By Brittany Goins, as part of her completion of her Bachelor’s of Arts in Physical Geography and Geographic Information Systems.

Abstract

Mountain lions (Puma concolor) play an important ecological role as apex predators, yet their presence increasingly overlaps with human activities across California, particularly in the wildland-urban interface (WUI) zones (Radeloff et al., 2005). The purpose of this research is to predict the spatial distribution of mountain lion presence in El Dorado County by identifying key environmental and landscape variables associated with their occurrence. A Random Forest (RF) classification model was used with presence data from confirmed mountain lion sightings (H. Molzahn & M. Summers, 2024) and pseudo-absence points generated across the county. The spatial datasets were sourced online and processed using ArcGIS Pro, with environmental attributes extracted for each pseudo-absence and known location point. The final Empirical Bayesian Kriging (EBK) model output emphasizes areas of high, moderate, and low predicted sighting probabilities, highlighting the influence of man-made and natural landscape features on the prediction of mountain lion presence within the county. 

Introduction

The mountain lion (Puma concolor), commonly referred to as the cougar or puma, is one of the most wide-ranging and adaptive mammals in the Western Hemisphere. Worldwide, thirty subspecies are recognized, with thirteen being found in North America (Tesky, 1995). Mountain lions are the fourth-largest cat in the world and are comparable in mass to the African leopard and in general, their body size increases with latitude (Logan & Sweanor, 2001). These animals were once commonly distributed across most of the Americas before the settlement of Europeans, but their populations have overall declined throughout the last 200 years due to an increase in hunting and urbanization (Tesky, 1995). Currently, the mountain lion dominates a broad but fragmented range across North America, with populations persisting from the Yukon territory of Alaska to the Andean Mountain Range in South America (LaBarge et al., 2022). 

These large carnivores are generally found throughout California in various elevations extending from sea level to 10,000 feet, and their preferred habitat largely depends on the habitat of their prey (United States Department of Agriculture, 2020). Furthermore, within these habitats, they tend to favor rocky cliffs or ridgetops with plenty of vegetation to provide cover for observing their prey. Mountain lions are more likely to use what’s referred to as edge habitats, which are the transition areas between two different vegetation types or habitats, like where forested areas meet open grasslands or human development (Holmes & Laundré, 2006). This interface, where two different habitats meet, is commonly used by deer and other prey animals, giving the mountain lions enough cover to watch and stalk from the underbrush. They are also known to utilize creek beds and riparian stream channels as travel corridors since these areas have plenty of dense vegetation cover. Throughout California, over half of the land is considered mountain lion habitat including vegetation communities like high elevation alpine, coniferous hardwood forests, open woodland and mixed chaparral foothills, and shrub grasslands (California Department of Fish and Wildlife, 2017). A study conducted in 2002 found that the average home range for mountain lions in the Sierra Nevada for females and males in the summer was 209 mi² and 279 mi², while in winter it averaged 135 mi² and 181 mi² respectively (Grigione et al., 2002). On average, males typically weigh up to 60% more than females and have a much larger home range due to territorial and reproductive needs (Logan & Sweanor 2001).  

Mountain lions play a critical ecological role as large carnivores, not only being a keystone species but also as apex predators that contribute to the structure and function of ecosystems. They do this by influencing and regulating prey populations and altering the flow of energy in biological communities by leaving behind carcasses for other animals to scavenge (LaBarge et al., 2022). They are active year-round and are solitary, elusive, and primarily understood to be nocturnal, crepuscular predators that exhibit remarkable flexibility with adaptability, diet, and movement patterns (McCain, 2008). Mountain lions are opportunistic generalists, meaning they are not picky with what they eat and if the chance arises for them to capture easy prey, they will take it. Contrary to popular belief, mountain lions hunt during diurnal hours when the sun is out, which is not what most published studies have shown (Allen et al., 2015). They will follow the daily activities of the prey they are hunting according to the availability of that prey species in a season, along with shifting their hunting hours based on temperature changes (McCain, 2008). A defining feature of mountain lion predation is the behavior of caching, involving the carcass of their prey being dragged to another area to be hidden under soil, leaf litter, or snow, to prevent other scavengers from taking their kill (Murphy et al., 1998). After a kill has been made, they then move the body of the deceased prey to another location, typically to an area where they can feed uninterrupted with enough concealment to keep them from being detected by other animals or birds.

 A primary source of prey for mountain lions in California is mule deer and big horn sheep, making up 60-80% of their diet, although they eat other small mammals and occasionally domestic livestock like goats and sheep when the chance arises (Ahlborn, 2000). The depredation of livestock has brought to light the issue of the Human-Carnivore Conflict (HCC), especially in California. HCC is technically any type of conflict between humans and carnivore populations, being that of wolf, bear, or any large predator. This type of conflict involving mountain lions can take several forms as attacks on human life, depredation of livestock or pets, vehicular fatality of lions, or any other type of strain between humans and mountain lions. HCC occurs for several reasons since these large animals have a protein-rich diet and have large home ranges that bring them into competition for natural resources with humans (Treves & Karanth, 2003). These conflicts are likely going to increase with the expansion of human development into rural and forested areas as both mountain lion and human populations continue to grow. 

As human development continues to expand, the spatial dynamics of mountain lion populations are going to be increasingly shaped by anthropogenic influences. As this expansion into rural and undeveloped areas increases, there has been a focus on the environmental and ecological conflicts that arise in what’s called the wildland-urban interface (WUI). WUI is defined as a transition zone between undeveloped natural habitats and human-developed areas, such as residential communities, agricultural lands, or other urban infrastructure (Radeloff et al., 2005). The areas that meet this definition are where there has been the most conflict between mountain lions and people, especially when those human populations are near national forests and designated wilderness areas. Additionally, deer populations are abundant in California’s foothills and mountains, and they adapt well to WUI environments, drawing mountain lions even closer to residential areas. This push of new development into rural areas brings people even closer into mountain lion habitat, leading to more visibility and conflict with humans.

In the western United States, some populations of mountain lions are considered stable, however, there are cases of isolated populations that are facing threats from low genetic diversity, as seen in studies conducted in southern California (Drummond, 2021). There has been speculation over the years that mountain lion populations have been steadily increasing in California due to the 1990 ban on hunting them and declassifying them as a game species (Lupo & Brinkhaus, 1996). However, a common threat to all mountain lion populations is habitat fragmentation and the ever-increasing pressure from expanding human development. Despite their adaptability to diverse environments, mountain lion populations have experienced significant pressures from this habitat loss and the conflict with humans (Vickers et al., 2015). 

In California, over the past 100 years, the management of mountain lion populations has had a long and varied history going from a bounty system from 1906 to 1963 to now holding a protected status (Dellinger & Torres, 2020). They are currently considered a “specially protected” species (Fish and Game Code Section 4800), which indicates that they cannot be possessed, injured, hunted, or transferred unless certain circumstances arise like depredation or loss of human life. This stems from the passage of the “The California Wildlife Protection Act of 1990” (Proposition 117) that permanently banned the hunting of mountain lions in the state of California (California Department of Fish and Wildlife, 2020). In the spring of 2013, the California Department of Fish and Wildlife (CDFW) had to update its policy from 2007 that dealt with the issue of depredation and public safety due to the rising concerns and increasing reports of human-mountain lion interactions. 

Given the complex ecological factors and the growing interactions between mountain lions and expanding human populations, there is an increasing need to gain deeper insight into where mountain lions are most likely to be present in the landscape, particularly in areas like El Dorado County, where human development continues to encroach upon natural habitats leading to an increase in sightings and depredation reports. Between 2020 and 2022, El Dorado County accounted for 15% and 23%, respectively, of all nonlethal and lethal mountain lion depredation permits issued statewide by the California Department of Fish and Wildlife (CDFW, 2017). For the actual numbers of annual depredation reports published by the CDFW, El Dorado County was the highest for annual depredation permit totals in all of California, going back to the early 2000s. Between 2001 and 2010, a total of 115 permits were issued. This number increased to 187 during the subsequent decade (2011–2020). In the most recent reporting period, covering only two years (2020–2022), 101 permits were officially issued. This shows that, according to CDFW’s reports, El Dorado County by far has the most depredation permits that were issued out of all the counties in the state. The other two highest counties were Mendocino and Siskiyou County, which also had high numbers, but did not come close to the amount of El Dorado (California Department of Fish and Wildlife, 2022). 

This study aims to use a Random Forest (RF) machine learning approach to model to predict mountain lion sightings based on a range of environmental variables, including important factors like land cover, elevation, vegetation type, and canopy closure among many other variables. Then use the interpolation method of Empirical Bayesian Kriging (EBK) to create a prediction surface raster by identifying areas with higher probabilities of mountain lion presence. By recognizing the key landscape features associated with the mountain lion’s presence, this research can provide constructive insights for wildlife managers and the public to better anticipate and mitigate potential human-carnivore conflicts. Beyond an academic interest, this work is critical for promoting coexistence and mitigation strategies for the public’s safety in El Dorado County. With the increasing development into the WUI, understanding these spatial patterns is essential for mountain lion conservation while also ensuring community safety.

Study Area

This study focuses on El Dorado County, the 27^th largest county in California, encompassing approximately 1,707.8 square miles of land. As of 2020, the total population was 191,185, and it is expected to continue to increase (United States Census Bureau, 2023). The county has a rich mining and timber history, rooted in the California Gold Rush in 1848, leading to an influx of people pouring into the foothills of the Sierra Nevada. The county was named El Dorado, meaning  “the gilded one” in Spanish, aptly named after the discovery of gold at Sutter’s Mill along the American River in Coloma (Broberg, 2019). The current county seat is the city of Placerville, which has retained this title since 1850 (El Dorado County Government, 2025). This county has a strong landowner rights culture where many residents place high value on rural living and ranching. Large areas of the county are composed of privately owned parcels, state parks, and conservation areas. The county is also bisected by major highways, including Highway 50 and State Route 49, and has a mix of both rural and urbanized areas that intersect with wildlife habitat. Most of the eastern portion is managed by the U.S. Forest Service and contains the Eldorado National Forest, which is rural with rugged terrain (Thrower, 2005). The county’s boundary stretches from the western foothills of the Sierra Nevada Mountains to the eastern crest near the Nevada border and includes portions of the Lake Tahoe Basin and the American River watershed. 

The underlying geology of the area is influenced by the Sierra Nevada batholith, which dominates the bedrock and is mainly composed of granitic and metamorphic rocks, which gives shape to the ridgelines and deep valleys (Bateman et al., 1963). There are distinct biotic zones observed within El Dorado County, and the climate varies across this gradient due to the change in elevation. These elevation differences are the dominant control of temperature and precipitation, especially in the mountainous regions of this county (Pepin et al., 2022). The pattern of weather here is influenced by the Sierra Nevada’s topography and position relative to the Pacific Ocean, where the moist air masses are pushed by the Westerlies east over the crest of the mountain range. El Dorado County primarily sits on the western side of this range, so it receives more precipitation, leading to diversity in the biotic communities. 

The general gradient of climate begins in the woodland foothills, along the western edge and boundary of the Central Valley, where grasslands and agriculture dominate around the southwestern edge. The typical Mediterranean climate dominates in the lower elevations, characterized by hot, dry summers and mild, wet winters (Cowling et al., 1996). Most of the precipitation arrives during winter, with the summer season primarily experiencing drought for 3 to 6 months, having a prominent influence on the plant communities found there (Safford et al., 2021). Along this boundary are the chaparral and oak woodlands, also known as the foothills. This biotic zone ranges in elevation from 400 to 1,500 feet, although occasional stands can be found as high as 3,500 feet (Van Wagtendonk et al., 2018). The lower and upper montane biotic zones make up the largest percentage of the county’s area, containing productive and temperate forests with mixed conifers and dense vegetation. In the lower zone, elevation ranges from 2,500 to 6,000 feet, and primary vegetation types include California black oak (Quercus kelloggii), ponderosa pine (Pinus ponderosa), firs (Abies concolor and Pseudotsuga menziesii ), and mixed evergreen forests (Van Wagtendonk et al., 2018). The upper zone’s elevation ranges from 2,500  to 11,500 feet, however, the highest elevations found within El Dorado County are around 10,000 feet. Within this zone, most winter precipitation falls as snow in these forests with the primary tree canopy containing jeffrey and lodgepole pine (Pinus jeffreyi and Pinus contorta). 

The county contains many different ecosystems, including oak woodlands, chaparral, coniferous forests, and alpine environments. These biotic zones and the distinct landscapes that are created from them provide a complex mixture of potential habitat types for mountain lions, which are known to typically inhabit large, undisturbed natural areas with access to prey. However, mountain lions have been coming into urban areas more and more, with county residents becoming more alarmed as the frequency of these visits increases. There are online community groups dedicated specifically to documenting the mountain lion sightings and depredations within El Dorado County. On average, there are at least three to four posts daily showing video and camera evidence that not only are mountain lions trying to enter residences and being found in major urban areas, but that the majority of these posts show that this is occurring during daylight hours. This is highly concerning for people with pets and small children, and more research will need to be conducted in El Dorado County specifically so that the public will stay protected and learn to manage with other coexistence strategies in the meantime.

Figure 1. Context map of El Dorado County, CA. Showing the extent within California, the Eldorado National Forest, main towns and cities, and the major highways that connect.

Methodology

The goal of this research is to identify the particular landscape characteristics that most significantly contribute to mountain lion presence, which then enables the spatial prediction of high-probability sighting areas throughout the county through machine learning, ultimately allowing the dissemination of this information to the residents of El Dorado County. This study uses a Random Forest (RF) classification approach to predict the likelihood of mountain lion sightings in El Dorado County, California, based on environmental and landscape features related to confirmed mountain lion sightings. RF is an algorithm that uses classification systems or regression trees and is widely used for modeling species distributions based on a set of input variables (Valavi, Lahoz‐Monfort, & Guillera‐Arroita, 2021). How this works is each tree in the “forest” is trained on a random subset of the data, termed “bootstrapping”, and at each node, the algorithm selects a random subset of predictor variables to determine the optimal split in the tree (Valavi, Lahoz‐Monfort, & Guillera‐Arroita, 2021). These decision trees are constructed by dividing the response data that are based on predictor variables. At each step, the algorithm evaluates possible split points on a selected predictor to divide the data into two groups, with the goal of maximizing similarity within each group, termed a node, and also maximizing the difference between them. All of the available predictors are tested at each split, and the one that produces the greatest increase in node separation is chosen. Every time a layer is split, it adds depth to the tree structure, advancing it further (Valavi, Lahoz‐Monfort, & Guillera‐Arroita, 2021). This algorithm handles both categorical and continuous data types, allowing for a smooth operation and analysis using ArcGIS Pro.  

Figure 2. Showing the visual representation of how the Random Forest algorithm works by constructing multiple decision trees and aggregating their results (Blakely et al., 2018 ).

Data Collection

This study employed a collaboration with a team of independent researchers and biologists located in El Dorado County who have been actively researching the mountain lion sightings in this county since 2024. The research team consisted of Haley Molzahn and Marie Summers, who created and collected the raw, unpublished mountain lion sighting data (H. Molzahn & M. Summers, 2024) as part of the El Dorado County Carnivore Project (https://carnivoreproject.org/).

The environmental predictor variables were selected based on the ecological relevance to mountain lion habitat preferences and the landscape factors found at those sites. The datasets came from primary and secondary sources. Land cover data was incorporated as a key predictor variable to represent the physical and biological characteristics of the landscape across El Dorado County. The dataset was compiled using layers from the California Department of Forestry and Fire Protection (CAL FIRE, 2024) and the LANDFIRE program’s Existing Vegetation Type (EVT) dataset (LANDFIRE, 2023), accessed via the LANDFIRE Map Viewer. These sources classify land cover based on vegetation structure and dominant plant species, including evergreen forest, shrubland, grassland, developed areas, and agriculture. Canopy cover was originally sourced from LANDFIRE program’s Existing Vegetation Cover (EVC) dataset (LANDFIRE, 2023), accessed via the LANDFIRE Map Viewer and used as an indicator of habitat structure and potential cover for stalking prey (open vs. closed canopy).

Datasets on protected lands and the Bureau of Land Management (BLM) areas were incorporated into the analysis. These datasets were obtained from the USGS Protected Areas Database of the United States (USGS, 2024). The datasets include national forests, state parks, designated wilderness areas, along with lands managed by BLM. These variables were included to explore whether mountain lion sightings were more likely to occur on lands with less human development, such as conservation areas, which offer large, untouched lands and contiguous habitat for wildlife. The distance to waterbodies and rivers was included as a hydrological predictor variable and was sourced from the U.S. Census Bureau (U.S. Census Bureau, 2023). The availability of water is necessary for wildlife and mountain lions are known to utilize these areas near rivers and streams for travel and hunting corridors, especially during the dry seasons in El Dorado County. This data was used to represent proximity to water sources and riparian corridors. A 0.5-mile buffer was employed to generate the distance using ArcGIS Pro.

The distribution of mule deer (Odocoileus hemionus) range, being one of the primary prey species for mountain lions, was used as a strong environmental indicator of potential mountain lion presence. By incorporating this dataset, the model was used to capture the spatial dynamics at a landscape-level pattern. The mule deer range was sourced from California Open Data Portal and created by CDFW (California Department of Fish and Wildlife, 2025). Elevation, aspect, and slope data were incorporated to capture the gradient and orientation across the county, which strongly influences climate, vegetation patterns, and mountain lion habitat structure. Higher elevations are usually known to support coniferous forests and seasonal deer migrations, while lower elevations feature mixed woodland and grassland habitats. This data was derived from LANDFIRE’s Map Viewer (LANDFIRE, 2023) and processed as a continuous variable.

Agricultural land use is taken into account as a potential factor influencing mountain lion presence due to the availability of livestock and open land. The data was obtained from the California Department of Conservation’s California Important Farmland Finder (CIFF) tool (California Department of Conservation, 2025). The dataset assigns a classification system to different categories of farmland based on soil health, type of farming land use, and local agriculture.

Zoning classification and land use data were sourced from El Dorado County’s government website and ArcGIS Web Application (El Dorado County 2024, GOTNET) and added to allow for the overlay of urban versus residential land-use patterns in relation to the reported mountain lion sighting data points. To account for anthropogenic influence on mountain lion distribution, the analysis included the U.S. Census Bureau’s data of all roads in El Dorado County as a landscape variable in the RF model (U.S. Census Bureau, 2022) since roads represent disturbances and play a role in mountain lion movement patterns. 

Data Preparation

The spatial data layers and datasets for each of the landscape variables were collected and then processed using ArcGIS Pro. Continuous variables, such as distance to water, protected areas and major roadways, elevation, aspect, and slope, were rasterized at a consistent spatial resolution of 30 meters. Categorical variables, such as land cover, vegetation type and cover, and zoning, were reclassified using the Raster Calculator tool for model input. Sightings data were overlaid with the environmental layers to extract values of each predictor at sighting locations using the Extract Multi Features to Points tool. In order to provide a balanced dataset, random background points (pseudo-absence data) were generated across the study area, ensuring representation of the full range of environmental conditions. A set of pseudo-absence points was generated using ArcGIS Pro’s Create Random Points tool to represent absence data. These points were randomly distributed within the boundary of El Dorado County, excluding known sighting locations, with an allowed minimum distance of 4 meters so that there was no overlap between the known locations and generate absence locations. A total of 792 known mountain lion sighting points and 810 background-absence points were used to train and test the RF model.

Results

Using the RF model, the dataset was divided into 70% training and 30% testing subsets to calculate the model’s performance. The predictor variables were then appropriately formatted to allow for the use of both categorical and continuous data. This model was chosen for its robustness in handling ecological datasets and its ability to assess the significance of various input variables (Valavi, Elith, Lahoz‐Monfort, & Guillera‐Arroita, 2021). The model is trained on the dataset to classify mountain lion sighting presence and absence points based on the environmental predictors mentioned in the data collection section. The model was run twice to compare results and validate the findings of the trained features.

 Each absence point was assigned a value of 0 in the response variable field of “Presence” and later merged with the known mountain lion presence dataset to create a binary response layer suitable for classification modeling. The known mountain lion sighting data was assigned a 1 to indicate confirmed presence. This binary coding scheme was necessary for supervising machine learning in the RF model, where the model is trained to distinguish between presence (1) and absence (0) based on associated environmental predictors from the other datasets. 

Variable of importance rankings (Figure 4) were generated to interpret the relative contribution of each environmental factor to the mountain lion presence on the prediction layer. The trained model was then applied across the entire study area of El Dorado County to generate a predictive map of mountain lion sighting probability. All environmental predictors were extracted from where the confirmed mountain lion sighting locations occurred and from randomly generated background points, which represented the available landscape conditions.  To visualize the spatial distribution of predicted sighting probability, the trained model was applied across El Dorado County by converting all layers into a raster format for all the input variables. The resulting probability prediction surface is shown in Figure 3, where the blue points represent a likely probability of presence, and the orange points represent where there’s a predicted absence.

     Figure 3. Modeled Likelihood of Mountain Lion Presence in El Dorado County.

In order to get a true prediction map, I then used the RF model’s prediction to interpolate areas within the county that had a higher likelihood of presence. This second analysis was done using Empirical Bayesian Kriging (EBK) to create a surface showing the relative probability of mountain lion presence. EBK is a geostatistical interpolation method that predicts values at unsampled locations based on the spatial autocorrelation, which means that observations closer to each other are more similar than those farther away from the sample points. This way of modeling has a main assumption that mountain lion sightings at one location influence or are similar to nearby locations, which is correct in a way for this study. Some of the sighting reports are from the same lion at different locations within the same neighborhood so the assumption fits. The output is a pseudo-probability surface that’s essentially modeling spatial trends in presence intensity according to the RF model. It shows a continuous surface predicting the likelihood of presence even in areas without direct observations. The higher values (orange/red) can be interpreted as hotspots or zones of high predicted mountain lion activity. Lower values (blue/teal) reflect lower predicted probability or likely absence zones.

Figure 4. Interpolated surface showing relative probability of mountain lion presence using Empirical Bayesian Kriging on Random Forest model-predicted presence data.

Discussion

The analysis showed clustering from both the RF and EBK models in the central-western region of El Dorado County of predicted mountain lion presence. This makes sense because this area is where the mountain lion data was reportedly taken and shows that the RF model was accurate in the way it was trained using the landscape and environmental variables. The reason for the clustering in this part of the county is the way the raw mountain lion data was collected, since it was obtained from county residents and was self-reported. The residents who made the initial reports live in areas near roads and in developed areas within the county, skewing the data to the more populated areas, especially on the western portion of the El Dorado County. This is why the two most significant variables are distance to developed areas and the land cover category roads, as shown in Figure 5. The majority of reports centered around the western half of the county because the eastern region is a national forest and sparsely populated, where no reports have been made yet. This analysis would have been better if the data for the mountain lion sightings had been collected over many years instead of just for 2024. With more research into mountain lion sightings and data collection, eventually we could create a finer picture of where the presence occurs across the whole county.

    Figure 5. The variable importance table is given once the random forest model creates the prediction layer.

There is a senate bill proposed for El Dorado County named SB 818 that would create a 5-year pilot program named “Tree and Free” allowing for qualified houndsmen to use non-lethal methods to proactively haze mountain lions that are considered a threat to livestock or human life (Alvarado-Gil, 2025). The aim of this bill is to establish and restore the mountain lions’ fear of humans again, causing them to become more cautious and to reduce depredation events while also keeping them safe from conflicts with humans by inducing aversive conditioning. The objective of the “tree and free” approach is to condition mountain lions to avoid human-populated areas and encourage their return to more remote, forested habitats while providing data collection on the efficacy of nonlethal hazing to guide future management strategies and aid in research across California. These mountain lion pursuits would occur only in areas considered appropriate by the California Department of Fish and Wildlife (CDFW) in coordination with the local law enforcement of El Dorado County (Alvarado-Gil, 2025). 

The reason this bill is important is that it was created after a tragic incident occurred in March 2024 in El Dorado County, where two brothers, Taylen and Wyatt Brooks were attacked, leaving Taylen fatally injured and Wyatt sustaining severe injuries but thankfully surviving the attack. This attack was carried out by a male mountain lion during the day, and was completely unprovoked, leaving lingering questions in the residents’ minds and bringing into focus the likelihood of another encounter. The young men knew the area well and knew what to do in a dangerous wildlife encounter, and still this tragedy happened. Sadly, this bill failed to pass through the Senate Committee with the original language proposed on Tuesday, April 22^nd, 2025. Many of the legislators who represent urban districts have never had to face living alongside these predators in a rural county, and ultimately altered the core amendments on the bill (Grimes, 2025). These alterations scrubbed the main point of the bill to promote public safety with the help of houndsmen keeping these predators away from residents’ homes and properties. The original language of the bill was replaced with what CDFW is already doing with the “three-step” policy, ensuring no immediate response or change in these laws and regulations. 

One of the main voices against this bill was the Mountain Lion Foundation, which claimed that the SB 818 bill was needless and would allow for the “ruthless harassment” of mountain lions (Bennett, 2025).  It is interesting that groups like the Mountain Lion Foundation, which state that their organization’s goal is to protect mountain lions and their habitat by advocating for coexistence strategies, would be opposed to something that has been shown to work and possibly prevent the loss of mountain lions on the landscape. In a study conducted in Washington state, researchers explored using GPS collars to test the efficacy of aversive conditioning on mountain lions with the use of hounds and found that after being hazed, mountain lions began to flee after seeing the researchers and were found to have run farther away (Parsons, George, & Young, 2024).

Conclusion

Mountain lions, as apex predators, appear to be growing increasingly comfortable in El Dorado County. Their presence may be influenced by both ecological factors, like the growing number of small-scale farming operations located in the area and certain policies, including the state’s “three-step” depredation protocol. The first step or strike is a confirmed mountain lion attack on livestock or pets by CDFW, then the agency will issue a non-lethal depredation permit (California Department of Fish and Wildlife, 2017). Then the landowner is told to apply deterrents such as protected enclosures, livestock guardian animals, and motion-activated lights. The second strike is when the landowner has applied all recommended non-lethal measures, and a second attack still occurs, then CDFW will issue another non-lethal permit. Only after a third incident takes place can the landowner obtain a lethal depredation permit to have the animal removed from the landscape and disposed of. This policy was established in 2017, requiring landowners to attempt non-lethal deterrents before a lethal depredation permit is issued by CDFW (California Department of Fish and Wildlife, 2017). These permits are issued only for 10 days and encompass a 10-mile radius around where the incident took place (Alvarado-Gil, 2025). During a presentation to the El Dorado County Board of Supervisors on October 8, 2024, the Agricultural Commissioner reported a noticeable increase in the number of livestock and domestic animals lost to mountain lion depredations, with 100 confirmed losses in 2023 and an even greater number anticipated for 2024 (Alvarado-Gil, 2025).

 This represents a substantial rise from the average of 30 to 40 animals lost annually between 2010 and 2022, showing something is drawing these predators to this particular county. The data for this study was collected during the year 2024, and the total number of encounters had indeed increased in frequency when compared to the 2010 and 2020 CDFW reports (California Department of Fish and Wildlife, 2022). In Figure 5, the majority of mountain lion encounters reported in El Dorado County were visual sightings (physically observed or recorded videos), which accounted for the majority of encounter types, representing 585 of all incidents. This was followed by signs or sounds of life with 92 reports, depredation events involving livestock or pets with 87 reports, 17 livestock or pet depredation attempts, and a smaller number of 10 mountain lion fatality reports (by car, professional trapper, or natural causes). Only one report involved a loss of human life.

Figure 6. Mountain lion sightings were derived from unpublished, raw data in CSV format, sourced from researchers Haley Molzahn and Marie Summers as part of the El Dorado County Carnivore Project (H. Molzahn & M. Summers, 2024).

In El Dorado County, the feeding habitats of these mountain lions are coming into direct conflict with livestock grazing and small-scale domestic farming. This risk is especially pronounced in transitional zones like the WUI, which comprises almost the entire county, where these rural properties border forests and undeveloped lands. As agricultural activity continues to grow in this county, this spatial overlap between mountain lion habitat and farming highlights the need for deterrent strategies to reduce human-wildlife conflict while preserving the integrity of animal life. These conflicts have resulted in great economic losses for ranchers and have escalated public safety concerns, not only prompting the issuance of depredation permits but also by prompting residents to take retaliatory action against these carnivores. Residents have begun to lose their patience waiting for something to change, and instead are taking measures into their own hands by implementing a method of: “shoot, shovel, and shut up” (SSS). Now, there are mountain lions being killed and buried without any oversight from CDFW, and the people of this county are losing their faith in the state’s regulations.

With the increase in frequency of mountain lion encounters in this county, more reports will need to be documented so that a full dataset can be compiled in the future. Future research in El Dorado County is necessary, where funding is provided for GPS-collared tracking to study the patterns of mountain lions’ movement. This will help determine how many lions live within the county and where their presence is most likely to occur. If this tracking data can be gathered, we could better predict their spatial patterns and enhance public understanding. The RF model did work and showed that mountain lion presence is clustered around populated areas, however, the data is inherently biased due to the reports coming from residents living in these areas because there is no data for the eastern region of the county since it’s sparsely populated and the reports come from where most people live. The EBK model was not as accurate as the RF model, but it did give a depiction of the hot spots for mountain lion activity. The data that was used for this EBK model was already biased and skewed to the western region, centering on populated and developed areas. This study could have been more accurate if there had been data that was collected over many years instead of one, and if the collection was spread across the whole county instead of one primary region. 

If the SB 818 bill had been passed as originally written, this research could have begun, benefiting conflict mitigation and restoring the natural fear lions should have of humans. The residents of El Dorado County do not want to manage these conflicts themselves or see mountain lions killed illegally and buried in the backyard; instead, they need assistance from agencies that will protect them and their property. Hopefully, in the near future, lawmakers can reach an agreement with county residents to prevent such events from occurring. The residents of El Dorado County have been taking these nonlethal preventive measures for years, and the current system of conflict mitigation is not making an impact in their lives or the lives of the mountain lions. If this bill had been pushed through with the original language, it could have saved the lives of many animals, including mountain lions, and allowed for the much-needed collection of data from the mountain lions for this research. 

References

Ahlborn, G. (2000). California Wildlife Habitat Relationships System – Mountain Lion (Puma concolor). California Department of Fish and Wildlife California Interagency Wildlife Task Group. https://nrm.dfg.ca.gov/FileHandler.ashx?DocumentID=211554

Allen, M. L., Elbroch, L. M., Casady, D. S., & Wittmer, H. U. (2015). Feeding and spatial ecology of mountain lions in the Mendocino National Forest, California. California fish and game, 101(1), 51-65.

Alvarado-Gil. (2025). Mountain Lions: pilot program: permitted houndspersons . Bill Search. https://leginfo.legislature.ca.gov/faces/billSearchClient.xhtml 

Bales, R. C., Rice, R., & Roy, S. B. (2015). Estimated loss of snowpack storage in the Eastern Sierra Nevada with climate warming. Journal of Water Resources Planning and Management, 141(2), 04014055.

Bateman, P. C., Clark, L. D., Huber, N. K., Moore, J. G., & Rinehart, C. D. (1963). The Sierra Nevada batholith: a synthesis of recent work across the central part (No. 414-D).

Bennett, E. (2025). Take action against california SB 818!. Mountain Lion Foundation. https://mountainlion.org/2025/04/01/take-action-against-california-sb-818/ 

Blakely, Logan & Reno, Matthew & Broderick, Robert. (2018). Evaluation and Comparison of Machine Learning Techniques for Rapid QSTS Simulations.

Broberg, G. (2019). The Golden Rush to the Golden State: An Analysis of the Development of California, 1845-1960.

Bolas, E. C., Pingatore, A. D., Mathur, M., Blumstein, D. T., Sikich, J. A., Smith, J. A., … & Blakey, R. V. (2025). Human recreation influences activity of a large carnivore in an urban landscape. Biological Conservation, 301, 110812.

California Department of Conservation. (2025). California Important Farmland Finder (CIFF). Division of Land Resource Protection. https://maps.conservation.ca.gov/DLRP/CIFF/

California Department of Fish and Wildlife. (2017). Human/Wildlife Interactions in California: Mountain Lion Depredation, Public Safety, and Animal Welfare – Amendment to Department Bulletin 2013-02.

California Department of Fish and Wildlife. Mountain Lions in California. CDFW Conservation. https://wildlife.ca.gov/Conservation/Mammals/Mountain-Lion/Depredation 

California Department of Fish and Wildlife. (2020). California Fish and Game Code. The California Biologist’s Handbook. https://biologistshandbook.com/regulations/state-regulations/state-fish-and-game-code/

California Department of Fish and Wildlife. (2022). Annual Mountain Lion Depredation Summary Table. Mountain Lions in California. https://wildlife.ca.gov/Conservation/Mammals/Mountain-Lion/Depredation 

California Department of Fish and Wildlife. (2025). Mule Deer Range – CWHR M181 [DS795] – Dataset – California Open Data. California Wildlife Habitat Relationships. https://data.ca.gov/dataset/mule-deer-range-cwhr-m181-ds7951 

California Department of Forestry and Fire Protection. (2024). Cal Fire Hub. CAL FIRE Hub. https://hub-calfire-forestry.hub.arcgis.com/ 

Cowling, R. M., Rundel, P. W., Lamont, B. B., Arroyo, M. K., & Arianoutsou, M. (1996). Plant diversity in Mediterranean-climate regions. Trends in Ecology & Evolution, 11(9), 362-366.

Dellinger, J. A., & Torres, S. G. (2020). A retrospective look at mountain lion populations in California (1906-2018). CALIFORNIA FISH AND WILDLIFE JOURNAL, 106(1), 66-85.

Drummond, O. (2021). A Geographical Analysis of Forecasted Threats–Fire And Land Use Change–On Conflict Between The Mountain Lion (Puma Concolor) And People In San Diego County, California (Doctoral dissertation).

El Dorado County Government. (2025). About Us – El Dorado County. https://www.eldoradocounty.ca.gov/County-Government/About-Us 

El Dorado County Government. (2024). Planning Division – El Dorado County. GOTNET. https://www.eldoradocounty.ca.gov/Land-Use/Planning-and-Building/Planning-Division 

Grigione, M. M., Beier, P., Hopkins, R. A., Neal, D., Padley, W. D., Schonewald, C. M., & Johnson, M. L. (2002). Ecological and allometric determinants of home‐range size for mountain lions (Puma concolor). Animal Conservation, 5(4), 317-324.

Grimes, K. (2025). Katy Grimes. California Globe. https://californiaglobe.com/fr/one-year-after-fatal-mountain-lion-attack-partisan-democrats-inflict-more-harm-on-brooks-family/ 

Harvey, J. A., Martins, Q., Dougherty, K. D., Hettena, A. M., Garcelon, D. K., & Benson, J. F. (2025). Functional habitat characteristics influencing spatial variation in mountain lion predation in a human-dominated landscape. Biological Conservation, 302, 110975.

Holmes, B. R., & Laundré, J. W. (2006). Use of open, edge and forest areas by pumas Puma concolor in winter: are pumas foraging optimally?. Wildlife Biology, 12(2), 201-209.

LaBarge, L. R., Evans, M. J., Miller, J. R., Cannataro, G., Hunt, C., & Elbroch, L. M. (2022). Pumas Puma concolor as ecological brokers: a review of their biotic relationships. Mammal Review, 52(3), 360-376.

LANDFIRE. (2023). LANDFIRE – Data. LANDFIRE Map Viewer. https://www.landfire.gov/viewer/ 

Logan, K. A., & Sweanor, L. L. (2001). Desert puma: evolutionary ecology and conservation of an enduring carnivore. Island Press.

Lorand, C., Robert, A., Gastineau, A., Mihoub, J. B., & Bessa-Gomes, C. (2022). Effectiveness of interventions for managing human-large carnivore conflicts worldwide: Scare them off, don’t remove them. Science of the Total Environment, 838, 156195.

Lupo, T., & Brinkhaus, A. (1996). Mountain lion and human activity in California: testing speculations. Wildlife Society Bulletin 1996, 24(3):451-460.

McCain, E. B. (2008). Daily activity patterns of mountain lions (Puma concolor) in relation to the activity of their prey species in southern Arizona.

McKinney, B. P. (2001). Mountain lions, deer, and predator control. In Proceedings of a symposium. The role of predator control as a tool in game management. Texas Agricultural Research and Extension Center, Kerrville, Texas, USA (pp. 70-73).

Mountain Lion Foundation. (2023). https://mountainlion.org/ 

Murphy, K. M., Felzien, G. S., Hornocker, M. G., & Ruth, T. K. (1998). Encounter competition between bears and cougars: some ecological implications. Ursus, 55-60.

Parsons, M. A., George, B. E., & Young, J. K. (2024). Evaluating The Efficacy of Aversive Conditioning of Mountain Lions with Hounds. Canadian Wildlife Biology & Management (CWBM), 13(2).

Pepin, N. C., Arnone, E., Gobiet, A., Haslinger, K., Kotlarski, S., Notarnicola, C., & Adler, C. (2022). Climate changes and their elevational patterns in the mountains of the world. Reviews of Geophysics, 60(1), e2020RG000730.

Radeloff, V. C., Hammer, R. B., Stewart, S. I., Fried, J. S., Holcomb, S. S., & McKeefry, J. F. (2005). The wildland-urban interface in the United States. Ecological Applications, 15(3), 799-805.

Safford, H. D., Butz, R. J., Bohlman, G. N., Coppoletta, M., Estes, B. L., Gross, S. E.,  & Wuenschel, A. (2021). Fire ecology of the North American Mediterranean-climate zone. Fire ecology and management: Past, present, and future of US forested ecosystems, 337-392.

Sun, F., Berg, N., Hall, A., Schwartz, M., & Walton, D. (2019). Understanding end‐of‐century snowpack changes over California’s Sierra Nevada. Geophysical Research Letters, 46(2), 933-943.

Suraci, J. P., Clinchy, M., Zanette, L. Y., & Wilmers, C. C. (2019). Fear of humans as apex predators has landscape‐scale impacts from mountain lions to mice. Ecology letters, 22(10), 1578-1586.

Tesky, Julie L. (1995). Puma concolor. In: Fire Effects Information System, [Online]. 

U.S. Department of Agriculture, Forest Service, Rocky Mountain Research Station, Fire Sciences Laboratory (Producer). Available: 

www.fs.usda.gov/database/feis/animals/mammal/puco/all.html

Thrower, J. (2005). Earth Island institute v. United States forest service: salvage logging plans in star fire region undermine sierra Nevada framework. Ecology Law Quarterly, 32(3), 721-728.

Treves, A., & Karanth, K. U. (2003). Human‐carnivore conflict and perspectives on management worldwide. Conservation biology, 17(6), 1491-1499.

U. S. Census Bureau. (2023). El Dorado County, California. Explore Census Data. https://data.census.gov/profile/El_Dorado_County,_Cal…?g=050XX00US06017

U.S. Census Bureau. (2023). U.S. Census Bureau, Department of Commerce – Tiger/Line Shapefile, 2023, county, El Dorado County, CA, linear hydrography. Catalog. https://catalog.data.gov/dataset/tiger-line-shapefile-2023-county-el-dorado-county-ca-linear-hydrography 

U.S. Census Bureau. (2022). U.S. Census Bureau, Department of Commerce – Tiger/Line Shapefile, 2021, county, El Dorado County, CA, all roads. Catalog. https://catalog.data.gov/dataset/tiger-line-shapefile-2021-county-el-dorado-county-ca-all-roads 

U. S. Census Bureau. (2024). U.S. Census Bureau, Department of Commerce Catalog. https://catalog.data.gov/dataset/tiger-line-shapefile-2022-county-el-dorado-county-ca-area-hydrography 

United States Department of Agriculture. (2020). Mountain Lions in the Central Sierra. https://www.fs.usda.gov/Internet/FSE_DOCUMENTS/fseprd706878.pdf 

United States Department of Agriculture. (2025). USDA Forest Service FSGeodata Clearinghouse – Download National Datasets. https://data.fs.usda.gov/geodata/edw/datasets.php?dsetCategory=boundaries 

U.S. Geological Survey (USGS) Gap Analysis Project (GAP). (2024). Protected Areas Database of the United States (PAD-US) 4.1: U.S. Geological Survey data release, https://doi.org/10.5066/P96WBCHS.

 Van Wagtendonk, J. W., Fites-Kaufman, J. A., Safford, H. D., North, M. P., & Collins, B. (2018). Sierra Nevada Bioregion. Fire in California’s ecosystems, 249-279.

Valavi, R., Elith, J., Lahoz‐Monfort, J. J., & Guillera‐Arroita, G. (2021). Modelling species presence‐only data with random forests. Ecography, 44(12), 1731-1742.

Vickers, T. W., Sanchez, J. N., Johnson, C. K., Morrison, S. A., Botta, R., Smith, T., & Boyce, W. M. (2015). Survival and mortality of pumas (Puma concolor) in a fragmented, urbanizing landscape. PloS one, 10(7), e0131490.

Wilmers, C. C., Nisi, A. C., & Ranc, N. (2021). COVID-19 suppression of human mobility releases mountain lions from a landscape of fear. Current Biology, 31(17), 3952-3955.

Disclaimer

This senior project paper, authored by Brittany Goins, is an undergraduate work submitted as part of their academic requirements in 2025. The paper is based on 2024 mountain lion sightings data provided by El Dorado County Carnivore Project. It is not a thesis and has not been peer-reviewed, though it may undergo peer review in the future.

The content of this paper is protected by copyright © 2025 Brittany Goins. All rights reserved. Reproduction, distribution, or use of any part of this paper without explicit permission from the author is prohibited.

The data used in this paper is the property of El Dorado County Carnivore Project. Any use, replication, or analysis of the data, in whole or in part, requires prior written consent from El Dorado County Carnivore Project. For permissions or inquiries, please use our contact form.