Captive breeding programs are increasingly being utilized as a tool for preventing the imminent extinction of threatened and endangered species, or as a means for rebuilding at-risk populations. Many populations of Pacific salmon (Oncorhynchus spp.) along the west coast of the U.S. are currently listed as threatened or endangered under the U.S. Endangered Species Act. As a result, various programs have been initiated to supplement at-risk populations with hatchery-reared fish. Despite this widespread, and sometimes contentious, practice, there has been no large-scale effort to evaluate the efficacy of hatchery programs in rebuilding depleted populations. We used a multiple before-after, control-impact (MBACI) design to measure the response of 10 supplemented populations of threatened Chinook salmon (O. tshawytscha) from the Columbia River basin relative to 11 reference populations from the same basin that had never received any hatchery fish. The time period (1960-1999) of record covered 25 years prior to, and 15 years following, the intervention. Our comparison used two different response variables: (1) population productivity measured as the number of surviving adults per spawning (breeding) adult, and (2) the density of spawning adults. 

Results/Conclusions

Using the standard ANOVA design, we found no statistical difference in the productivity of populations following supplementation relative to the reference group. In the case of density, however, the data did not meet the a priori assumptions of the standard ANOVA design in that there were obvious downward trends in the time series (i.e. the primary reason the populations were “listed”). Therefore, we first fit a multivariate time series model to the reference-only populations to extract a common trend that we could use to describe common changes in habitat, hydropower development, fishing pressure, and climate. We then subtracted this common trend from all of the time series and repeated the assessment. Subsequently, we found no significant increase in the density of spawning adults. Although these analyses were not designed to identify the mechanism, our analyses do call into question whether hatchery supplementation programs are achieving their intended goal in this region.