Duck Band Recovery Models and Data Analysis

Eric Patterson

Eric Patterson

Moderator
Staff member
Ever wonder about what happens when you report a band? Ever wonder what the USF&WS does with that data? Ever wonder what statistics are generated from your band report?

Well, there just happens to be a handbook on analyzing band data that can satisfy your curiosity. Maybe! Assuming you understand statistics at the graduate level. Bummer. It's interesting reading but not easy. Skipping past the math stuff you can still gain insight to what analysis is performed. I did bounce this document off the methods in the AHM report and they are different, so I suspect this document covers models used in the past or maybe pre-input to AHM. But when you read the models and assumptions there is definitely insight to be gained.

STATISTICAL INFERENCE FROM BAND RECOVERY
 
Last edited:
Just had a flashback to Bio Statistic 520.
I got a C in that class and was ecstatic.

A win's a win Carl!
 
My brother's father-in-law, Lester, was a senior at Miss State and he put off his mandatory algebra class to his last semester. Going into the final he was failing, meaning he would not graduate. It was a large auditorium class and the exam was on the last day of exams. This was before computers aided Profs with grading and reporting. Given the prof's workload and short time to grade and enter grades in the university system he announces to the class while he's handing out the final exam that anyone who wanted to skip the final would automatically receive a D for the class. Lester said he about knocked folks down racing for the auditorium door!
 
Lmao. A "D" for skipping the final, hmmm. Why attend at all?

Eric, you are probably the only one here who would read this for fun. AHM uses bayesian hierarchal models to predict population size but survival rate and recovery rates from banding data is used in the integrated population modeling.

To give a very simplistic summary, you can estimate the survival rate of ducks from bands by knowing how many bands were put on ducks and how many are recovered. If this is done carefully and every year it's possible to determine the survival rates of each age and sex cohort (adult males, adult females, juvenile males and juvenile females). This information can be used to assess population response to hunting and other factors. I am about at the end of my ability to understand and explain....but I know there a number of really smart and dedicated people running these analyses that back up the hunting regulations.
 
The current telemetry studies are amazing. As technology advances the data collection for waterfowl is so cool to follow.

I don't hunt for bands but after 48 years of hunting ducks I have never harvested a migratory bird with a band. I have harvested local two Blackbellied Whistling ducks in S. La and a local Canada Goose here in SC with a band but not one migratory duck. Both of my sons have harvested banded ducks and one Snow Goose. Not that it matters, just thought y'all might find that interesting.
 
Brad

You may be correct about me reading it for fun. Although the years of my career where I actually develop mathematical models are behind me, I do enjoy seeing models from other fields, especially when they are related to ducks. One thing I came to the conclusion of many years ago is that the best way to truly develop an understanding of a model is to write code for it. Are the developers and maintainers of these models on staff? Or are these folks from educational institutions with grants? Or something else?
 
Eric, I believe all are government employees either in USFWS or USGS. Most write in code in R while 30 years ago we used SAS.
 
Brad

There has been a huge transition away from SAS, SPSS, Minitab, etc. to open-source R across industry. I don't do stat programming anymore but did pick up a few books on R and installed it on my machines. I'm just waiting for an excuse to learn it because I want to be hip and cool again :) I do hope to add that skill before I retire because in retirement I plan to fiddle with models for fun. Matlab is king in the lab I presently support and I'm still an old school perl programmer. Nearly everyone has moved to Python but I'm sticking with perl. I've yet to see anything Python can do that can't be done in perl just as easily. Plus, there is A LOT of legacy code in perl and occasionally I'm tasked to figure things out the Python whippersnappers can't. I love that work when I get it.
 
You must log in or register to reply here.
Back
Top