Evaluating a Fish Monitoring Protocol Using State-Space Hierarchical Models

Robin E. Russell1, *, David A. Schmetterling2, Chris S. Guy3, Bradley B. Shepard2, #, Robert McFarland1, Donald Skaar4
1 Montana Fish, Wildlife and Parks, 3201 Spurgin Road, Missoula, MT 59801, USA
2 Montana Fish, Wildlife and Parks, 1400 S. 19th Ave, Bozeman, MT 59718, USA
3 US Geological Survey, Montana Cooperative Fishery Research Unit, Montana State University, Bozeman, MT 59715, USA
4 Montana Fish, Wildlife and Parks, 1420 East 6th Ave, Helena, MT 59620, USA

Article Metrics

CrossRef Citations:
Total Statistics:

Full-Text HTML Views: 419
Abstract HTML Views: 551
PDF Downloads: 0
Total Views/Downloads: 970
Unique Statistics:

Full-Text HTML Views: 255
Abstract HTML Views: 351
PDF Downloads: 0
Total Views/Downloads: 606

© 2012 Russell et al.

open-access license: This is an open access article distributed under the terms of the Creative Commons Attribution 4.0 International Public License (CC-BY 4.0), a copy of which is available at: This license permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

* Address correspondence to this author at US Geological Survey, 6006 Schroeder Rd, Madison WI, 53711, USA; Tel: 608-2742474; Fax: 608-270-2415; E-mail:
* Current address: Wildlife Conservation Society, 301 North Willson Avenue, Bozeman, Montana 59715; USA


Using data collected from three river reaches in Montana, we evaluated our ability to detect population trends and predict fish future fish abundance. Data were collected as part of a long-term monitoring program conducted by Montana Fish, Wildlife and Parks to primarily estimate rainbow (Oncorhynchus mykiss) and brown trout (Salmo trutta) abundance in numerous rivers across Montana. We used a hierarchical Bayesian mark-recapture model to estimate fish abundance over time in each of the three river reaches. We then fit a state-space Gompertz model to estimate current trends and future fish populations. Density dependent effects were detected in 1 of the 6 fish populations. Predictions of future fish populations displayed wide credible intervals. Our simulations indicated that given the observed variation in the abundance estimates, the probability of detecting a 30% decline in fish populations over a five-year period was less than 50%. We recommend a monitoring program that is closely tied to management objectives and reflects the precision necessary to make informed management decisions.

Keywords: Bayesian, Gompertz models, Montana, Oncorhynchus mykiss, Salmo trutta, Trout.