessing_submission.R code.
The folders are:
land_use_map - current land use map
presettlement_habitat - historial potential ecosystem layer
carbon - potential for carbon sequestration
carbon_max - potential for carbon sequestration upper limit
carbon_min - potential for carbon sequestration lower limit
oppcost - opportunity cost
pnr - potential for natural regeneration
actcost - establishment cost (assuming active regeneration)
species - species data
The stages of the preprocessing_submission.R code is:
- Land use
- Presettlement ecosystems
- Area available for restoration
- Planning units
- Area currently forested
- Extract - Carbon and cost
- Species
- Presettlement habitat
- Create habitat matrix (which species is associated with which ecosystem)
- Create species dataframe
The mapdata folder contains some basic spatial data to create maps to display the results of the optimisation including country boundaries and topographic rasters
## Sharing/Access information
Links and citations to other publicly accessible locations of the data (all other data is available through the published article, species distribution models not yet publicly available for download are available upon request from Instituto Humboldt (the Humboldt Institute) jochoa@humboldt.org.co):
Land use
IDEAM. (2018). Mapa de Coberturas de la Tierra Metodología Corine Land Cover Escala 1:100.000 Periodo. http://www.siac.gov.co/catalogo-de-mapas
Presettlement ecosystems
Etter, A., Andrade, A., Saavendra, K., Amaya, P., Arevalo, P., Andrade, Á., & Arévalo, P. (2017). Risk assessment of Colombian ecosystems: An application of the Red List of Ecosystems methodology (Vers. 2.0).
Carbon
Broadbent, E., & Zambrano, A. (2021). Global Aboveground biomass Potential (GAP). http://www.speclab.org/global-aboveground-biomass-potential-gap.html
Opportunity cost
SEPAL. (2023). Cost data layers. https://docs.sepal.io/en/latest/modules/dwn/seplan.html#opportunity-cost
Species
Velásquez-Tibatá, J., Olaya-Rodríguez, M. H., López-Lozano, D., Gutiérrez, C., González, I., & Londoño-Murcia, M. C. (2019). BioModelos: A collaborative online system to map species distributions. PloS One, 14(3), e0214522.
Astorquiza Onofre, J. M. (2022). Patrones biogeográficos de diversidad Alfa, Beta y funcional de especies de murciélagos (mammalia, chiroptera) y su representatividad en el sistema nacional de áreas protegidas en Colombia. Masters Thesis, Universidad de Nariño. https://sired.udenar.edu.co/8075/
Ayerbe-Quiñones, F. A. (2018). Guía Ilustrada de la Avifauna Colombiana. Panamericana Formas e Impresos S. A, Bogota, Colombia.
Vélez, D., Tamayo, E., Ayerbe-Quiñones, F., Torres, J., Rey, J., Castro-Moreno, C., Ramírez, B., & Ochoa-Quintero, J. M. (2021). Distribution of birds in Colombia. Biodiversity Data Journal, 9, e59202.
Ramírez Chávez, H., Muñoz Rodríguez, C. J., Chacón Pacheco, J., Cepeda Duque, J. C., Pérez Torres, J., Vides Avilez, H. A., Castaño Salazar, J. H., Torres Martínez, M. M., Mejía Fontecha, I. Y., Mejía Correa, J. S., Concha Osbahr, D. C., Osbahr Hansen, K., Rojano Bolaño, C., Lizcano, D. J., Noguera Urbano, E. A., & Cruz Rodríguez, C. A. (2022). Atlas de la Biodiversidad de Colombia. Grandes Roedores: Mejores modelos con el apoyo de expertos. Instituto de Investigación de Recursos Biológicos Alexander von Humboldt. Bogotá DC, Colombia., 33.
Londoño, M., Olaya, M. H., Bello, C., González, I., Gutiérrez, C., López, D., & Velásquez, J. (2015). Regiones bióticas delimitadas utilizando como unidad de análisis los polígonos resultados del proceso de delimitación por la unidad ejecutora. Un mapa para cada grupo taxonómico: aves, mamíferos y herpetos y un mapa consenso. Laboratorio de Biogeografía Aplicada y Bioacústica. Instituto de Investigación de Recursos Biológicos Alexander von Humboldt. https://github.com/LBABHumboldt/BIOGEOGRAPHICAL_REGIONALIZATION
## Code/Software
R - https://www.r-project.org/
Gurobi - https://www.gurobi.com/ (there is an alternative ranking algorithm provided in the code where gurobi is not needed)
R packages used in code
gurobi, rgdal, raster, sp, rgeos, stringr, foreign, maptools, sf, rgeos, fasterize, dplyr, terra, sjmisc, rredlist, redlistr, tidyverse, data.table, rjson, spatialEco
Methods
We used spatial prioritisation, the process of using computational tools for the informed spatial allocation of actions or placement land uses, to achieve an objective of restoring forest to maximise biodiversity and carbon sequestration benefits within selected priorities, while considering establishment and opportunity cost. Tree planting and extensive site preparation are popular restoration strategies and can be effective, but implementation can be prohibitively expensive for some sites or at large scales. Where ecological conditions are such that forests can grow back on their own or with low-cost assistance, natural regeneration methods can be less costly. To leverage these potential costs our establishment cost estimates account for the potential for natural regeneration by adjusting values relative to a spatially explicit random forest model.
Usage notes
R - https://www.r-project.org/
Gurobi - https://www.gurobi.com/ (there is an alternative ranking algorithm provided in the code where gurobi is not needed)