A fundamental tenet of maternal effects assumes that maternal variance over time should have discordant consequences for offspring traits across litters. Yet, seldom are parents observed across multiple reproductive bouts, with few studies considering anthropogenic disturbances as an ecological driver of maternal effects. We observed captive coyote (Canis latrans) pairs over two successive litters to determine whether among-litter differences in behavior (i.e., risk-taking) and hormones (i.e., cortisol and testosterone) corresponded with parental plasticity in habituation. Thus, we explicitly test the hypothesis that accumulating experiences of anthropogenic disturbance reduces parental fear across reproductive bouts, which should have disparate phenotypic consequences for first- and second-litter offspring. To quantify risk-taking behavior, we used foraging assays from 5–15 weeks of age with a human observer present as a proxy for human disturbance. At 5, 10, and 15 weeks of age, we collected shaved hair to quantify pup hormone levels. We then used a quantitative genetic approach to estimate heritability, repeatability, and between-trait correlations. We found that parents were riskier (i.e., foraged more frequently) with their second versus first litters, supporting our prediction that parents become increasingly habituated over time. Second-litter pups were also less risk-averse than their first-litter siblings. Heritability for all traits did not differ from zero (0.001-0.018); however, we found moderate support for repeatability in all observed traits (r = 0.085-0.421). Lastly, we found evidence of positive phenotypic and cohort correlations among pup traits, implying that cohort identity (i.e., common environment) contributes to the development of phenotypic syndromes in coyote pups. Our results suggest that parental habituation may be an ecological cue for offspring to reduce their fear response, thus emphasizing the role of parental plasticity in shaping their pups’ behavioral and hormonal responses toward humans.
Coyote Population Pedigree
This datafile contains pedigree information for all individual pups and parents included in the study. The 'pedantics' package was used on these data to determine pedigree specifications and summary statistics for the pedigree used in all Bayesian animal models. These data can be found in the Supporting Information Table S1.
Pedigree.A_Long.csv
Coyote Risk-Taking Assays
This CSV file contains all foraging (i.e. risk-taking) assay data, for both pups and parents, over the 5-15 week study period. Each row represents a single observation for each individual in order to assess repeated measures in pups/parents. The "animal", "dam", and "sire" columns are synonymous to those found in the 'Coyote Pedigree' file attached; this was done as a requirement by the 'MCMCglmm' package, which needs all three columns to structure data around the pedigree. The column titled "Lit.ID" denotes the litter identity of each pup, the variable used to observe common environment effects using MCMCglmm. The "ID" column is identical to the "animal" column, but is necessary in order to partition additive genetic effects from permanent environment effects (see Wilson et al. 2010 - An Ecologist's Guide to the Animal Model). The column "Age" denotes the age of the pup/litter when the observation occurred, the "life.stage" partitions pup observations from adult observations, and the "Dev.Period" column partitions the weaning and post-weaning periods (i.e. 5-9 week observations and 10-14 week observations, which can be traced back to the "Age" column). The "risk.success" column represents whether an individual foraged or not during an observation (i.e. "0" for no, and "1" for yes), whereas the "risk.total" column is always equal to "1", denoting the total number of observations. This column is only important insofar as the way in which you choose to analyze the data in the MCMCglmm package. If using the distribution family argument "multinomial2", then the "risk.total" column is used as the second variable in that dependent variable form. Further - "Lit.size" denotes litter size of the pup/adult, "Year" (either 1st or 2nd) denotes the year of observation and the parity of the parent, "Sex" is self-explanatory, and "Trmt" indicates the PARENTAL odor treatment group, which occurred BEFORE the pups were born (see Schell et al. 2016 - Olfactory attractants and parity affect prenatal androgens and territoriality of coyote breeding pairs).
Coyote Risk-taking Assays_Dryad.csv
Coyote Pup Hormones_pt.1
This datafile contains information for both pup cortisol and testosterone, to perform univariate animal models used to measure repeatability and partition components of variance in each hormone. This file was also used for univariate model selection analyses, which are detailed in the Supporting Information Table S3. This file, however, was not used for multivariate analyses to determine correlation estimates among traits; that file is designated "Coyote Pup Hormones_pt.2". In this file, all of the identical columns from the "Coyote Risk-taking Assays.csv" file are include, sans the columns that pertain to coyote risk-taking. The values for the "Cortisol" and "Testo" columns are in units "ng/g", and pertain specifically to the values from the hair sample collected at the particular developmental time (i.e. AGE=WK5, WK10, and WK15).
Coyote Pup Hormones_pt.2
This aggregated file contains information on pup cortisol, testosterone, and risk-taking behavior to perform the multivariate analyses used to determine correlation estimates among traits (risk-taking with cortisol; cortisol with testosterone; and risk-taking with testosterone). Data for each individual were binned according to the period of hair sample collection. Hence, hair collected at 10 weeks of age were paired with the aggregate number of foraging "successes" (i.e. "1") over the 5-10 week developmental period. To use a "multinomial2" distribution, the column "FR.total" was included with "FR.success", which creates a proportional matrix of how often the individual ate in front of an observer. All other columns are identical to the files "Coyote Pup Hormones_pt.2" and "Coyote Risk-taking Assays"
R Code_Coyote Parental Fear
Complete and annotated R script used for all analyses. The data files denoted in this Dryad repository are named and linked to all of the code in this file. Thus, Bayesian animal models performed, model selection analyses, correlation estimates, highest posterior density intervals (HPDI), and model assessment lines of code are all contained in this file.
Coyote Parental Fear_Dryad.R