Introduction
Population distribution and gene flow depends heavily on habitat connectivity. Fragmentation of natural areas causes genetic differentiation in urban landscapes. The consequences of fragmentation depends on the species; generalist species (wild boars) adapt better to artificialisation by exploiting resources available around smaller fragments. Wild boars prefer environments with dense vegetation, which provides them both shelter and food. The mobility and use of space of wild boars reflects their decisions balancing the benefits of exploiting a resource and the dangers it takes to reach them. Human disturbance here is the real danger. Human disturbance and habitat fragmentation leads to higher mobility and higher nocturnality. SDM (Species distribution models) are powerful tools to identify the distribution area of a species. These models generate statistical relationships between the probability of presence between species and environmental variables, such as land use, landscape metric, climate or topography. Using survey data this method can predict the ranges of species based on current or future environmental factors. This method was used in this study predict urban boar’s range and assess landscape features selected or avoided by this species. The first hypothesis being that wild boar’s presence in urban areas is determined by access of resources and modulated by avoiding densely built up areas. The second hypothesis is that Bordeaux’s urbanistic green frame is suitable for wild boars.
Methods
This study took place in Bordeaux Metropolis (southwest France). The study area was divided into 100-hectares cells, equaling 524 observation units. A third of these units (174) were randomly selected and explored from 25th March to 10th May 2019 to detect wild boar presence. Observations included sightings, rooting sites, wallows and trails. Other clues like footprints, rubs, hairs, or beds confirmed wild boar frequent the area. Cells where boars were present in were given an observation score of 1 (corresponding to a 100% probability that animals have been there) and the cell assigned 0 corresponded to absence of species. Then the probability of wild boar’s presence in each grid cell was studied using logistic regression models (R 4.03 software and QGIS) incorporating several predictor variables. These environmental variables included 1) physical obstacles to animal movements; 2) degree of human impact; 3) availability and accessibility of resources (food, water and shelter).The Garonne river and Bordeaux’s ring road were also seen as physical obstacles for the wild boar. Therefore, every grid cell was labeled based on its position to the river and its centrality to the road ring. The density of buildings was and roads was calculated in each grid cell and a French topographic database was implemented for metric precision. Hosting potential of a grid by a wild boar was estimated by calculating the surface area of lands that tend to provide resources and by assessing the level of fragmentation of forest habitats included in the cell. Overall by categorizing grid cells in three categories on level of amount refuge and food it was determined that 53.6% of the total study area had either refuge or food if not both. After removing correlated variables, they tested multiple model combinations and selected the best fit using AIC, BIC, and TSS metrics. The strongest model (TSS = 0.79) showed that wild boar presence increased with larger green areas (WBGF) and decreased with higher building density, confirming that access to resources and avoidance of dense urban zones best explained their distribution. Predicted probabilities were mapped to identify areas with high (>50%) or low (≤50%) likelihood of wild boar presence. Results showed that nearly 90% of the urban green frame was suitable for wild boars, especially outside the ring road and on the right bank of the Garonne River. A year-long camera trap survey beneath the ring road recorded 114 wild boars, confirming that animals use this underpass as a corridor between suburban and inner-city green spaces, validating the spatial model’s predictions about connectivity.

Results
Model analyses showed that wild boar presence in Bordeaux was mainly driven by the availability of green habitats and limited by building density. The best logistic model confirmed that wild boars were most likely to occur in areas with extensive wild boar green frame (WBGF) and fewer urban structures, especially outside the ring road and on the right bank of the Garonne River. Overall, 68.9% of grid cells had a predicted boar presence probability above 0.5, closely matching field data where presence was observed in 67.8% of surveyed cells. Nearly 90% of the green frame showed a high probability of use. Inside the ring road, boars were mostly present on the right bank (70.5% of green area likely occupied) but rare on the left bank (22%). Camera trap data supported these predictions, showing frequent use of an underpass beneath the ring road, with boars moving both into and out of the city. This confirmed that while boars prefer peripheral green zones, they can cross barriers and access central urban areas through existing corridors.
Critique
The study was successful in using statistical modeling intertwined with field observations to understand urban wildlife niches. The author’s use of statistical models like SDMs and QGIS and R provided a thorough, clear, and replicable method for assessing environmental variables. Overall the variety in methods like field observations, camera monitoring, and spatial analysis makes the research highly reliable. However, there were some limitations to the study. For one, the field survey took place over a short period of time during a single season (spring). This small temporal scope may not capture the variation in boar behavior across seasons. Additionally, using whole 100-hectare grid cells may miss fine-scale behavioral patterns and small habitat use within fragmented areas. Furthermore, the cells with no seen boar behaviour within were marked as “absent”. This is tricky, because assuming absence of evidence is one thing , but this does not give us the right to assume actual absence of the presence of boars in that grid. Lastly, it would’ve been beneficial to include social attitudes to implement in the coexistence between boars and humans. All in all, the paper provides valuable knowledge about how generalist species exploit urban environments and how we must acknowledge the coexistence of these species and us.

Introduction

The global lockdown beginning in 2020 allowed some wildlife to venture into cities, like foxes, coyotes, and birds. The paper “What can we learn from wildlife sightings during the COVID-19 global shutdown?” by Zeller et al. (2020) in Ecosphere encompasses both an ecological and social study. The author uses citizen science data to explore changes in wildlife observations during the pandemic. This paper caught my attention due to the discussion in class about the pause in human activity during the pandemic and how urban wildlife behavior is intertwined with our presence.

Summary

Zellmer and colleagues analyzed citizen reported wildlife sightings from platforms like iNaturalist and eBird, comparing data from the early months of the pandemic to previous years. The purpose was to figure out whether wildlife actually became more abundant in cities during lockdown or if people only noticed a greater abundance of animals because they were at home and actually paying attention.

Results

They found mixed evidence. Reports increased of wildlife sightings in many urban areas, but this does not mean that there were necessarily more animals. Species that appeared more frequently in urban environments were coyotes, deer, and foxes. These were species that typically avoid dense human spaces. The paper emphasized that changes in behavior varied by species, city, and the degree of lockdown strictness. They ultimately, highlighted, that human movement can drastically alter the visibility and distribution of urban wildlife. A big portion of the paper is dedicated to acknowledging what the data cannot prove due to observer effort bias, detection vs. abundance, and heterogeneity among cities. The article never comes to an exact conclusion that wildlife increased in urban areas during lockdowns, but rather raises possibilities and promotes cautious interpretation.

Critical Reflection

This study is strong in that it used this rare global event as a large-scale natural experiment. The use of citizen science allowed for fast data collection and broad spatial coverage during a difficult time. The article is also rightfully cautious in not claiming “nature reclaimed cities” and instead acknowledging biases in observation and sampling. This underlines that this study is more of an analysis of human behavior rather than animal abundance. All in all this reflects on their credibility, making it a more reliable source. However, the authors could have gone further in correcting biases, rather than simply acknowledging it. For example they could have implemented observation models and categorization, which could have helped them separate real life wildlife increases from high detection rates. Furthermore, the study focuses on the early stages of lockdown in 2020, when in reality animals responses to the lockdown evolved over time. A study comparing all years of lockdown would test whether behavioral flexibility persists. Further studies could combine citizen science with automates monitoring, such as camera traps, sensors or tracking collars. Additionally, comparing results across continents could reveal cultural or infrastructural influences on how wildlife adapted depending on the density of buildings, size of city, noise pollution etc.

Work Cited

Zellmer, A. J., Wood, E. M., Surasinghe, T., Putman, B. J., Pauly, G. B., Magle, S. B., Lewis, J. S., Kay, C. A. M., & Fidino, M. (2020). What can we learn from wildlife sightings during the COVID-19 global shutdown? Ecosphere, 11(8), e03215. https://doi.org/10.1002/ecs2.3215

Background and purpose: As urbanization grows, the overlap between cougar and human populations increases. Large carnivores are highly susceptible to habitat modification, because they tend to have low population density and wide ranging travel for their food requirements. The effects of urban development on the way cougars forage has rarely been studied and researched. This article investigated variation in cougar use of three prey types (synanthropes, ungulates, and rodents) along a wildland–urban gradient in western Washington to determine how urbanization affects the foraging ecology of this apex predator.

Methods: This study used trained dogs and cage traps to capture and radio-tag cougars throughout the 4450km^2 study site. They did this from 2004-2008 and again from 2013-2016. Once captured the animals were immobilized and given a physical examination and outfitted a GPS radio collar. Kill sites were located and prey identified. The surrounding urbanization of kill sites was measured as building density (structure per hectare). The diets of twenty individual cougars and their 568 kills were analyzed using statistical models.

Results: Firstly synanthropic prey use increased and odds of cougars preying on synanthropes (animals living in close association with humans) rose nearly fivefold with each additional building per hectare. However, only certain individual cougars specialized in synanthropes. Black-tailed deer and elk did remain the dominant prey throughout. Cougars remained to rely on ungulates as prey, which suggests that predator-ungulate systems can survive near human settlements. Additionally, as building density increased, kills of beaver and mountain beaver decreased, likely due to habitat loss and management practices reducing rodent presence. And it was also discovered that male cougars hunted rodents more often than females. Lastly, the study showed that some cougars deviated extensively in prey choice. For example, one male accounted for half the rodent predation.

Criticisms: Overall I thought this study was an insightful and interesting read. The research did have some limitations however. For one, building density was the only measure of urbanization, which oversimplifies the the complexity of human disturbance. Furthermore, the difference in collar technology between study periods may also have affected kill detection, making temporal comparisons less reliable. The lack of direct data on prey abundance makes it difficult to analyze whether dietary preferences of the cougar was a result of prey preference or prey abundance. Lastly, the limited sample size of twenty cougars makes it hard to apply this data across populations. Individual variation most likely strongly influenced results, especially if one male accounted for half the rodent kills. Future work could benefit from combining prey availability survey with kill site data to distinguish between preference and availability. Additionally standardized tracking technology, multiple measurements of urbanization, and a broadened sample size would yield more thorough results.

Reference:

Robins, C. W., Kertson, B. N., Faulkner, J. R., & Wirsing, A. J. (2019). Effects of urbanization on cougar foraging ecology along the wildland–urban gradient of western Washington. Ecosphere, 10(3), e02605.https://esajournals.onlinelibrary.wiley.com/doi/10.1002/ecs2.2605