Testing the potential of streamflow data to predict spring migration of ungulate herds

Area of Science:

• Ecology
• Hydrology
• Wildlife Biology

Background:

• Migratory animals in mountains and high latitudes use vegetation green-up, linked to snowmelt and rising temperatures.
• Snowmelt creates a 'spring pulse' in rivers, measurable by streamgages with long-term data.
• Streamgage data offers potential for predicting animal migration timing.

Purpose of the Study:

• To test streamflow data's potential for predicting mule deer spring migration timing.
• To compare streamflow-based models with traditional temperature-based models.
• To analyze long-term trends in mule deer migration onset.

Main Methods:

• Used non-parametric linear regression to assess variable stationarity.
• Employed logistic regression to model migration initiation probabilities.
• Compared predictive performance of streamflow and temperature-informed models.

Main Results:

• Streamflow data models performed comparably to temperature data models for predicting past migration.
• Streamflow models showed nearly equal potential for forecasting future migrations.
• Analysis revealed a trend towards earlier spring migration in mule deer over 69 years.

Conclusions:

• Streamflow data is a viable predictor of animal migration timing, comparable to temperature data.
• Integrating hydrologic and biological data can enhance understanding of past and future migration patterns.
• Mule deer spring migration is occurring earlier, highlighting the need for adaptive conservation strategies.