An integrated population model reveals source-sink dynamics for competitively subordinate African wild dogs linked to anthropogenic prey depletion

Creel, Scott1 ; Merkle, Johnathan1; Goodheart, Ben1; Mweetwa, Thandiwe2; Mwape, Henry2; Simpamba, Twakundine3; Becker, Matthew2

Published Jan 17, 2024 on Dryad. https://doi.org/10.5061/dryad.tdz08kq61

Data files

Jan 17, 2024 version files 212.78 KB

bayesian_cjs_Luangwa_WD_data_object.csv

205.45 KB
lve_awd_fecundity_by_area

2.68 KB
README.md

4.65 KB

Abstract

Many African large carnivore populations are declining due to decline of the herbivore populations on which they depend. The densities of apex carnivores like the lion and spotted hyena correlate strongly with prey density, but competitive subordinates like the African wild dog benefit from competitive release when the density of apex carnivores is low, so the expected effect of a simultaneous decrease in resources and dominant competitors is not obvious.
Wild dogs in Zambia’s Luangwa Valley Ecosystem occupy four ecologically similar areas with well-described differences in the densities of prey and dominant competitors, due to spatial variation in illegal offtake.
We used long-term data to fit a Bayesian integrated population model (IPM) of the demography and dynamics of wild dogs in these four regions. The IPM used Leslie projection to link a Cormack-Jolly-Seber model of area-specific survival (allowing for individual heterogeneity in detection), a zero-inflated Poisson model of area-specific fecundity, and a state-space model of population size that used estimates from a closed mark-capture model as the counts from which (latent) population size was estimated.
The IPM showed that both survival and reproduction were lowest in the region with the lowest density of preferred prey (puku, Kobus vardonii, and impala, Aepyceros melampus), despite little use of this area by lions. Survival and reproduction were highest in the region with the highest prey density, and intermediate in the two regions with intermediate prey density. The population growth rate (λ) was positive for the population as a whole, strongly positive in the region with the highest prey density, and strongly negative in the region with the lowest prey density.
It has long been thought that the benefits of competitive release protect African wild dogs from the costs of low prey density. Our results show that the costs of prey depletion overwhelm the benefits of competitive release and cause local population decline where anthropogenic prey depletion is strong. Because competition is important in many guilds and humans are affecting resources of many types, it is likely that similarly fundamental shifts in population limitation are arising in many systems.

README: An integrated population model reveals source-sink dynamics for competitively subordinate African wild dogs linked to anthropogenic prey depletion

https://doi.org/10.5061/dryad.tdz08kq61

Description of the data and file structure

Scott Creel, 4 January 2024

Data and JAGS code for the Bayesian Integrated Population Model of Creel et al. 2014. An integrated population model reveals source-sink dynamics for competitively subordinate African wild dogs linked to anthropogenic prey depletion. Journal of Animal Ecology.

The data are in two comma-delimited files:

bayesian_cjs_Luangwa_WD_data_object.csv

This file contains capture histories and associated covariates for the survival component model. All of the data codes and columns are described below under code/software, as comments in R code that will read the data in the correct format for the integrated population model code. That code will create four R objects used in the model:

CH is a dataframe that holds capture histories, with one row per individual (463 individuals) and 63 columns. Each column represents one time interval, with nine intervals per year over a period of 7 years as described in the paper. In each cell, a zero indicates that an individual was not detected in that interval, and a 1 indicates that it was detected.
sex is a vector with 463 values, one for each individual in CH. It holds character data with values of 'Female' and 'Male'.
x is a matrix that holds data on each individual's age at each time period considered in the analysis, with 463 rows, one for each individual in CH, and 63 columns, each holding data for the same time bins as in CH. It holds numeric data with codes as follows: 1 = pup (less than 1 year old), 2 = yearling (one year old), 3 = adult (two years old or older).
g is a matrix that holds data on each individual's location at each time period considered in the analysis, with 463 rows, one for each individual in CH, and 63 columns, each holding data for the same time bins as in CH. It holds numeric data with codes as follows: 1=Lower Lupande, 2 = Main Game, 3 = North, 4 = Nsefu, 5 = other regions. Individuals with code 5 are not included the the final analysis of the paper.

lve_awd_fecundity_by_area.csv

This file contains data on annual fecundity and associated covariates for the reproduction component model. All of the data codes and columns are described below under code/software, as comments in R code that will read the data in the correct format for the integrated population model code. There is one row per individual, with two variables in each row.

X.Recruit1yr holds numeric data, giving the number of offspring that a female raised to one year of age.
area holds character data, giving the area in which the female lived in that year. These are the same 4 areas described above.

Code/Software

awd_ipm.txt is the JAGS code to fit the IPM to data from African wild dogs in the Luangwa Valley Ecosystem using the jags() function of the R package jagsUI. Prior to running the jags() function, the data must be read into R objects and bundled into a list, for example with:

# Read capture histories and associated covariates to fit CJS model

# CH = 0/1, detection matrix of 463 rows (animals) and 63 columns (occasions)

# sex = 'Female', 'Male', vector with 463 animals

# x 1=pup, 2=yearling, 3=adult, age matrix of 463 rows (animals) and 63 columns (occasions)

# g 1=Lower Lupande, 2 = Main Game, 3 = North, 4 = Nsefu, region matrix, 463 x 63

data <- read.csv("bayesian_cjs_Luangwa_WD_data_object.csv", header = TRUE)

CH <- data.matrix(data[,2:64])

sex <- data.matrix(data[,65])

x <- data.matrix(data[,66:128])

g <- data.matrix(data[, 129:191])

# Read fecundity and associated covariate to fit CJS model

fecundity <- read.csv("lve_awd_fecundity_by_area.csv", header = TRUE)

# Bundle data and inspect

# first line is for CJS model of apparent survival

# second line is for ZIP model of fecundity

# third line is for the SSM model of population size

jags.data <- list(y=CH, f=f, nind=dim(CH)[1], n.occasions=dim(CH)[2], z=known.state.cjs(CH), area=g, nareaclass=4,

C = fecundity$X.Recruit1yr, n = length(fecundity$X.Recruit1yr), region=area,

Pop=Pop, n.occ= length(Pop))

str(jags.data)

In the code above, the known.state.cjs() function is provided on page 182 of Kery, M & Schaub, M. 2012. Bayesian Population Analysis using WinBUGS. Academic Press, and the object Pop is a vector of integers = [115, 109, 112, 102, 110, 88, 61].