Agricultural use metrics were calculated for 66 elephants as part of a study to characterize crop use tactics in the Mara-Serengeti ecosystem in Kenya and Tanzania. Metrics were calculated to capture mean agricultural use, maximum use from a moving average, and the difference between mean and max use. These metrics were used to classify agricultural use tactics for each elephant using Gaussian mixture models. Tables are provided with metrics and tactic classifications for the lifetime track (TableS3) and individual years (TableS4). Data contained in these files can be used to reproduce and further investigate Guassian mixture model clustering and cutpoint calculation, agricultural use linear mixed models, and tactic change generalized logistic mixed models in Hahn et al. 2021. 

Elephant tracking data was collected Masai Mara, Kenya (IACUC Protocol No. 1458) and western Serengeti, Tanzania (IACUC Protocol No. xx) between 2011 and 2019. For each individual, metrics were calculated over the lifespan of the GPS collar (n = 66 individuals) and on a yearly basis (n = 202 individual-years). For each time scale, we used R to calculate the mean use (aggregate.mean, year.mean), the max 90-day moving average, and the difference between mean and max (aggregate.delta, year.delta). These values were used in Gaussian mixture models to identify tactic clusters in the dataset using different combinations of these values. Tactic classification results are presented for the top three models identified in Hahn et al. 2021. Subject age class was assigned to all collared elephants as young adults (15-34) and old adults (>35) based on ageing criteria established from known elephant populations (Moss, 2001). Home ranges were produced using minimum convex polygons. Mean daily displacement is the average of the total distance moved during a 24-hour period for each individual.  

See the README file for a detailed description of column values.