David A. Patterson

Power of Sign Surveys to Monitor Population Trends

The urgent need for an effective monitoring scheme for grizzly bear (Ursus arctos) populations led us to investigate the effort required to detect changes in populations of low-density dispersed animals, using sign (mainly scats and tracks) they leave on trails. We surveyed trails in Glacier National Park for bear tracks and scats during five consecutive years. Using these data, we modeled the occurrence of bear sign on trails, then estimated the power of various sampling schemes. Specifically, we explored the power of bear sign surveys to detect a 20% decline in sign occurrence. Realistic sampling schemes appear feasible if the density of sign is high enough, and we provide guidelines for designs with adequate replication to monitor long-term trends of dispersed populations using sign occurrences on trails.

A Linear Programming Approach to Discriminant Analysis with a Reserved-Judgment Region

A linear-programming model is proposed for deriving discriminant rules that allow allocation of entities to a reserved-judgment region. The size of the reserved-judgment region, which can be controlled by varying parameters within the model, dictates the level of aggressiveness (cautiousness) of allocating (misallocating) entities to groups. Results of simulation experiments for various configurations of normal and contaminated normal three-group populations are reported for a variety of parameter selections. Results of cross-validation experiments using real data sets are also reported. Both the simulation and cross-validation experiments include comparison with other discriminant analysis techniques. The results demonstrate that the proposed model is useful for deriving discriminant rules that reduce the chances of misclassification, while maintaining a reasonable level of correct classification.

Comment on Cameron's censored logistic regression model for referendum data

Abstract T. A. Cameron (A new paradigm for valuing non-market goods using referendum data: Maximum likelihood estimation by censored logistic regression, J. Environ. Econom. Management 15, 355–379 (1988)) presents what she terms a new paradigm for interpreting referendum data derived from contingent valuation surveys for nonmarket resources. We show that since her model is a reparameterization of the usual logistic regression model, asymptotic standard errors can be derived for the maximum likelihood estimators of the parameters in either model from the other. We also argue that computational convenience in deriving demand relationships or elasticities depends more on the choice of welfare measure than on the choice of a parameterization.

Toward estimation of map accuracy without a probability test sample

The time and effort required of probability sampling for accuracy assessment of large-scale land cover maps often means that probability test samples are not collected. Yet, map usefulness is substantially reduced without reliable accuracy estimates. In this article, we introduce a method of estimating the accuracy of a classified map that does not utilize a test sample in the usual sense, but instead estimates the probability of correct classification for each map unit using only the classification rule and the map unit covariates. We argue that the method is an improvement over conventional estimators, though it does not eliminate the need for probability sampling. The method also provides a new and simple method of constructing accuracy maps. We illustrate some of problems associated with accuracy assessment of broad-scale land cover maps, and our method, with a set of nine Landsat Thematic Mapper satellite image-based land cover maps from Montana and Wyoming, USA.

Ideal bootstrap estimation of expected prediction error for k-nearest neighbor classifiers: Applications for classification and error assessment

Euclidean distance k-nearest neighbor (k-NN) classifiers are simple nonparametric classification rules. Bootstrap methods, widely used for estimating the expected prediction error of classification rules, are motivated by the objective of calculating the ideal bootstrap estimate of expected prediction error. In practice, bootstrap methods use Monte Carlo resampling to estimate the ideal bootstrap estimate because exact calculation is generally intractable. In this article, we present analytical formulae for exact calculation of the ideal bootstrap estimate of expected prediction error for k-NN classifiers and propose a new weighted k-NN classifier based on resampling ideas. The resampling-weighted k-NN classifier replaces the k-NN posterior probability estimates by their expectations under resampling and predicts an unclassified covariate as belonging to the group with the largest resampling expectation. A simulation study and an application involving remotely sensed data show that the resampling-weighted k-NN classifier compares favorably to unweighted and distance-weighted k-NN classifiers.

Positive effects of fire on birds may appear only under narrow combinations of fire severity and time-since-fire

We conducted bird surveys in 10 of the first 11 years following a mixed-severity fire in a dry, low-elevation mixed-conifer forest in western Montana, United States. By defining fire in terms of fire severity and time-since-fire, and then comparing detection rates for species inside 15 combinations of fire severity and time-since-fire, with their rates of detection in unburned (but otherwise similar) forest outside the burn perimeter, we were able to assess more nuanced effects of fire on 50 bird species. A majority of species (60%) was detected significantly more frequently inside than outside the burn. It is likely that the beneficial effects of fire for some species can be detected only under relatively narrow combinations of fire severity and time-since-fire. Because most species responded positively and uniquely to some combination of fire severity and time-since-fire, these results carry important management implications. Specifically, the variety of burned-forest conditions required by fire-dependent bird species cannot be created through the application of relatively uniform low-severity prescribed fires, through land management practices that serve to reduce fire severity or through post-fire salvage logging, which removes the dead trees required by most disturbance-dependent bird species.

Effects of wildfire on national park visitation and the regional economy: a natural experiment in the Northern Rockies

Federal wildland fire management policy in the United States directs the use of value-based methods to guide priorities. However, the economic literature on the effect of wildland fire on nonmarket uses, such as recreation, is limited. This paper introduces a new approach to measuring the effect of wildfire on recreational use by utilising newly available long-term datasets on the location and size of wildland fire in the United States and observed behaviour over time as revealed through comprehensive National Park Service (NPS) visitor data. We estimate travel cost economic demand models that can be aggregated at the site-landscape level for Yellowstone National Park (YNP). The marginal recreation benefit per acre of fire avoided in, or proximate to, the park is US

Modeling the behavior of marlin anglers in the Western Pacific.

43.82 per acre (US

Discriminant analysis applied to aerial radiometric data and its application to Uranium favorability in South Texas

108.29 per hectare) and the net present value loss for the 1986-2011 period is estimated to be US

A three-population constrained discrimination procedure

206 million. We also estimate marginal regional economic impacts at US