Data from: Testing the benefits of conservation set-asides for improved habitat connectivity in tropical agricultural landscapes

Scriven, Sarah A.; Carlson, Kimberly M.; Hodgson, Jenny A.; McClean, Colin J.; Heilmayr, Robert; Lucey, Jennifer M.; Hill, Jane K.

Published Aug 30, 2019 on Dryad. https://doi.org/10.5061/dryad.600vs50

Data files

Aug 30, 2019 version files 13.26 MB

Abstract

1. Habitat connectivity is important for tropical biodiversity conservation. Expansion of commodity crops, such as oil palm, fragments natural habitat areas, and strategies are needed to improve habitat connectivity in agricultural landscapes. The Roundtable on Sustainable Palm Oil (RSPO) voluntary certification system requires that growers identify and conserve forest patches identified as High Conservation Value Areas (HCVAs) before oil palm plantations can be certified as sustainable. We assessed the potential benefits of these conservation set-asides for forest connectivity. 2. We mapped HCVAs and quantified their forest cover in 2015. To assess their contribution to forest connectivity, we modelled range expansion of forest-dependent populations with five dispersal abilities spanning those representative of poor dispersers (e.g., flightless insects) to more mobile species (e.g., large birds or bats) across 70 plantation landscapes in Borneo. 3. Because only 21% of HCVA area was forested in 2015, these conservation set-asides currently provide few connectivity benefits. Compared to a scenario where HCVAs contain no forest (i.e., a no-RSPO scenario), current HCVAs improved connectivity by ~3% across all dispersal abilities. However, if HCVAs were fully reforested, then overall landscape connectivity could improve by ~16%. Reforestation of HCVAs had the greatest benefit for poor to intermediate dispersers (0.5-3 km per generation), generating landscapes that were up to 2.7 times better connected than landscapes without HCVAs. By contrast, connectivity benefits of HCVAs were low for highly mobile populations under current and reforestation scenarios, because range expansion of these populations was generally successful regardless of the amount of forest cover. 4.Synthesis and applications. The RSPO requires that HCVAs be set aside to conserve biodiversity, but HCVAs currently provide few connectivity benefits because they contain relatively little forest. However, reforested HCVAs have the potential to improve landscape connectivity for some forest species (e.g., winged insects), and we recommend active management by plantation companies to improve forest quality of degraded HCVAs (e.g., by enrichment planting). Future revisions to the RSPO’s Principles and Criteria (P&C) should also ensure that large (i.e., with a core area >2 km2) HCVAs are reconnected to continuous tracts of forest to maximise their connectivity benefits.

Data and R code for running IFM simulations

Zip folder containing raw data and R code for running Incidence Function Model (IFM) simulations. The 'LoopedIFM_SetupScript.R' script runs the IFM simulations in a loop for the 70 New Planting Procedure (NPP) plantation landscapes. Model parameters (e.g., dispersal distance and population density) can be changed in this script. The set-up script also loads the raw data (raster) files ('FinalRaster70NPPsBuffer.tif' and 'FinalRaster70NPPsCurrentBuffer.tif'), which are converted into a dataframe before the IFM simulations are run. The set-up script calls three functions: the 'IFM2_Function.R' script, which is the IFM simulation code, as well as two 'Cardinal Points function' scripts ('CardinalPoints_WithHCVFunctionParallel.R' and 'CardinalPoints_WithNoHCVFunctionParallel.R'), which set up the modeled landscapes over which to run the simulations. Finally, the 'NPP70Centroids.csv' file is required to run the loop and provides NPP plantation IDs and X/Y coordinates of each plantation centroid.

Connectivity_simulations.zip

Data and R code for logistic regression analyses

Zip folder containing data and R code for running the logistic regression analyses. The 'LogisticRegression_FinalScript.R' script contains R code for the General Additive Models (GAMs) and the Generalized Linear Mixed Models that were used to analyse the Incidence Function Model (IFM) simulation results. The summarised IFM simulation outputs used in the statistical analyses are contained within the 'SuccessGLMDataNew.csv' file. Definitions of column headings within this file are provided in the R script.

Logistic_regression.zip

HCVA and NPP plantation summary data

Zip folder containing three Excel spreadsheets with summary information for the 70 New Planting Procedure (NPP) plantations and associated High Conservation Value Areas (HCVAs) used in our study. The 'SummaryData_ForHistograms.xlsx' file contains the data used for Fig. 3 in the main text, and includes the area and forest cover of HCVAs within each NPP plantation. It also includes the size of each NPP plantation estate and the area of non-HCVA forest cover within each NPP plantation. The 'SummaryData_HCVAPatchSize.xlsx' file contains the size of all spatially discrete HCVA patches within each NPP plantation and the 'SummaryData_HCVAType.xlsx' contains the size and forest cover of all spatially discrete HCVA patches within each NPP plantation that could be identified to Type (e.g., HCV 1-6). Descriptions of column headings are provided in the Excel files.

HCVA_summary_data.zip