Recovery Plan for the Northern Spotted Owl - DRAFT

however, we know little about the behavior of these nonterritorial birds, sodetermining their importance in maintaining stable populations is difficult.Monitoring a small sample of birds, perhaps followed continuously afterbeing banded as fledglings, could provide much of the needed information.The critical question is how widely and continuously they search for vacanciesin territories. Other issues, such as how their survival rates compare toterritorial adults, should also be investigated.As this information is obtained, it will be important to challenge the model withtests to reveal the reliability of its predictions. For example, predictions by themodel of how dispersing juveniles will move across real landscapes can becompared with actual movements by transmittered birds. The environment in1950, when public land was largely covered by old-growth and presumably hadrather continuous populations of owls, can be reproduced in the computer andthen altered during a 40-year simulation to resemble its current state. Distributionof owls at the end of the simulation then can be compared to actualdistribution. Other tests of this sort can be carried out, to increase our knowledgeof what the model can, and cannot, reveal about owl trends. Three broadoutcomes of this work might be distinguished: the model may not make anyreliable predictions, it may permit an "ordering" of trends under differentproposals, or it may permit estimates of trends of sufficient accuracy to beuseful. It is too early for confidence about how well the landscape model willperform, but it appears to have a better chance of predicting order, and perhapsabsolute trends, than any other approach for modeling populationdynamics of spotted owls.Early Warning MethodsTwo other methods for making inferences about population trends warrantconsideration. Neither will provide reliable estimates of trends, but both mayprovide an "early warning" that population declines are imminent.The first method is monitoring the age of first-time territory holders. Therationale of the method is that if the nonterritorial population is large, thenmany of the first-time territory holders probably will not be young (1- or 2-yearold)birds. However, if all adults hold territories then all first-time territoryholders must be young birds (unless adults move after establishing a territorywhich is thought to be uncommon). These facts suggest that as thenonterritorial population declines, the proportion of first-time territory holdersthat are young birds eventually must increase to 1.0. Accordingly, monitoringthis proportion might provide a basis for crude inferences about thenonterritorial population. If the ratio was 1.0 in one population and 0.3 inanother, we might reasonably infer that the first population had morenonterritorial birds, and if the proportion began increasing, we might suspectthat the overall population was declining even if the number of territorial birdsremained stable.The exact relationship between the proportion of first-time territory holdersthat are young and either the size of the nonterritorial population or the rate ofpopulation change appears complex and difficult to model. Computer simulations,however, can be used to indicate general trends. We have prepared sucha simulation and used it to investigate the behavior of this age ratio. Ifnonterritorial birds are assumed to have wide 'knowledge" of vacancies, andparticularly if older birds are assumed to have a competitive advantage infilling vacancies, then the age ratio, in a population undergoing overall decline,remains rather constant until only a year or two before a decline begins in theterritorial population. If each floater only fills vacancies occurring within asmall area then the proportion drifts upwards much more slowly as population268