Adana Dewlap pigeons, native to southern Turkey, are renowned for their distinctive skeletal structure, striking appearance, and cultural significance. While breeders and enthusiasts have valued them for generations, their biological growth processes have remained largely unexplored. Our recent study 

bridges this gap by applying both classical and advanced statistical models to capture the hidden dynamics of pigeon growth.

Why Study Pigeon Growth?

Growth models ar

e central to animal science: they inform breeding programs, optimize feeding strategies, and 

p

rovide insights into physiological development. For pigeons, especially rare regional breeds like th

e Adana Dewlap, understanding growth dynamics is also a matter of conservation and cultu

ral preservation.

From Classi

cal Models to Adaptive Filtering

We collected 43-day body weight measurements from 88 pigeon chicks over seven years. Traditional models such as the Richards, Gompertz, and Logistic functions were tested, each offering a mathematical description of the typical S-shaped growth curve. Among them, the Richards model provided the best static fit.

However, biological systems are rarely static. Measurement noise, environmental variability, and genetic diversity demand a more flexible approach. This is where the Adaptive Kalman Filter (AKF) comes into play. Unlike static models, AKF can track weight (position), rate of weight gain (velocity), and changes in growth speed (acceleration) in real time.

What Did We Find?

- Rapid early growth: During the first two weeks, velocity was highest, reflecting a critical post-hatch development stage.
- Slowing phase: Growth decelerated gradually, suggesting genetic and metabolic constraints.
- Transition detection: AKF successfully identified the shift from rapid growth to the asymptotic plateau, outperforming static models in accuracy.

Quantitatively, the AKF reduced mean squared error (MSE) and improved predictive accuracy compared to all classical models.

Why Does It Matter?

The findings show that growth is not just about final size but about the trajectory of how size changes over time. With AKF, we gain a layered perspective—revealing latent transitions invisible to traditional methods.

Such insights matter for:
- Precision poultry farming: real-time monitoring of growth for improved animal welfare and productivity.
- Veterinary diagnostics: early detection of growth anomalies.
- Biological sciences: a methodological advance that can be extended to other species.

Looking Ahead

Our study demonstrates how combining classical growth models with adaptive filtering offers both accuracy and interpretability. Future work will explore nonlinear extensions, machine learning integrations, and cross-species applications.

By focusing on a culturally significant yet understudied breed, the Adana Dewlap pigeon, this research not only enriches avian science but also highlights the value of applying advanced statistical tools to biological questions.
I welcome your thoughts and examples of where adaptive filtering has helped uncover hidden dynamics in your own fields.

Levent Özbek – Ankara University, Faculty of Science, Department of Statistics