Recovery Plan for the Northern Spotted Owl - DRAFT

Recovery Plan for the Northern Spotted Owl - DRAFT Recovery Plan for the Northern Spotted Owl - DRAFT

10.07.2015 Views

Conclusionsthe year effect can be ignored, and that formulas appropriate with simplerandom sampling are used in the analyses.Little difference occurs in sample size requirements if the proportions arebetween approximately 0.3 and 0.7. Productivity (which can probably bemodeled as a proportion since production of two female fledglings by a pair isreasonably rare) and the age ratio of first-time breeders are both probably inthe 0.3 to 0.7 range. Suppose sites were visited for 5 years. The number ofsites per group per year and associated 90 percent confidence intervals forestimated differences are approximately as follows: 50 sites, 0.09; 100 sites,0.06: 200 sites, 0.044; and 400 sites, 0.03 1. For example, if 100 sites in eachof two groups were each visited for 5 years, then the confidence interval for thedifference would be approximately 0.044. Average productivities of 0.30 and0.35, would be significantly different. In general, the sample size required todetect differences among groups would probably vary from 50 to 200 dependingon the level of precision required. For estimating or comparing adultsurvival rates, a larger sample might be needed because, while the standarderror is smaller with large probabilities, the difference one might wish to detectalso would be smaller.The brief analysis just discussed suggests that a reasonable target for theactivity site monitoring program would be 200 sites per year in each state (40to 60 sites per province). We-assume that these would be in addition to sites inthe density study areas. Such a program would permit numerous investigationsof how productivity, adult survival, age ratios, and replacement timesvaried across habitats or other environmental features, across the range, andthrough time. The results would be of great value in developing the populationmodeling approach, they would probably be useful in the capture-recaptureestimates, and they would provide confidence that populations were healthy orhelp identify problems if any occurred. The cost, 5 person-years per state for 5years, seems acceptable in light of current expenditures for monitoring activities.We stress again, however, that these are 'order of magnitude" estimatesintended solely to give a first impression of how large the sample sizes shouldbe. More detailed estimates can be provided after detailed objectives, and somepreliminary results, are available.Sample size guidelines for the other studies and cannot be provided until thespecific objectives and methods have been identified.The analyses and recommendations in this appendix are intended to provide aframework and a starting place for designing the monitoring program. Ourmain goal has been to demonstrate that such a program could be useful andcould satisfy the requirement in the Delisting Criteria for a "scientificallycredible sampling plan" at acceptable costs. Even if our estimates turn out tobe low, it seems clear that the monitoring program described could be carriedout at costs substantially less than those currently being incurred for spottedowl monitoring programs.278

Literature CitedBalser, J.P. 1981. Confidence interval estimation and tests for temporaryoutmigration in tag-recapture studies. Unpublished PhD Dissertation CornellUniversity, Ithaca, New York.Franklin, A. B. In press. Population regulation in northern spotted owls:theoretical implications for management. In D. McCullough and R. Barrett. ed.Wildlife 2001: Populations. Elsevier Press, Essex, England.McKelvey, K. 1991. A spatially explicit life-history simulator for the northernspotted owl. USDA Forest Service, Pacific Southwest Forest and RangeExperiment Station.Thomas, J.W., E.D. Forsman, J.B. Lint, E.C. Meslow, B.R. Noon, and J.Verner. 1990. A conservation strategy for the northern spotted owl. Report ofthe Interagency Scientific Committee to address the conservation of the northernspotted owl. Forest Service, Fish and Wildlife Service, Bureau of LandManagement, and National Park Service, Portland, Oregon, U.S. GovernmentPrinting Office. 1990-791-171/20026.Yates, R. 1960. Sampling methods for censuses and surveys. 3rd ed.London:Charles Griffin and Company.279

Conclusions<strong>the</strong> year effect can be ignored, and that <strong>for</strong>mulas appropriate with simplerandom sampling are used in <strong>the</strong> analyses.Little difference occurs in sample size requirements if <strong>the</strong> proportions arebetween approximately 0.3 and 0.7. Productivity (which can probably bemodeled as a proportion since production of two female fledglings by a pair isreasonably rare) and <strong>the</strong> age ratio of first-time breeders are both probably in<strong>the</strong> 0.3 to 0.7 range. Suppose sites were visited <strong>for</strong> 5 years. The number ofsites per group per year and associated 90 percent confidence intervals <strong>for</strong>estimated differences are approximately as follows: 50 sites, 0.09; 100 sites,0.06: 200 sites, 0.044; and 400 sites, 0.03 1. For example, if 100 sites in eachof two groups were each visited <strong>for</strong> 5 years, <strong>the</strong>n <strong>the</strong> confidence interval <strong>for</strong> <strong>the</strong>difference would be approximately 0.044. Average productivities of 0.30 and0.35, would be significantly different. In general, <strong>the</strong> sample size required todetect differences among groups would probably vary from 50 to 200 dependingon <strong>the</strong> level of precision required. For estimating or comparing adultsurvival rates, a larger sample might be needed because, while <strong>the</strong> standarderror is smaller with large probabilities, <strong>the</strong> difference one might wish to detectalso would be smaller.The brief analysis just discussed suggests that a reasonable target <strong>for</strong> <strong>the</strong>activity site monitoring program would be 200 sites per year in each state (40to 60 sites per province). We-assume that <strong>the</strong>se would be in addition to sites in<strong>the</strong> density study areas. Such a program would permit numerous investigationsof how productivity, adult survival, age ratios, and replacement timesvaried across habitats or o<strong>the</strong>r environmental features, across <strong>the</strong> range, andthrough time. The results would be of great value in developing <strong>the</strong> populationmodeling approach, <strong>the</strong>y would probably be useful in <strong>the</strong> capture-recaptureestimates, and <strong>the</strong>y would provide confidence that populations were healthy orhelp identify problems if any occurred. The cost, 5 person-years per state <strong>for</strong> 5years, seems acceptable in light of current expenditures <strong>for</strong> monitoring activities.We stress again, however, that <strong>the</strong>se are 'order of magnitude" estimatesintended solely to give a first impression of how large <strong>the</strong> sample sizes shouldbe. More detailed estimates can be provided after detailed objectives, and somepreliminary results, are available.Sample size guidelines <strong>for</strong> <strong>the</strong> o<strong>the</strong>r studies and cannot be provided until <strong>the</strong>specific objectives and methods have been identified.The analyses and recommendations in this appendix are intended to provide aframework and a starting place <strong>for</strong> designing <strong>the</strong> monitoring program. Ourmain goal has been to demonstrate that such a program could be useful andcould satisfy <strong>the</strong> requirement in <strong>the</strong> Delisting Criteria <strong>for</strong> a "scientificallycredible sampling plan" at acceptable costs. Even if our estimates turn out tobe low, it seems clear that <strong>the</strong> monitoring program described could be carriedout at costs substantially less than those currently being incurred <strong>for</strong> spottedowl monitoring programs.278

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!