Article

Generalized site occupancy models allowing for false positive and false negative errors.

U.S. Geological Survey, Patuxent Wildlife Research Center, Laurel, Maryland 20708, USA.
Ecology (Impact Factor: 5). 05/2006; 87(4):835-41. DOI: 10.1890/0012-9658(2006)87[835:GSOMAF]2.0.CO;2
Source: PubMed

ABSTRACT Site occupancy models have been developed that allow for imperfect species detection or "false negative" observations. Such models have become widely adopted in surveys of many taxa. The most fundamental assumption underlying these models is that "false positive" errors are not possible. That is, one cannot detect a species where it does not occur. However, such errors are possible in many sampling situations for a number of reasons, and even low false positive error rates can induce extreme bias in estimates of site occupancy when they are not accounted for. In this paper, we develop a model for site occupancy that allows for both false negative and false positive error rates. This model can be represented as a two-component finite mixture model and can be easily fitted using freely available software. We provide an analysis of avian survey data using the proposed model and present results of a brief simulation study evaluating the performance of the maximum-likelihood estimator and the naive estimator in the presence of false positive errors.

0 Bookmarks
 · 
90 Views
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: As freshwater habitats are among the most endangered, there is an urgent need to identify critical areas for conservation, especially those that are home to endangered species. The Pyrenean desman (Galemys pyrenaicus) is a semi-aquatic mammal whose basic ecological requirements are largely unknown, hindering adequate conservation planning even though it is considered as a threatened species. Species distribution modelling is challenging for freshwater species. Indeed, the complexity of aquatic ecosystems (e.g., linear and hierarchical ordering) must be taken into account as well as imperfect sampling. High-quality and relevant hydrological descriptors should also be used. To understand the influence of environmental covariates on the occupancy and detection of the Pyrenean desman, we combine both a robust sign-survey data set (i.e. with genetic validation ensuring true presence information) and a hydrological model to simulate the flow regime across a whole catchment. Markovian site-occupancy analysis, taking into account sign detection and based on spatially adjacent replicates, indicated a positive influence of heterogeneity of substrate and shelters, and a negative influence of flow variability on Pyrenean desman detection. This valuable information should help to improve monitoring programs for this endangered species. Our results also highlighted a spatially clustered distribution and a positive influence of stream flow and number of tributaries on occupancy. Hence, modifications of flow regime (e.g. hydropower production, irrigation, climate change) and habitat fragmentation appear to be major threats for this species, altering the connectivity between tributaries and the mainstream river as well as between adjacent sub-catchments.
    Biological Conservation 04/2015; 184. DOI:10.1016/j.biocon.2015.01.019 · 4.04 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Data quality is a common source of concern for large-scale citizen science projects like eBird. In the case of eBird, a major cause of poor quality data is the misiden-tification of bird species by inexperienced contributors. A proactive approach for improving data quality is to discover commonly misidentified bird species and to teach inexperienced birders the differences between these species. To accomplish this goal, we develop a latent variable graphical model that can identify groups of bird species that are often confused for each other by eBird participants. Our model is a multi-species extension of the classic occupancy-detection model in the ecology literature. This multi-species extension requires a structure learning step as well as a computationally expensive parameter learning stage which we make efficient through a variational approximation. We show that our model can not only discover groups of misidentified species, but by including these misidentifications in the model, it can also achieve more accurate predictions of both species occupancy and detection.
    Proceedings of the 28th Conference on Artificial Intelligence (AAAI), Quebec City, Quebec, Canada; 07/2014
  • [Show abstract] [Hide abstract]
    ABSTRACT: Automated recorders and occupancy models can be used together to monitor population trends of multiple avian species across a large geographic region. Automated recorders are an attractive method for monitoring birds, because they leave a record that can be independently validated and multiple units can be programmed to repeatedly survey different locations at the same daily times. We assessed the use of automated recorders and single-species, single-season occupancy models to monitor common forest birds across a 5.4-million-ha region of northern California. Using a survey protocol of 5-minute recordings at 3 times of the morning repeated over 3 consecutive days at 453 sites, we detected 32 species at >10% of these sites. Five of these species (Steller's jay [Cyanocitta stelleri], mountain chickadee [Poecile gambeli], red-breasted nuthatch [Sitta canadensis], dark-eyed junco [Junco hyemalis], and western tanager [Piranga ludoviciana]) were dominant with occupancies >0.5. We also modeled occupancy associations with elevation and canopy cover for brown creeper (Certhia americana), MacGillivray's warbler (Geothlypis tolmiei), and western tanager and found the environmental conditions at which occupancy was maximized differed by up to 399 m in elevation and 17.9% canopy cover for these species. Given a sampling effort of 100 new sites per year, we demonstrated 80% power (α = 0.1) to detect occupancy declines as small as 2.5% per year over 20 years for the 32 most common species. The effective radius of automated recorder surveys was approximately 50 m. In a field test, surveys conducted concurrently using automated recorders and point counts yielded similar occupancy estimates despite differences in detection probability. Our results suggest that automated recorders, used alone or in conjunction with point counts, can provide a practical means of monitoring common forest birds across a large geographic area. © 2014 The Wildlife Society.
    Journal of Wildlife Management 12/2014; DOI:10.1002/jwmg.821 · 1.64 Impact Factor