Faecal nutrient deposition of domestic and wild herbivores in an alpine grassland
The Dataset.xlsx contains data of the faecal nutrient content of four herbivores species (chamois, sheep, horse, and cow) in a alpine grassland in the Spanish Pyrenees. Other information is also added, like the body size, feeding strategy, and the dung dry weight of each species. Our main main objectives to study were 1) explore if herbivore dung stoichiometry differed between species and body size groups, 2) determine if seasonal differences in herbivore dung stoichiometry persisted in our study system, and 3) quantify the amount of nutrients released annually by each herbivore. The quantification of the amount of nutrient released annually (objective 3) can be found in the Dataset.xlsx in the datasheet annual_input. The fie R_script.zip contains the dataset used in a csv delimited by commas, and the R script used to answer our objectives 1) and 2). We have found differences in faecal nutrient content due to seasonality and body size. In addition, cattle is by far the herbivore with more fertilization effect in the ecosystem.
Steps to reproduce
Fresh faecal samples of the studied herbivores were collected monthly from June to October for two years (2011 and 2012) in both study areas (Fontalba and Costabona) following a transsect of 5 km. A group of droppings collected in one bag was considered a single faecal sample, even if the droppings could have belonged to different individuals. Samples were transported to the laboratory and stored frozen at -20oC. Before processing, samples were thawed for 24 hours and then dried in an oven at 80 oC for 48 h. To analyse nutrient content we selected two faecal samples for each site, month, and year. Some months had a lower number of samples because no fresh faeces were found. A total of 121 samples (33 for chamois, 23 for sheep, 31 for horse, and 34 for cattle) . Faecal samples were ground and homogenized using an MM400 Retsch mixer ball mill. Three replicate aliquots of each sample were analysed for total C and N concentrations by gas chromatography using a LECO TruSpec Micro CNHS autoanalyzer. We used linear mixed-effect models (LMMs) to analyse the relationship between faecal nutrient concentrations (C, N, P, K) and stoichiometry (C/N) and herbivore species (chamois, sheep, horse, and cattle), body size (small or large herbivores), and seasonality (from June to October). Prior to analysis, we excluded one faecal nitrogen data point for chamois, which was a clear outlier 4 standard deviations away from the mean. We first built models for each of the nutrients to assess if the average concentration for each nutrient (C, K, N, P, C/N) was different among herbivores. We included a nested random effect of the sampling site and a variable for each independent month across the two years. To compare each species, we then performed a posthoc Tukey test on these models using the package ‘multcomp.’ To assess if average nutrient values were different among body size, we then divided the herbivore species into two groups: small herbivores (chamois and sheep) and large herbivores (horse and cattle). We classified herbivores per body size and not per feeding strategy (i.e., grazer, browser, and mixed feeder) due to our low sample size of each guild (see Table 1). We then built a set of repeated measures ANOVAs to explore if seasonal differences existed for each species for each of the nutrients (C, N, P, K C/N). For the models with significant seasonal differences, we performed a posthoc Tukey test to evaluate if any seasonal patterns persisted across nutrients. All statistical analyses were conducted using the ‘lme4’ and ‘lmerTest’ packages in the R Statistical Software 3.6.2 version (Kuznetsova et al., 2017; R Core Team, 2021).