Combining local ecological knowledge with camera traps to assess the link between African mammal life history traits and their occurrence in anthropogenic landscapes
Data files
Jul 12, 2024 version files 178.13 KB
-
baboon_data_integrated.rda
10.82 KB
-
bushbuck_data_integrated.rda
10.97 KB
-
bushpig_data_integrated.rda
10.90 KB
-
caracal_data_integrated.rda
10.79 KB
-
duikerblue_data_integrated.rda
10.71 KB
-
genetcapelargespotted_data_integrated.rda
10.97 KB
-
grysbokcape_data_integrated.rda
10.71 KB
-
honeybadger_data_integrated.rda
10.73 KB
-
leopard_data_integrated.rda
10.69 KB
-
mongooselargegrey_data_integrated.rda
10.72 KB
-
mongoosesmallcapegrey_data_integrated.rda
10.80 KB
-
mongoosewatermarsh_data_integrated.rda
10.70 KB
-
monkeyvervet_data_integrated.rda
10.78 KB
-
ottercapeclawless_data_integrated.rda
10.69 KB
-
porcupine_data_integrated.rda
10.89 KB
-
README.md
16.26 KB
Abstract
Understanding what influences species and trait composition is critical for predicting changes in communities driven by landscape transformation.
We explored how life history traits are associated with the persistence of mammal species in human-dominated habitats within the Garden Route Biosphere Reserve, South Africa. We combined data from a camera trap and a local ecological knowledge-based survey in an integrated occupancy model to analyze species occurrence along a gradient of anthropogenic landscape transformation.
Results confirmed that mammal occurrence in human-modified habitats was related to specific life history traits. Species with more specialist diets, as well as larger body mass species were more likely to stay in protected areas. Species with slow reproductive strategies occupied more natural areas.
Our study also showed that combining different monitoring methods enabled us to increase spatial coverage and mammal sighting numbers. This approach fostered research participation by various stakeholders, an important step for co-designing wildlife-friendly anthropogenic spaces.
Synthesis and applications: Integrating data from a standard ecological protocol and structured participatory citizen knowledge allowed us to identify the species functional traits associated with mammal species occurrence in anthropogenic landscapes at a local scale. These results advocate for wisely combining methods, and will guide conservation orientated land-use planning towards the protection of natural habitats in the Garden Route Biosphere Reserve. This methodological approach will enable managers and conservationists to use data obtain from diverse protocols. This should catalyze the involvement of citizens in biodiversity monitoring and conservation.
https://doi.org/10.5061/dryad.41ns1rnph
In this study, we explored how life history traits were associated with the occurence of mammal species in human-dominated habitats within the Garden Route Biosphere Reserve, South Africa. We combined data from a camera trap and a local ecological knowledge-based survey in an integrated occupancy model to analyze species occurrence along a gradient of anthropogenic landscape transformation. Results confirmed that mammal occurrence in human-modified habitats was related to specific life history traits. Species with more specialist diets, as well as larger body mass species were more likely to stay in protected areas. Species with slow reproductive strategies occupied more natural areas.
Description of the data and file structure
There is one dataset for each of the 16 species presented in the paper. The dataset are RDA files, which contains four lists: 1) the occurrence of a species, separately for the two methods of data collection (from the camera trap and the local ecological knowledge-based survey) ; 2) the occupancy variables used for the single-season, single-species occupancy model ; 3) the detection variables, specific to each method and 4) the list of sites ID.
1) List of species occurrence (y)
The first list of the RDA file contains two table : a) a table with presence/absence of a species based on the camera trap data, b) a table with the presence/absence of the same species based on the local ecological knowledge-based survey. Each lign of these tables correspond to a site and each column to a repeated occurence at a site.
2) Occupancy variables (occ.covs)
The second list of the RDA file contains a table with the different occupancy variables used in the occupancy analysis. The variables are named :
- hmi: the mean human modification index (HMI) value per 5km² cell
- prop_pa: the percentage of protected areas in a 5km² cell
-
cultivated: the percentage of culivated areas (i.e. cultivated land, pastures and orchards) in a 5km² cell
- naturality: the percentage of natual habitats (i.e. forest, fynbos, thicket and wetland) in a 5km² cell
For the three percentage variables (natural habitats, cultivated areas and protected areas), we used the arcsine square root transformation to account for the percentages being skewed towards 0%. All the variables were scaled.
3) Detection variables (det.covs)
The third list of the RDA file contains two lists: a) one with the detection variables for the camera traps and one fot the detection variables for the local ecological knowledge survey.
a) For data from the CTs, we controlled for variables which could explain heterogeneity in the detection of the species between the sites: (1) the visibility in front of the CT (variable named “visibility”), measured as the distance in meters from the camera to the surrounding brush using a rangefinder; (2) the model of the CT (Spypoint or Bushnell; variable named “model”); and (3) whether the CT was on a road, on a trail, on a game trail or outside of any identified trail (variable named “trail”). We measured the height, in centimeters, at which each CT was attached to each tree/pole (variable named “height”), but ultimately did not include it in the final analysis.
b) For data from the local ecological knowledge survey and to correct for respondents’ detection bias, both variables 1) the ecological knowledge score (each respondent was allocated a knowledge score based on their percentage of correct answers to the four ecological knowledge questions, between 0%, wherein no answer was correct, and 100%, wherein all the answers were correct ; variable named “knowledge”) and 2) the frequency of detection of a species (between 0 and 3 ; variable named “freq”), were considered as detection variables.
4) List of sites ID (sites)
The last list of the RDA file contains a list where each element is the site indices for the given data source. the first list contains the site indices for the camera traps site and the second for the local ecological knowledge survey)
The RDA files are formatted to fit the structure needed for integrated model using the package “spOccupancy” and “intPGOcc” function (Doser et al. 2022).
The structure of each file is as follow :
List of 4
$ y :List of 2 # OCCURRENCE DATA
..$ : tibble [66 × 16] (S3: tbl_df/tbl/data.frame) # OCCURRENCE DATA FROM THE CAMERA TRAPS
.. ..$ o13: int [1:66] 0 0 0 0 0 0 0 0 0 0 ...
.. ..$ o14: int [1:66] 0 0 0 0 0 1 0 0 0 0 ...
.. ..$ o15: int [1:66] 0 0 0 0 0 1 0 0 0 0 ...
.. ..$ o16: int [1:66] 0 0 1 0 0 1 0 0 0 0 ...
.. ..$ o17: int [1:66] 0 0 0 0 0 1 0 1 0 0 ...
.. ..$ o18: int [1:66] 0 1 1 0 0 0 1 0 0 0 ...
.. ..$ o19: int [1:66] 0 0 0 0 0 0 0 0 0 0 ...
.. ..$ o20: int [1:66] 0 0 1 0 0 0 0 1 0 0 ...
.. ..$ o21: int [1:66] 0 0 0 0 0 0 0 0 0 0 ...
.. ..$ o22: int [1:66] 0 0 0 0 0 0 0 0 0 0 ...
.. ..$ o23: int [1:66] 0 0 0 0 0 0 0 0 1 0 ...
.. ..$ o24: int [1:66] 0 1 1 0 0 0 0 0 1 0 ...
.. ..$ o25: int [1:66] 0 0 1 0 0 0 0 1 0 0 ...
.. ..$ o26: int [1:66] 0 0 1 0 0 0 0 0 0 0 ...
.. ..$ o27: int [1:66] 0 0 0 0 0 1 0 1 0 0 ...
.. ..$ o28: int [1:66] 0 0 1 0 0 1 0 0 0 0 ...
..$ : tibble [144 × 26] (S3: tbl_df/tbl/data.frame) # OCCURRENCE DATA FROM THE ONLINE SURVEY
.. ..$ occ1 : int [1:144] 1 1 0 0 1 0 0 0 1 1 ...
.. ..$ occ2 : int [1:144] 1 NA NA 1 NA NA 1 1 NA NA ...
.. ..$ occ3 : int [1:144] NA NA NA 0 NA NA 0 NA NA NA ...
.. ..$ occ4 : int [1:144] NA NA NA 1 NA NA NA NA NA NA ...
.. ..$ occ5 : int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. ..$ occ6 : int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. ..$ occ7 : int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. ..$ occ8 : int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. ..$ occ9 : int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. ..$ occ10: int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. ..$ occ11: int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. ..$ occ12: int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. ..$ occ13: int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. ..$ occ14: int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. ..$ occ15: int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. ..$ occ16: int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. ..$ occ17: int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. ..$ occ18: int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. ..$ occ19: int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. ..$ occ20: int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. ..$ occ21: int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. ..$ occ22: int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. ..$ occ23: int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. ..$ occ24: int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. ..$ occ25: int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. ..$ occ26: int [1:144] NA NA NA NA NA NA NA NA NA NA ...
$ occ.covs: num [1:182, 1:4] -0.404 -1.281 -0.209 -0.321 -1.192 ... # OCCUPANCY VARIABLES
..- attr(*, "dimnames")=List of 2
.. ..$ : NULL
.. ..$ : chr [1:4] "hmi" "prop_pa" "cultivated" "naturality"
$ det.covs:List of 2 # DETECTION VARIABLES
..$ :List of 4 # DETECTION VARIABLES FOR THE CAMERA TRAPS
.. ..$ height : tibble [66 × 16] (S3: tbl_df/tbl/data.frame)
.. .. ..$ H15: num [1:66] 0.371 1.156 0.707 -0.19 0.819 ...
.. .. ..$ H16: num [1:66] 0.371 1.156 0.707 -0.19 0.819 ...
.. .. ..$ H17: num [1:66] 0.384 1.16 0.716 -0.171 0.827 ...
.. .. ..$ H18: num [1:66] 0.384 1.16 0.716 -0.171 0.827 ...
.. .. ..$ H19: num [1:66] 0.387 1.164 0.72 -0.168 0.831 ...
.. .. ..$ H20: num [1:66] 0.395 1.176 0.729 -0.163 0.841 ...
.. .. ..$ H21: num [1:66] 0.395 1.176 0.729 -0.163 0.841 ...
.. .. ..$ H22: num [1:66] 0.395 1.176 0.729 -0.163 0.841 ...
.. .. ..$ H23: num [1:66] 0.395 1.176 0.729 -0.163 0.841 ...
.. .. ..$ H24: num [1:66] 0.395 1.176 0.729 -0.163 0.841 ...
.. .. ..$ H25: num [1:66] 0.395 1.176 0.729 -0.163 0.841 ...
.. .. ..$ H26: num [1:66] 0.395 1.176 0.729 -0.163 0.841 ...
.. .. ..$ H27: num [1:66] 0.395 1.176 0.729 -0.163 0.841 ...
.. .. ..$ H28: num [1:66] -0.375 1.186 0.74 -0.152 0.852 ...
.. .. ..$ H29: num [1:66] -0.375 1.186 0.74 -0.152 0.852 ...
.. .. ..$ H30: num [1:66] -0.375 1.186 0.74 -0.152 0.852 ...
.. ..$ model : tibble [66 × 16] (S3: tbl_df/tbl/data.frame)
.. .. ..$ C12: chr [1:66] "Spypoint" "Spypoint" "Bushnell" "Bushnell" ...
.. .. ..$ C13: chr [1:66] "Spypoint" "Spypoint" "Bushnell" "Bushnell" ...
.. .. ..$ C14: chr [1:66] "Spypoint" "Spypoint" "Bushnell" "Bushnell" ...
.. .. ..$ C15: chr [1:66] "Spypoint" "Spypoint" "Bushnell" "Bushnell" ...
.. .. ..$ C16: chr [1:66] "Spypoint" "Spypoint" "Bushnell" "Bushnell" ...
.. .. ..$ C17: chr [1:66] "Spypoint" "Spypoint" "Bushnell" "Bushnell" ...
.. .. ..$ C18: chr [1:66] "Spypoint" "Spypoint" "Bushnell" "Bushnell" ...
.. .. ..$ C19: chr [1:66] "Spypoint" "Spypoint" "Bushnell" "Bushnell" ...
.. .. ..$ C20: chr [1:66] "Spypoint" "Spypoint" "Bushnell" "Bushnell" ...
.. .. ..$ C21: chr [1:66] "Spypoint" "Spypoint" "Bushnell" "Bushnell" ...
.. .. ..$ C22: chr [1:66] "Spypoint" "Spypoint" "Bushnell" "Bushnell" ...
.. .. ..$ C23: chr [1:66] "Spypoint" "Spypoint" "Bushnell" "Bushnell" ...
.. .. ..$ C24: chr [1:66] "Spypoint" "Spypoint" "Bushnell" "Bushnell" ...
.. .. ..$ C25: chr [1:66] "Spypoint" "Spypoint" "Bushnell" "Bushnell" ...
.. .. ..$ C26: chr [1:66] "Spypoint" "Spypoint" "Bushnell" "Bushnell" ...
.. .. ..$ C27: chr [1:66] "Spypoint" "Spypoint" "Bushnell" "Bushnell" ...
.. ..$ trail : tibble [66 × 16] (S3: tbl_df/tbl/data.frame)
.. .. ..$ T15: num [1:66] 0.149 0.149 2.535 0.149 -1.044 ...
.. .. ..$ T16: num [1:66] 0.149 0.149 2.535 0.149 -1.044 ...
.. .. ..$ T17: num [1:66] 0.149 0.149 2.535 0.149 -1.044 ...
.. .. ..$ T18: num [1:66] 0.149 0.149 2.535 0.149 -1.044 ...
.. .. ..$ T19: num [1:66] 0.149 0.149 2.535 0.149 -1.044 ...
.. .. ..$ T20: num [1:66] 0.149 0.149 2.535 0.149 -1.044 ...
.. .. ..$ T21: num [1:66] 0.149 0.149 2.535 0.149 -1.044 ...
.. .. ..$ T22: num [1:66] 0.149 0.149 2.535 0.149 -1.044 ...
.. .. ..$ T23: num [1:66] 0.149 0.149 2.535 0.149 -1.044 ...
.. .. ..$ T24: num [1:66] 0.149 0.149 2.535 0.149 -1.044 ...
.. .. ..$ T25: num [1:66] 0.149 0.149 2.535 0.149 -1.044 ...
.. .. ..$ T26: num [1:66] 0.149 0.149 2.535 0.149 -1.044 ...
.. .. ..$ T27: num [1:66] 0.149 0.149 2.535 0.149 -1.044 ...
.. .. ..$ T28: num [1:66] 0.149 0.149 2.535 0.149 -1.044 ...
.. .. ..$ T29: num [1:66] 0.149 0.149 2.535 0.149 -1.044 ...
.. .. ..$ T30: num [1:66] 0.149 0.149 2.535 0.149 -1.044 ...
.. ..$ visibility: tibble [66 × 16] (S3: tbl_df/tbl/data.frame)
.. .. ..$ V15: num [1:66] -0.597 0.171 0.171 -0.501 -0.213 ...
.. .. ..$ V16: num [1:66] -0.597 0.171 0.171 -0.501 -0.213 ...
.. .. ..$ V17: num [1:66] -0.597 0.171 0.171 -0.501 -0.213 ...
.. .. ..$ V18: num [1:66] -0.601 0.168 0.168 -0.505 -0.216 ...
.. .. ..$ V19: num [1:66] -0.601 0.168 0.168 -0.505 -0.216 ...
.. .. ..$ V20: num [1:66] -0.601 0.168 0.168 -0.505 -0.216 ...
.. .. ..$ V21: num [1:66] -0.601 0.168 0.168 -0.505 -0.216 ...
.. .. ..$ V22: num [1:66] -0.586 0.189 0.189 -0.489 -0.198 ...
.. .. ..$ V23: num [1:66] -0.586 0.189 0.189 -0.489 -0.198 ...
.. .. ..$ V24: num [1:66] -0.586 0.189 0.189 -0.489 -0.198 ...
.. .. ..$ V25: num [1:66] -0.586 0.189 0.189 -0.489 -0.198 ...
.. .. ..$ V26: num [1:66] -0.586 0.189 0.189 -0.489 -0.198 ...
.. .. ..$ V27: num [1:66] -0.586 0.189 0.189 -0.489 -0.198 ...
.. .. ..$ V28: num [1:66] -0.601 0.168 0.168 -0.505 -0.216 ...
.. .. ..$ V29: num [1:66] -0.601 0.168 0.168 -0.505 -0.216 ...
.. .. ..$ V30: num [1:66] -0.594 0.174 -0.306 -0.498 -0.21 ...
..$ :List of 2 # DETECTION VARIABLES FOR THE ONLINE SURVEY
.. ..$ knowledge: tibble [144 × 26] (S3: tbl_df/tbl/data.frame)
.. .. ..$ K1 : int [1:144] 2 3 0 0 0 0 0 0 1 1 ...
.. .. ..$ K2 : int [1:144] 1 NA NA 2 NA NA 2 0 NA NA ...
.. .. ..$ K3 : int [1:144] NA NA NA 0 NA NA 0 NA NA NA ...
.. .. ..$ K4 : int [1:144] NA NA NA 1 NA NA NA NA NA NA ...
.. .. ..$ K5 : int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. .. ..$ K6 : int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. .. ..$ K7 : int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. .. ..$ K8 : int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. .. ..$ K9 : int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. .. ..$ K10: int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. .. ..$ K11: int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. .. ..$ K12: int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. .. ..$ K13: int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. .. ..$ K14: int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. .. ..$ K15: int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. .. ..$ K16: int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. .. ..$ K17: int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. .. ..$ K18: int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. .. ..$ K19: int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. .. ..$ K20: int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. .. ..$ K21: int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. .. ..$ K22: int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. .. ..$ K23: int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. .. ..$ K24: int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. .. ..$ K25: int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. .. ..$ K26: int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. ..$ freq : tibble [144 × 26] (S3: tbl_df/tbl/data.frame)
.. .. ..$ K1 : int [1:144] 25 25 25 25 75 25 50 50 0 0 ...
.. .. ..$ K2 : int [1:144] 75 NA NA 50 NA NA 50 25 NA NA ...
.. .. ..$ K3 : int [1:144] NA NA NA 50 NA NA 25 NA NA NA ...
.. .. ..$ K4 : int [1:144] NA NA NA 75 NA NA NA NA NA NA ...
.. .. ..$ K5 : int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. .. ..$ K6 : int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. .. ..$ K7 : int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. .. ..$ K8 : int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. .. ..$ K9 : int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. .. ..$ K10: int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. .. ..$ K11: int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. .. ..$ K12: int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. .. ..$ K13: int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. .. ..$ K14: int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. .. ..$ K15: int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. .. ..$ K16: int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. .. ..$ K17: int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. .. ..$ K18: int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. .. ..$ K19: int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. .. ..$ K20: int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. .. ..$ K21: int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. .. ..$ K22: int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. .. ..$ K23: int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. .. ..$ K24: int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. .. ..$ K25: int [1:144] NA NA NA NA NA NA NA NA NA NA ...
.. .. ..$ K26: int [1:144] NA NA NA NA NA NA NA NA NA NA ...
$ sites :List of 2 # UNIQUE SITE ID
..$ : int [1:66] 68 69 72 73 74 75 76 77 78 79 ... # SITE ID FOR SITES WITH CAMERA TRAPS DATA
..$ : int [1:144] 1 2 3 4 5 6 7 8 9 10 ... # SITE ID FOR SITES WITH ONLINE SURVEY DATA
Sharing/Access information
More information are available from the corresponding author, Alice Bernard (alice.bernard14@free.fr), if needed.
Code/Software
All the analyses were performed on R (R Core Team 2020).
Doser, J.W., Finley A.O., Kéry M., & Zipkin E.F. (2022). spOccupancy: An R Package for Single-Species, Multi-Species, and Integrated Spatial Occupancy Models. *Methods in Ecology and Evolution *13, no 8: 1670‑78.
R Core Team. (2020). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria.
Data were collected from a camera trap survey and an online questionnaire. The research was approved by Nelson Mandela University Human Research Ethic Committee (H20-SCI-SRU-002), SANParks (BERN-A/2020-008) and CapeNature (CN44-87-16198).