Dear MARMAM colleagues,
Detecting population decline before it's too late: What dolphin teeth revealed 
about a hidden crisis
We're excited to share our new study in Conservation Letters demonstrating the 
first evidence of declining viability in common dolphins (Delphinus delphis) in 
the Bay of Biscay. Using age-at-death data from 759 stranded specimens 
(1997-2019), we documented a dramatic decline in female longevity from 24 to 17 
years - representing 2-3 fewer breeding events per female and a 2.4% reduction 
in population growth rate. This occurred despite stable abundance estimates, 
highlighting the Bay of Biscay as a demographic sink.
The challenge: Making the most of stranding data
Many of us collect teeth from stranded marine mammals and invest significant 
resources in aging specimens. Yet analyzing these cross-sectional age-at-death 
data has traditionally required restrictive assumptions (particularly 
stationarity) or complex models that are difficult to fit and don't easily 
accommodate covariates or allow temporal trend detection. This often means 
pooling data across years and potentially missing critical early warning 
signals of population decline.
Our solution: A flexible, accessible framework
We developed a pragmatic modeling approach using stratified random sampling and 
a flexible linear regression framework that can:

  *   Detect temporal trends in survivorship without assuming stationarity
  *   Work with realistic sample sizes (accurate estimates with n≥300)
  *   Accommodate individual covariates (sex, area, etc.)
  *   Provide operational demographic indicators for management

The complete methodology is published and fully reproducible, with R code 
available on GitHub. Our Monte Carlo validation study (Ecological Modelling 
2024) demonstrates the approach works even with sparse data and model 
misspecification.
Full text (open access): Rouby et al. 2025. Conservation Letters: 
https://doi.org/10.1111/conl.13142
Methods paper:
Rouby et al. 2024. Ecological Modelling: 
https://doi.org/10.1016/j.ecolmodel.2024.110647
Code repository: https://github.com/erouby/dolphin_viability_cons_letter
We hope this framework proves useful for others working with stranding data to 
assess marine mammal population viability. We welcome questions and feedback!
Best regards,
Etienne Rouby
---
Etienne Rouby, Ph.D.
Postdoctoral Associate, University of Colorado Boulder
Traditional Territory of the Arapaho, Cheyenne, and Ute
www.etiennerouby.com<http://www.etiennerouby.com>
Guest Editor BMC Environmental Science
[cid:133d5c3c-4987-4507-bc95-e4067e9971f5]
_______________________________________________
MARMAM mailing list
[email protected]
https://lists.uvic.ca/mailman/listinfo/marmam

Reply via email to