An expert-based system to predict population survival rate from health data

Area of Science:

• Veterinary science
• Wildlife conservation
• Marine mammal health

Background:

• Effective wildlife management requires timely detection of population decline causes.
• Assessing population size trends is standard, but monitoring health may be more effective.
• Bottlenose dolphin populations face threats necessitating innovative conservation strategies.

Purpose of the Study:

• To develop a method for estimating survival probability using health measures in bottlenose dolphins.
• To compare survival estimates from a health-based model with traditional capture-mark-recapture methods.
• To identify specific health indicators predictive of short-term mortality in dolphins.

Main Methods:

• Collated health data from 7 bottlenose dolphin populations in the southeastern U.S.
• Utilized logistic regression and a Bayesian framework to implement the veterinary expert system for outcome prediction (VESOP).
• Employed capture-mark-recapture (CMR) analyses for comparative survival estimates.

Main Results:

• Health measures, especially inflammation markers, predicted 1- and 2-year survival.
• Low alkaline phosphatase indicated high 1-year mortality risk (OR=10.2); elevated globulin indicated high 2-year mortality risk (OR=9.60).
• VESOP-predicted population survival rates strongly correlated with CMR estimates (1-year r=0.99, 2-year r=0.94).

Conclusions:

• Monitoring animal health, particularly inflammation, can predict mortality risk and inform conservation.
• The VESOP approach offers a viable alternative to traditional methods for assessing wildlife population health and survival.
• Advancements in remote sampling could enhance the application of this health-monitoring approach to other species.