Nathan P. Snow

Potential attractants for detecting and removing invading Gambian giant pouched rats (Cricetomys gambianus)

BACKGROUND Native to Africa, Gambian giant pouched rats (Gambian rats; Cricetomys gambianus Waterh.) are a threatening invasive species on a Florida island, Grassy Key. Gambian giant pouched rats shifted from a domestic pet to invading species after suspected release from a pet breeder. Because of the large size of Gambian rats (weighing up to 2.8 kg), they pose a serious threat to native species (particularly nesting species) and agricultural crops, especially if Gambian rats invade mainland Florida. Also, Gambian rats pose a threat from disease, as they were implicated in a monkeypox outbreak in the midwestern United States in 2003. The United States Department of Agricultures Wildlife Services has initiated eradication and detection efforts in the Florida Keys, but trapping the sparse population of Gambian rats has proven difficult. RESULTS Fifteen attractants that could be used in traps for capturing or detecting single or paired Gambian rats were tested. It was found that conspecific scents (i.e. feces and urine) from other Gambian rats were the best treatment for attracting single and paired Gambian rats. Single Gambian rats explored more attractant types than paired Gambian rats. CONCLUSIONS Effective attractants for use with Gambian rats have been identified, and multiple attractant types should be used to capture or detect the sparse population. It is recommended that mainly urine and feces from Gambian rats be used, but peanut butter, anise, ginger and fatty acid scent could also be useful for attracting the currently small population of Gambian rats on Grassy Key.

A field evaluation of a trap for invasive American bullfrogs

Native to the eastern United States, American bullfrogs (Rana catesbeiana (Lithobates catesbeianus)) have been introduced in many countries throughout the world. There have been relatively few effective and efficient control methods developed to manage bullfrogs. Particularly in the Hawaiian Islands, Pacific coast of North America, and Japan, finding effective methods for controlling invasive bullfrogs is needed with special emphasis on low impacts for sensitive native species. We conducted a field study to examine the efficacy of a newly designed live trap for capturing invasive bullfrogs. We found that our trap was successful at capturing bullfrogs because we captured up to seven in a single trap overnight. Fishing lures, live crickets, and lights were used as attractants and all capture bullfrogs, however more research is needed for finding effective attractants. We captured one known non-target frog that was released. Our findings suggest that the multiple capture traps could effectively be used as part of an integrated pest management strategy for controlling invasive bullfrog populations.

The effects of vitamin K1-rich plant foods on the efficacy of the anticoagulant rodenticides chlorophacinone and diphacinone, used against Montane Voles (Microtus montanus)

Voles can cause significant losses to agriculture and wood fibre production. California growers typically rely on baits containing chlorophacinone and diphacinone to reduce vole population densities, but the efficacy of those rodenticides has been decreasing. One hypothesis suggests that voles are consuming high levels of an antidote (vitamin K1) to the anticoagulants, contained within green leafy plants. We tested that hypothesis by first feeding Montane Voles (Microtus montanus) diets that were high in vitamin K1, and then providing the animals with either: (1) chlorophacinone-containing bait, (2) diphacinone-containing bait, or (3) a control diet. We found that the chlorophacinone-containing bait remained efficacious (100% mortality), whereas the diphacinone-containing bait had a much lower efficacy (60% mortality). When only the diphacinone-containing bait was presented, the efficacy was somewhat better (80%). We infer that a diet rich in vitamin K1 did not negate the effects of the chlorophacinone for voles, and so we recommend its continued use in California unless anticoagulant resistance is known to have developed in the vole population. We hypothesise that: (1) diphacinone has a relatively low efficacy against Montane Voles when compared to chlorophacinone, and (2) this lower efficacy could be further reduced by a vitamin K1-rich diet.

Machine learning to classify animal species in camera trap images: applications in ecology

Motion-activated cameras (“camera traps”) are increasingly used in ecological and management studies for remotely observing wildlife and have been regarded as among the most powerful tools for wildlife research. However, studies involving camera traps result in millions of images that need to be analyzed, typically by visually observing each image, in order to extract data that can be used in ecological analyses. We trained machine learning models using convolutional neural networks with the ResNet-18 architecture and 3,367,383 images to automatically classify wildlife species from camera trap images obtained from five states across the United States. We tested our model on an independent subset of images not seen during training from the United States and on an out-of-sample (or “out-of-distribution” in the machine learning literature) dataset of ungulate images from Canada. We also tested the ability of our model to distinguish empty images from those with animals in another out-of-sample dataset from Tanzania, containing a faunal community that was novel to the model. The trained model classified approximately 2,000 images per minute on a laptop computer with 16 gigabytes of RAM. The trained model achieved 98% accuracy at identifying species in the United States, the highest accuracy of such a model to date. Out-of-sample validation from Canada achieved 82% accuracy, and correctly identified 94% of images containing an animal in the dataset from Tanzania. We provide an R package (Machine Learning for Wildlife Image Classification; MLWIC) that allows the users to A) implement the trained model presented here and B) train their own model using classified images of wildlife from their studies. The use of machine learning to rapidly and accurately classify wildlife in camera trap images can facilitate non-invasive sampling designs in ecological studies by reducing the burden of manually analyzing images. We present an R package making these methods accessible to ecologists. We discuss the implications of this technology for ecology and considerations that should be addressed in future implementations of these methods.

Evaluation of movement behaviors to inform toxic baiting strategies for invasive wild pigs (Sus scrofa): Movement and baiting for invasive wild pigs

BACKGROUND Invasive wild pigs damage agriculture, property, and natural ecosystems. To curtail damage, an effective and humane toxic bait containing microencapsulated sodium nitrite is under development. Strategies for delivering the toxic bait are needed to establish adequate spacing of bait sites, and for simultaneously accustoming wild pigs to the novel bait and wild pig-specific bait stations designed to exclude non-target species. RESULTS We monitored movements of 32 Global Positioning System (GPS)-collared wild pigs relative to 41 bait sites containing placebo bait. Among the bait sites, we compared three experimental baiting strategies (and a control) to evaluate which strategy led to the most wild pigs accessing the placebo bait inside bait stations. We found that bait sites should be spaced 0.5-1 km apart to maximize opportunities for all wild pigs to find and utilize the bait sites. Baiting strategies that allowed ≥ 15 days for accustoming wild pigs to bait stations were most effective and resulted in nearly 90% of wild pigs accessing the placebo bait inside the bait stations. Bait stations excluded all non-target animals, except one instance with a raccoon (Procyon lotor). CONCLUSION These results demonstrate the potential for toxic bait to be an effective tool for reducing populations of wild pigs with minimal risks to non-target species, if optimized delivery procedures are followed.