Endangered Species III

Contributed Oral

Informing Monarch Butterfly Conservation Strategies Using Deployable Computer Vision Technologies
Thomas Chen

The famous monarch butterflies (Danaus plexippus) and their migration patterns are an iconic phenomenon of the natural world. The species’ unique behavior has warranted a variety of in-depth studies over the years. In recent years, the monarch butterfly’s iconic migration patterns have come under threat from a number of factors, including climate change, deforestation, droughts, loss of habitat, and loss of milkweed. To track trends in their populations, scientists as well as citizen scientists must identify individuals accurately. This is key because there exist other species of butterfly, such as viceroy butterflies (Limenitis archippus and soldier butterflies (Danaus eresimus), that are “look-alikes,” having similar phenotypes (external appearances). To tackle this problem and to aid in more efficient identification, we present MonarchNet, the first comprehensive dataset consisting of butterfly imagery for monarchs and five look-alike species. We train a baseline deep-learning classification model (convolutional neural network) to serve as a tool for differentiating monarch butterflies and its various look-alikes. The input consists of a labeled butterfly image, with the species and bounding box coordinates labels. The output is a digit representing the predicted species of the butterfly in the image. The model architecture is ResNet50, pretrained on ImageNet data. The criterion for optimization is the cross-entropy loss function. We train on a randomly selected 80% of the dataset with a batch size of 32. The Adam optimizer with a learning rate of 0.01 is utilized, achieving a weighted-F1 score of 0.824. We seek to contribute to the study of biodiversity and butterfly ecology by providing a novel method for computational classification of these particular butterfly species. The ultimate aim is to help scientists track monarch butterfly population and migration trends in the most precise and efficient manner possible.

Long-Term Demography of a Reintroduced and Isolated Population of Endangered Falcons
Brian Rolek, Benjamin Pauli, Alberto Macias-Duarte, Brian Mutch, Paul Juergens, Tim Anderson, Chris Parish, Jeff Johnson, Brian Millsap, Christopher McClure, Leah Dunn

We examined demography of an endangered subspecies, the Northern Aplomado Falcon (Falco femoralis septentrionalis), in South Texas, USA. The Peregrine Fund has intensively managed and monitored this population since reintroductions began in 1993. Data spanning 1993–2018 enabled us to build an integrated population model to estimate survival for all life stages, abundance, fecundity, immigration, and emigration. Models revealed that male falcons survived at lower rates than females during their first year; Hurricane Harvey caused a decline in survival rates; and fecundity increased after 2012 coinciding with improvements in management. Both immigration and emigration were negligible indicating an isolated population that probably lacked rescue effects from neighboring populations. Population growth rates were greatest during years having greater numbers of released captive-bred young and greater probabilities of breeder survival. Importantly, breeder survival declined sharply during 2006 and remained lower until 2008; however, the cause of decline in breeder survival was unknown. Determining the cause of this low breeder survival period could be important for informing the management of populations. Our study greatly improves knowledge and understanding of demographics for a reintroduced, isolated, and intensively managed population of Northern Aplomado Falcons. This modeling framework will enable adaptive monitoring and management of the South Texas population by providing annual evaluations of vital rates, and will allow a nimble assessment and deployment of monitoring and management. Lastly, these demographic estimates can be used to inform population viability analyses and evaluate the relative persistence of populations.

An Assessment of the Behavior of Terrestrial Rodents in Marshes of the San Francisco Estuary during Extreme Tides
Katie Smtih

The marshes of the San Francisco Estuary are home to several terrestrial rodents, including the endangered salt marsh harvest mouse (Reithrodontomys raviventris, SMHM) which is endemic there. Over 90% of tidal wetlands in the estuary have been lost or converted, with the marshes remaining altered, fragmented, and reduced in size. Further, many are now bordered by cities, roads, steep levees, or other unnatural features, dramatically reducing available adjacent upland habitat. It has long been assumed that SMHM must move into higher elevation habitat fringing marshes (and do so en masse) during the most extreme tides. Tidal restoration strategies rely heavily upon this assumption despite a lack of comprehensive evidence supporting it. In the first study to specifically investigate high tide refugia throughout the range of the species I tested how far into “uplands” SMHM move during extreme tide events. I was surprised to find that not only was there not a strong directed movement by SMHM toward higher elevation areas during extreme flooding, only one individual moved completely out of a tidal wetland during the king tides that were examined, while competing rodent species were more commonly found in the “uplands.” Most SMHM remained on the flooded marsh plain, indicating a that strategy focusing primarily on upland enhancements to support the species in the face of sea level rise may be inadequate without also incorporating habitat complexity on the marsh plain, and that the larger community should be considered when implementing recovery actions. Further, with very little adjaent habitat available to accomodate marsh migration in response to sea level rise, refugia on the marsh plain will be essential for allowing SMHM to persist in the face of the more frequent and extreme flooding that is predicted to drown many marshes in the estuary.

Habitat Selection of Ocelots Using Lidar-Derived Canopy Metrics
Maksim Sergeyev, Michael Tewes, Daniel Crawford, Jason “Jay” Lombardi, Tyler Campbell

Many species depend on specific features on the landscape to persist and as such, characterizing available vegetation can be essential for managing these specialist species. One such habitat specialist is the ocelot (Leopardus pardalis), a medium-sized wild felid keenly adapted to dense thornshrub vegetation. Availability of this habitat in South Texas has plummeted due to agriculture and urbanization, leading to population declines and an eventual listing as federally endangered. As such, quantifying the characteristics of vegetation most suitable for ocelots has become an essential aspect to conservation of the species. Light detection and ranging (LiDAR) differs from other remote sensing techniques in its ability to penetrate through the canopy surface and describe the inner structure of the vegetation. Using LiDAR, we quantified overall canopy height and percentage cover at 1 m increments. We captured 8 ocelots on the East Foundation’s El Sauz Ranch in Southern Texas and fitted individuals with global positioning system (GPS) collars that recorded locations every 30 minutes. We used a step selection function to determine selection of canopy height and cover. Ocelots selected for greater canopy cover at lower heights (1 – 2m above ground). By combining accurate, fine-scale measurements derived from LiDAR data with high-frequency GPS locations, a more detailed understanding of habitat selection can used to inform conservation strategies by preserving existing patches and reforesting habitat.

Range-Wide Declines of Northern Spotted Owl Populations: on the Road to Extinction?
Alan B. Franklin, Katie Dugger, Damon Lesmeister, Raymond Davis, J. David Wiens, Gary C. White, James Nichols, James Hines, Charles Yackulic, Carl Schwarz, Larissa Bailey, Kenneth Burnham

The northern spotted owl (Strix occidentalis caurina) inhabits older coniferous forests in the Pacific Northwest and has been at the center of forest management issues in this region. The immediate threats to this federally listed species include habitat loss and competition with barred owls (Strix varia), which invaded from eastern North America. We conducted a prospective meta-analysis to assess population trends and factors affecting those trends in northern spotted owls using 26 years of survey and capture-recapture data from 11 study areas across the owls’ geographic range to analyze demographic traits, rates of population change, and occupancy parameters for spotted owl territories. We found that northern spotted owl populations experienced significant declines of 6–9% annually on 6 study areas and 2-5% annually on 5 other study areas. Annual declines translated to <35% of the populations remaining on 7 study areas since 1995. Barred owl presence on spotted owl territories was the primary factor negatively affecting apparent survival, recruitment, and ultimately, rates of population change. Analysis of spotted and barred owl detections in an occupancy framework corroborated the capture-recapture analyses with barred owl presence increasing territorial extinction and decreasing territorial colonization of spotted owls. While landscape habitat components reduced the effect of barred owls on these rates of decline, they did not reverse the negative trend. Our analyses indicated that northern spotted owl populations potentially face extirpation if the negative effects of barred owls are not ameliorated while maintaining northern spotted owl habitat across their range.

Habitat Use of Native Mesocarnivores: Guigna, Culpeo and Chilla Foxes after the Fire Storm in Southern-Central Chile
Christian Osorio, Marcella Kelly, Carlos Tomás Urrutia Valdés

Landscape changes in Chile are threatening wildlife both directly through habitat destruction and fragmentation, and indirectly through increasing the likelihood of catastrophic fires. In summer of 2017, mega-wildfires burned 183,946 hectares in central and southern-central Chile. To assess the impacts of the mega-fire, 4 years post burn, we used occupancy modeling of camera trapping data to assess factors influencing mesocarnivores distribution across the landscape. Mesocarnivores play important roles in structuring wildlife communities, and in this context, there has been little research conducted on mesocarnivores of central Chile, and the effects of fire have not yet been assessed. We used a suite of covariates that included effects of fire, native vegetation, exotic plantations, and other anthropic, climate, and landscape variables to assess both detection and occupancy for Culpeo (Lycalopex culpaeus) and chilla (Lycalopex griseus) foxes, and the vulnerable small felid the guigna. We found variable responses among species with chilla foxes exhibiting the most tolerance to human modified landscapes with high detection rates near rivers and human habitation and showing little effect of fire on habitat use. Culpeo foxes were intermediate in being highly detectable in homogeneous plantations but exhibited tendency toward lower occupancy in burned areas. Guignas were highly detectable near secondary roads and exhibited tendencies toward lower occupancy as burn severity increased and higher occupancy as native forest increased. We projected estimated site occupancy probabilities for each species across the landscape as a way to prioritize conservation efforts by identifying high occupancy areas for habitat protection or to target areas for restoration that link high occupancy areas. This is not only a way forward for conserving biodiversity, but since native forests are less fire-prone, will aid in proactively preventing future devastating mega-wildfires.

Contributed Oral
Location: Virtual Date: November 5, 2021 Time: 4:00 pm - 5:00 pm