Estimating herbivore abundance through fecal pellet counts in arid environments is complicated by the slow decay of feces, which results in mixed-age samples and uncertainty in population density estimates. We tested whether age-related changes in pellet color and texture due to desiccation could serve as reliable temporal markers to infer fecal age structure and improve the accuracy of density estimation. Using real-time observations from 70 pellet groups of black-tailed jackrabbit (Lepus californicus) monitored over 33-37 days in Baja California Sur, Mexico (2022, 2024), we developed logistic models that accurately captured desiccation dynamics across sun and shade conditions (R2 = 0.87-0.97). These models were then applied in a simulation framework with hypothetical pellet counts to estimate population density via inverse modeling. Despite some interannual and microhabitat variation, desiccation followed a consistent and quantifiable trajectory, enabling robust inference of recent fecal input. Additionally, we estimated and corrected for early-stage spoilage, pellets rendered unsuitable for monitoring before full desiccation, resulting in a 20-22% increase in density estimates. This highlights the importance of accounting for early pellet loss in field applications. Our findings demonstrate that pellet desiccation traits can be used to model freshness decay and support density estimation under dryland conditions. The approach refines the fecal standing crop method by integrating condition-specific decay modeling and spoilage-aware corrections, offering a practical and scalable tool for population monitoring where pellet persistence allows temporal inference from visible traits.