Fine-scale body condition dynamics in elk revealed by implanted bioimpedance biologgers

Body condition shapes ungulate survival and reproduction, but temporally sparse capture-based measurements limit inference about when individuals gain or lose fat reserves in response to environmental change. We evaluated implanted biologgers that measure bioimpedance, based on the electrical conductivity of tissue, as a continuous signal of body condition dynamics in captive elk. Adult female elk (n = 4) were implanted with subcutaneous biologgers that recorded bioimpedance every 4 h during a ~6-month feeding experiment inducing gradual condition loss and recovery. We quantified correspondence between bioimpedance and maximum rump fat thickness measured approximately monthly, lagged relationships between daily caloric intake and bioimpedance and whether 4-h bioimpedance variability was associated with camera-derived feeding and drinking time budgets. Weekly bioimpedance showed strong within-individual correlation with maximum rump fat thickness, with Spearman correlations of 0.75-0.96 during dietary restriction (mean = 0.90) and 0.96-0.98 during the normal feeding schedule for three elk (mean = 0.97). Bioimpedance also covaried with caloric intake, but the timing of the association varied among elk: correlations between first differences of smoothed daily calories and smoothed daily impedance peaked at lags spanning 0-60 days, with the fattest individual showing the minimum lag (0 day). Coupling was more consistent during restriction than during the gain phase, suggesting reserve loss may act faster than rebuilding. At the finest scale, feeding and drinking behaviour explained only a portion of 4-h impedance variation and was most consistent with multi-day (72 h) summaries. Together, these results indicate that biologger-derived bioimpedance provides a continuous, within-individual indicator of condition dynamics across hours to weeks. Although additional calibration will be needed to translate raw impedance into absolute fat estimates across animals, this approach enables inference about the timing, rate and extrema of fat-reserve change, with applications for monitoring nutritional status and evaluating habitat and management outcomes in free-ranging wildlife.