Data from: Epigeal arthropods in small-scale oil palm plantations of Central America driven by landscape-scale habitat disturbance
Data files
Mar 09, 2026 version files 11.25 KB
-
data_iNext.csv
2.37 KB
-
data_models.csv
5.51 KB
-
README.md
3.37 KB
Abstract
With anthropogenic pressures driving global biodiversity declines, there is an urgent need to better understand how agricultural practices shape species communities. This study evaluated the potential benefits of diversified oil palm plantations on biodiversity by comparing epigeal arthropod (EA) communities in the oil palm monoculture and polyculture field sites of the Experimental African Palm Laboratory (LAPA) project in southwest Costa Rica. EA were sampled using pitfall traps to assess the effects of cropping system (monoculture vs. polyculture), landscape-scale habitat disturbance, sampling session (dry vs. rainy season), and site-specific soil characteristics on arthropod diversity. Specimens were sorted to order level, with Coleoptera further sorted to species to detect potential taxon-specific responses. General additive mixed models indicated higher EA diversity in oil palm monocultures, with a similar, non-significant trend observed for Coleoptera, and diversity of both EA and Coleoptera was higher in the dry season. Interestingly, landscape-scale habitat disturbance negatively affected EA and Coleoptera diversity in monocultures. This finding suggests that, in smallholder oil palm plantations, maintaining landscape integrity can support arthropod diversity and that polyculture practices may buffer the impacts of landscape-scale habitat disturbance at patch level. Given the limitations associated with pitfall trapping and coarse taxonomic resolution, we recommend future research to employ complementary sampling methods and time- and cost-effective sorting techniques to further explore the patterns observed in our study.
Dataset DOI: 10.5061/dryad.1vhhmgr84
Description of the data and file structure
(1) data_iNext.csv
Excel file including the epigeal arthropod abundance matrix for use in biodiversity analysis (iNext).
Description of column and row names of the abundance matrix:
-Row names: Scientific names of identified arthropod orders.
-Column names: Unique codes referring to the sampling units which includes a combination of (i) sampling session (S1 = sampling session 1, S2 = sampling session 2, S3 = sampling session 3, S4 = sampling session 4), (ii) sampling site (CO = Coto 63, LG = La Guaria, PB = Piedras Blancas, CH = Chacarita, SA = Salama, RQ = Rancho Quemado, BA = Banegas, LP = La Palma), and (iii) type of cropping system (M = monoculture, P = polyculture).
(2) data_models.csv:
Excel file including biodiversity Hill índices and metadata for use in GAMM models.
Description of column names of the dataset:
-ID:
Unique codes referring to the sampling units, aligns with the codes used as column names in the epigeal arthropod abundance matrix (data_iNext.csv).
-Response variables:
hill1_arthropods (= hill number 1 calculated using the epigeal arthropod order diversity matrix; representing the effective number of arthropod oders in an assemblage, calculated as the exponential of the Shannon entropy index)
hill1_beetles (= hill number 1 calculated using the Coleoptera species diversity matrix; representing the effective number of Coleoptera species in an assemblage, calculated as the exponential of the Shannon entropy index)
-Predictor variables:
Season: categorical variable referring to the sampling season, with two categories; dry = dry season and rainy = rainy season.
Session: categorical variable referring to the sampling session, with four categories; session 1 = first sampling session (March 2023), session 2 = second sampling session (April 2023), session 3 = third sampling session (May 2023), session 4 = fourth sampling session (June 2023).
Site: categorical variable referring to the sampling site, with eight categories; CO = Coto 63, LG = La Guaria, PB = Piedras Blancas, CH = Chacarita, SA = Salama, RQ = Rancho Quemado, BA = Banegas, LP = La Palma.
System: categorical variable referring to the cropping system, with two categories; M = monoculture, P = polyculture.
Disturbance: numeric variable referring to the calculated habitat disturbance in the surrounding landscape (proportion).
PCA1: numeric variable referring to the first axis principal component analysis (PCA) applied to the soil chemical variables dataset, including variables related to soil acidity (pH in water, soil exchangeable acidity, exchangeable iron, cation exchange capacity).
PCA2: numeric variable referring to the second axis principal component analysis (PCA) applied to the soil chemical variables dataset, including variables contributing to macronutrient cycling (total phosphorous and exchangeable potassium).
PCA3: numeric variable referring to the third axis principal component analysis (PCA) applied to the soil chemical variables dataset, including variables contributing to micronutrient cycling (exchangeable magnesium and zinc).