Long-term data reveal fitness costs of anthropogenic prey depletion for a subordinate competitor, the African wild dog (Lycaon pictus)
Data files
Jul 03, 2024 version files 244.52 KB
-
BetaReg_LVE_Recruitment_Ratio.R
3.58 KB
-
CH_2019.csv
1.57 KB
-
Density_Estimates.csv
1.16 KB
-
LVE_CJS_Survival.R
11.67 KB
-
LVE_CJS.csv
204.98 KB
-
LVE_Density.R
3.78 KB
-
LVE_Littersize_2008-2021.csv
973 B
-
LVE_Littersize_2008-2021.R
3.26 KB
-
LVE_Recruit_to_1yr_2008-2021.csv
1.10 KB
-
LVE_Recruit_to_1yr_2008-2021.R
3.28 KB
-
README.md
8.17 KB
-
Recruitment_Ratio.csv
996 B
Abstract
Within carnivore guilds, dominant competitors (e.g., lions, Panthera leo) are limited primarily by the density of prey, while subordinate competitors (e.g., African wild dogs, Lycaon pictus) have been limited by the density of dominant competitors. Historically, the fitness and population density of subordinate competitors have not been tightly linked to prey density. However, populations of large herbivores have declined substantially across sub-Saharan Africa due to human impacts, and where prey depletion is severe, fitness costs for competitive subordinates may begin to outweigh the benefits of competitive release. Using long-term intensive monitoring of African wild dogs in Zambia’s Luangwa Valley Ecosystem (LVE), we tested the effects of prey depletion on survival and reproduction. Our study area included four contiguous regions that varied in protection, prey density, and lion density. We fit Bayesian Cormack-Jolly-Seber and closed-capture models to estimate effects on survival and population density, and generalized linear models to estimate effects on reproductive success. We found that the LVE is a stronghold for wild dogs, with an estimated median density of 4.0 individuals/100 km2. Despite this high density, survival and reproduction differed among regions, and both components of fitness were substantially reduced in the region with the lowest prey density. Anthropogenic prey depletion is becoming an important limiting factor for African wild dogs. If prey depletion (or any other form of habitat degradation) becomes severe enough that its fitness costs outweigh the benefits of competitive release, such changes can fundamentally alter the balance between limiting factors for competitively subordinate species.
https://doi.org/10.5061/dryad.qbzkh18q0
Data was collected through long-term intensive monitoring of African wild dogs (Lycaon pictus).
Capture histories were created from sighting and re-sighting of identified individuals through time.
Reproductive data was recorded from observation of litter size and the pups’ subsequent survival.
For more information on methods and analysis techniques, please refer to the article.
Description of the data and file structure
The R code files that generate the analysis and figures for the manuscript can be used as follows:
LVE_Density.R: Generates the mean median densities and provides an example capture history (CH) to show how annual abundance was estimated.
Associated files:
Density_Estimates.csv
| Heading | Description |
|---|---|
| LL_Density | The lower limit of 95% credible interval for density estimate (number of adult & yearling African wild dogs/ 100 km^2). |
| Mean_Density | The mean density estimate (number of adult & yearling African wild dogs/ 100 km^2). |
| Median_Density | The median density estimate (number of adult & yearling African wild dogs/ 100 km^2). |
| (UL_Density | The upper limit of 95% credible interval for density estimate (number of adult & yearling African wild dogs/ 100 km^2). |
| Input | The number of individuals included in the capture history for a respective year. Adults & yearlings only. |
| Added | The number of individuals (adults & yearlings) within the study area that were excluded from the estimation of abundance as they either died or dispersed during the annual monitoring period. These individuals were added back to the model’s output estimate of annual abundance. |
| outputLL | The lower limit of the 95% credible interval for estimated annual abundance (number of adult & yearling African wild dogs) plus the individuals in the “Added” column. |
| outputMean | The mean estimated annual abundance (number of adult & yearling African wild dogs) plus the individuals in the “Added” column. |
| outputMedian | The median estimated annual abundance (number of adult & yearling African wild dogs) plus the individuals in the “Added” column. |
| outputUL | The upper limit of the 95% credible interval for estimated annual abundance (number of adult & yearling African wild dogs) plus the individuals in the “Added” column. |
| Area(Km2) | The estimated area of the annual demographic monitoring area in kilometers squared based on the merged annual home ranges. |
| Type | The type of method used to estimate annual demographic monitoring area; KUD = Kernal Utilization Distribution or DBBMM = Dynamic Brownian Bridge Movement models. |
| 4yrs | Yes or No variable, to filter out which years were included in the four-year analysis. |
| 7yrs | Yes or No variable, to filter out which years were included in the seven-year analysis. |
CH_2019.csv
- The seven columns are monthly detection bins.
- Rows are different individuals.
- Binary for detected (1) or undetected (0).
LVE_CJS_Survival.R: Generates the 1) estimates of region-specific annual apparent survival; 2) estimates of age & sex-specific annual apparent survival.
Associated file:
LVE_CJS.csv
Columns:
- ch.1 to ch. 63: The 63 detection bins (9 bins/year across 7 years); with binary for detected (1) or undetected (0).
- group: Factor with two levels (“Female”, “Male”) based on the gender of the individuals.
- X1 to X63: The 63 bins (9 bins/year across 7 years) track individuals’ age class, across three levels (1=Pup; 2=Yearling; 3= Adult).
- X1.1 to X63.1: The 63 (9 bins/year across 7 years) bins track the region used by individuals, across 4 levels (1=Lower Lupande, 2=Main Game, 3=North, 4=Nsefu).
LVE_Littersize_2008-2021.R: Generates the estimated regional effect on litter size while accounting for pack size (adults).
Associated file:
LVE_Littersize_2008-2021.csv
Columns:
- Region: 4 levels (Lower Lupande, Main Game, North, Nsefu)
- Littersize: The number of pups (i.e., litter size) born to a respective pack year.
- Adults: The number of adult individuals in the pack.
LVE_Recruit_to_1yr_2008-2021.R: Generates the estimated regional effect on the number of pups recruited to one year while accounting for pack size (adults).
Associated file:
LVE_Recruit_to_1yr_2008-2021.csv
Columns:
- Region: 4 levels (Lower Lupande, Main Game, North, Nsefu)
- Recruit1yr: The number of pups that survive to 1 year of age for a respective pack-year.
- Adults: The number of adult individuals in the pack.
BetaReg_LVE_Recruitment_Ratio.R: Generates the estimated regional effect on the recruitment ratio while accounting for pack size (adults).
Associated file:
Recruitment_Ratio.csv
Columns:
- Region: 4 levels (Lower Lupande, Main Game, North, Nsefu)
- RecruitRatio: The proportion of pups born that survive to one year of age, only for pack-years with both known litter size and the number of pups at one year of age.
- Adults: The number of adult individuals in the pack.
Code/Software
R & JAGS:
R script, JAGS model, and data used to estimate mean density, apparent survival (Φ), and measures of reproductive success for African wild dogs in the Luangwa Valley Ecosystem.
R scripts are attached as data in the data files.