Mountain Lions & Freeways

Acknowledgments

Dr. Sanchayeeta Adhikari

Thank you for believing that I could do that which even I thought I could not.  During some of the most academically challenging times during my college career, your words of encouragement gave me the confidence to know that I could succeed.

Dr. Regan Maas

Thank you for the enthusiasm that you showed for this project when it was initially proposed.  I must admit that I wasn’t sure if my compass was accurate enough to get me where I wanted to go.  Your guidance throughout the process always left me feeling that I was on the right track.

Dr. Soheil Boroushaki

Thank you for providing me with not only the GIS skills required to do a project like this but, more importantly, for teaching me how to think through geospatial problems.

Luis DeVera

Thank you for always being available for our group and for reminding us to breathe throughout the semester.  Also, thanks for your feedback on my writing.  It means a lot to me that you felt it was worthy.

Melanie Hoffman

Thank you for taking the time to look over my writing to make sure it was as grammatically sound as it could be.  Your corrections and suggestions are very much appreciated.

Table of Contents

Acknowledgments

Abstract

1. Introduction

          1.1 Biodiversity

                    1.1.1 Global Biodiversity

                    1.1.2 North American Biodiversity

                    1.1.3 California Biodiversity

2. Background

          2.1 Large Carnivore in a Fragmented Landscape

                    2.1.1 Mountain Lion (Puma concolor)

          2.2 The Problem

          2.3 Effects of Habitat Fragmentation

                    2.3.1 Vehicle Collision Mortality

                    2.3.2 Intraspecific Strife

                    2.3.3 Inbreeding Depression

          2.4 Fragmentation Mitigation Measures

                    2.4.1 Warning Signs

                    2.4.2 Bridges and Tunnels

3. Project

          3.1 Project Aims

                    3.1.1 Identify Habitat Characteristics and Potential Linkage Corridors

4. Methods

          4.1 Study Area

                    4.1.1 Northwest Los Angeles Region

          4.2 Data and Sources

                    4.2.1 GIS Feature Data

                    4.2.2 Existing Map Data

          4.3 Analyses

                    4.3.1 Habitat Suitability Analysis

                    4.3.2 Least-Cost Path Analysis

5. Results

          5.1 Home Range Composition

                    5.1.1 Land Cover Composition

                    5.1.2 Land Cover Reclassification

                    5.1.3 Elevation Composition

                    5.1.4 Elevation Reclassification

                    5.1.5 Slope Composition

                    5.1.6 Slope Reclassification

6. Discussion

          6.1 Summary

          6.2 Limitations

                    6.2.1 Data Limitations

                    6.2.2 Time Constraints

          6.3 Conclusions

Bibliography

Appendix A – Home Ranges

Appendix B – Home Range Land Cover Composition

Appendix C – Home Range Elevation Composition

Appendix D – Home Range Slope Composition

Appendix E – Telemetry Locations

Appendix F – Vehicle Collision Mortality Locations

Appendix G – Home Range Overlaps Contributing to Intraspecific Killing

Appendix H – Home Range Overlaps Contributing to Inbreeding

Abstract

1. Introduction

            1.1 Biodiversity

1.1.1 Global Biodiversity

There is consensus in the scientific community that Earth is currently experiencing a sixth mass extinction crisis (Moseley et al. 2014, 287).  The previous extinctions were all caused by natural events such as cataclysmic volcanic activity, extreme glaciation, or asteroid strikes (as was likely the case with the Cretaceous-Paleogene event which decimated the dinosaurs).  The current extinction, however, can be attributed solely to anthropogenic causes.  Since the beginning of the Industrial Age, not only has there been an exponential rise in global pollution of the world’s oceans and atmosphere but also rampant development which continues to this day virtually unchecked.  The spread of human development and the associated encroachment on the natural landscape is leaving the habitats of many animal species fragmented into patches that are no longer large enough to sustain them.  Further, these changes are happening at rates faster than the biota is able to adapt causing more vulnerable species to die off.

1.1.2 North American Biodiversity

In the United States of America (U.S.) in particular, the effects of anthropogenic species extinction (or more often, local extirpation) are more apparent.  During the 19th century, in the U.S., human society generally held an anthropocentric attitude towards the environment – one in which human beings were considered to be separate from nature and that the environment was meant to be used by humankind.  Furthermore, along with the westward expansion of settlement came large growth in ranching and farming.  There were two major impacts of this expansion.  First, the presence of confined livestock provided potentially easy prey for predatory animals that may have already inhabited the land.  Animals such as the gray wolf (Canis lupus), the mountain lion (Puma concolor), and the North American brown bear (Ursus arctos) are some of the keystone species that inhabited much of the American west while settlers moved towards the western coast in fulfillment of “manifest destiny.”  The perceived threat of predation on livestock by these large carnivores caused them to be thought of as vermin to be exterminated.  The second environmental challenge that emerged was the rampant deforestation of the 1800s.  By the early-20th century, the vast majority of virgin forest in the U.S. had been cleared out.  Unbeknownst to those responsible for the deforestation that occurred, massive tracts of the ecosystem, along with the biota that depended on it, were destroyed.  It was during this period of time that a paradigm shift occurred in the U.S. government’s attitude toward the environment.  In 1916, the U.S. National Park Service (NPS) was formed to manage the country’s wilderness lands.  Even though it had already been declared a national park in 1872, the first wilderness land to be managed under the jurisdiction of the new NPS was Yellowstone National Park.  Yellowstone National Park served as a living laboratory to demonstrate the detrimental effect of the elimination of the region’s apex predator – the gray wolf – and the positive effect the reintroduction of the animal had on the ecosystem.  In the absence of its primary predator to keep its own population in check, the wolves’ natural prey – the elk (Cervus canadensis) – effectively had free reign on the producer level of the trophic structure.  The result was massive overgrazing of the region’s flora which had damaging effects on biotic and abiotic elements of the ecosystem.  When wolves were reintroduced in the mid-1990s, measurable improvement and restoration to the region’s ecosystem was seen (Beschta and Ripple 2009).  The Yellowstone case proved how important apex predators are for ecosystems.

1.1.3 California Biodiversity

Recognizing the reality of species loss and the negative ecological impacts these losses can incur, the state of California emerged as a leader in environmental protection when it enacted the California Endangered Species Act (CESA) of 1970.  This law preceded the United States government’s enactment of a similar federal law in 1973.  Seeing a need for further protection of its wildlife, the state passed the California Wildlife Protection Act (CWPA) in 1990 which required the state to spend at least $30 million per year to protect the state’s wildlife habitat (Mountain Lion Foundation 2019).  In 1996, the California Floristic Province (CFP) was designated a biodiversity hotspot by Conservation International (Environmental Science 2019).  However, in spite of these advances in environmental stewardship, California is still threatened with biodiversity loss.  Examples of past large mammal extirpation include the California brown bear (Ursus arctos californicus), the gray wolf, and the jaguar (Panthera onca) (Critical Ecosystem Partnership Fund 2019).  The California condor (Gymnogyps californianus) was functionally extirpated from California by 1987 but conservation efforts have since successfully reintroduced them to the wild (California Department of Fish & Wildlife 2019).  According to the Critical Ecosystem Partnership Fund (CEPF), twenty of the CFP’s 150 native mammal species are found nowhere else in the world (Critical Ecosystem Partnership Fund 2019).  As of August 2019, the CESA lists forty-nine of the state’s animals as endangered, forty as threatened, and eight as candidates to be listed as either endangered or threatened (California Department of Fish & Wildlife 2019).  Although a particular species may not be listed as endangered or threatened on a regional or global level, it may still be afforded some legal protection at the state level.  Aside from the broader benefit of the state-mandated wildlife habitat protection fund, the CWPA also gives the state the authority to designate a species as “specially protected,” as was the case with California’s mountain lion (California Department of Fish & Wildlife 2019).

2. Background

            2.1 Large Carnivore in a Fragmented Landscape

2.1.1 Mountain Lion (Puma concolor)

Description

Mountain lions, also commonly known as cougar, puma, or catamount, are the largest species of the lesser cats (subfamily Felinae).  They generally have a slender, athletic build with males having a head and body length in the range of 1020 mm to 1540 mm (with tails of approximately 680 mm to 960 mm in length) and weighing between 36 kg to 120 kg.  As mountain lions are a sexually dimorphic species, head and body length for females range from just under 860 mm to 1310 mm (with tails of approximately 630 mm to 790 mm) while weights range between 29 kg to 64 kg.  The fur is generally short and coarse with coloration ranging from light tan to tawny brown.  Mountain lions are highly secretive animals and live solitary lives with socialization only occurring with mother and cubs or with mates during short (one to six-day) breeding periods (Dewey and Shivaraju 2003).

Prey

In Southern California, the primary prey item for mountain lions is the mule deer (Odocoileus hemionus).  However, due to the abundance of such smaller mammals in the region, mountain lions may opportunistically prey on coyote (Canis latrans), raccoons (Procyon lotor), and Virginia opossum (Didelphis virginiana) as well (Dickson and Beier 2002).  Mountain lions will consume approximately 860 kg to 1300 kg of prey per year.  This is roughly equal to forty‑eight deer consumed annually (Dewey and Shivaraju 2003).

Habitat & Range

Mountain lions can be found throughout North and South America.  As such, they are highly adaptable animals and can be found in many different environments.  In California, their home range includes the vast majority of the state with the exception of the Central Valley, the Mojave Desert, and the Colorado Desert.  Dickson and Beier (2002) found that mountain lions that inhabited the Santa Ana Mountain Range (SAMR), just southeast of the Los Angeles area, tended to show highest preference to landscapes abundant in chaparral.  Second and third highest selection went to scrub and riparian zones respectively.  The study further outlines slight variability in preference between wet and dry seasons but, overall, there was no significant difference.  Furthermore, their findings showed that males and females were fairly consistent regarding their range choice from season to season.  This particular study also showed that topography was generally rugged and that elevation preference was between 275 m and 310 m with slope preference between 12° and 13° (21.26% – 23.09%).  However, research by Logan and Sweanor (2001, 15) describe North American mountain lion presence at elevation ranges between sea level up to 3350 m.  Further, Beier (2010, 205), in his contribution to Cougar: Ecology and Conservation, similarly described lion presence from sea level up to 4000 m.  Nicholson, Bowyer, and Kie (1997) showed high abundance of mule deer at elevations well in excess of 2500 m.  Since mountain lions will often follow their prey to higher elevations (Ruth and Murphy 2010, 170), it can be assumed that they have high tolerance to such altitudes.  With respect to slope preference, LaRue (2007) noted that mountain lions hold a similar disregard for wide variance in slope.  For that particular study, slope was given “minor importance” in terms of cost value (~0.14 normalized weighting) in the suitability analysis that was performed.  Further, it was stated that slope values >15% and 5% – 15% were given equal weighting while <5% slope was valued at approximately half the weight of the other two classes.   Intermittent streams exist in the study area as well as springs, seeps, and other water sources.  The animals in the study also showed a tendency to avoid paved roads by preferring areas that were greater than 1500 m from high-speed roadways.  However, once a home range was established, there was less of a tendency to avoid roads within the home range.  This was especially true when these roads intersected with riparian zones.  In Dickson and Beier’s study, adult male mountain lions (n=2) had an average multiyear home range size of 470 km² whereas adult females (n=13) had an average range of 81 km².  Grigione et al. (2002) studied mountain lion home range sizes in three areas: The Diablo Range (southeast of the San Francisco Bay area), the Sierra Nevada, and the Santa Ana Mountains.  In this study, range size data from the Diablo Range and the Santa Ana Mountains was combined and referred to as the Coastal Ranges.  This study found that Coastal Range males (n=5) had home range averages of 350 km² in the winter and 300 km² in the summer while females (n=22) had home range averages of 100 km² in the winter and 90 km² in the summer.  The Sierra Nevada population, however, enjoyed much larger ranges.  The study found that Sierra Nevada males (n=11) had home range averages of 469 km² in the winter and 723 km² in the summer while females (n=19) had home range averages of 349 km² in the winter and 541 km² in the summer.  Regarding range size, there tends to be differing findings in the literature due to the variability in study areas.  As previously stated, mountain lions are ubiquitous animals who have managed to adapt to a wide variety of habitats.  However, due to their predation requirements, they still require ranges with large enough ungulate carrying capacities to sustain them. 

            2.2 The Problem

Over the course of the 20th century, the rise in urban development has become a major threat to biological diversity.  Nowhere is this more evident than in the Los Angeles basin where the human-built environment, its freeway system in particular, has severely fragmented the habitat of the region’s apex predator, the mountain lion.  The Santa Monica Mountains National Recreation Area (SMMNRA), home to a population of Los Angeles mountain lions, has a total area of approximately 622 km² (National Park Service 2018).  By comparison with the home range sizes noted in the previously discussed study, conducted by Dickson and Beier, this is enough space for one or two male lions to live peacefully.  Naturally, females occupy the area as well, so problems arise when offspring reach dispersal age.  With the Santa Monica Mountains (SMM) being bound virtually on all sides by highly trafficked roadways, vehicle collision mortality, intraspecific strife, and inbreeding depression are serious threats to the long-term viability of the population.

            2.3 Effects of Habitat Fragmentation

2.3.1 Vehicle Collision Mortality

A major consequence of the fragmentation caused by the region’s freeway system is the deaths of these animals caused by motor vehicle collisions when attempting to cross these barriers.  As of 2018, there have been eighteen mountain lions killed by motor vehicle collisions since NPS started researching them in 2002 (National Park Service 2018).  This count was increased to nineteen when a male mountain lion, P-61, was struck and killed on Interstate 405 in September 2019 (Figure 1).  NPS biologists believe that the animal may have been attempting to leave the area after an altercation with another male that was presumably defending its territory (National Park Service 2019).  Further, in a thirteen-year study (2001-2013) conducted by Vickers et al. (2015), it was reported that with an already low 55.8% overall annual survival rate for mountain lions, 28% of the deaths were due to vehicle collisions.  According to the same study, the population of the SAMR mountain lions contains only seventeen to twenty-seven individuals.  Vehicle collisions as well as the inbreeding experienced by this population surely puts it at similar risk of extirpation as the SMM population.

2.3.2 Intraspecific Strife

Another consequence of fragmented habitat is the frequency of intraspecific killing that can occur when mountain lions come into contact.  As previously stated, mountain lions are solitary animals that tend to socialize only in mother/cub denning scenarios or during mating sessions.  During the course of ten years, Riley et al. (2014) tracked twenty-six radio-collared mountain lions.  Of these twenty-six, six died due to intraspecific strife with the additional death of an uncollared young male bringing the total to seven.  In five of the seven cases, the researchers were able to identify the aggressor in the killings, which was always an adult male with the victim being either offspring, a brother, or a previous mate.  Additionally, they reported that 50% of overall mortalities of known cause were due to deaths caused by intraspecific killing.  Considering the range requirements for mountain lions, adult males in particular, it should come as no surprise that a confined habitat that is pushing the limits of mountain lion carrying capacity – and the territory overlap that is intrinsically linked to such a scenario – is likely to promote mortal violence within the species.  Even if deaths are not directly caused by fighting, the example of P­-61 noted previously shows that high tension between individual animals has the potential to lead to other causes of death.  The high death rates, low life expectancy, and low levels of genetic diversity have caused the researchers to deem SMM a population sink – where death rates exceed birth rates (Riley et al. 2014).  For examples of death caused by intraspecific strife within the study area, see Appendix G.

2.3.3 Inbreeding Depression

Finally, a primary consequence of the lack of ability to disperse is inbreeding depression.  Under normal circumstances (i.e. no fragmentation), young mountain lions will move out away from their native range to establish home ranges of their own.  However, if they are unable to do so, there is a good chance that they will mate with lions that are closely related.  This is especially true in the SMM lion population.  Riley et al. (2014) genotyped forty-two individual lions and found that the “genetic diversity in SMM lions, prior to 2009, was lower than that for any population in North America except in southern Florida, where inbreeding depression led to reproductive failure.”  Due primarily to the SMMNRA’s tightly bounded 622 km² of land (contrasted with previously reported individual home ranges), the level of inbreeding in the SMM mountain lion population is so high that it is predicted that the population will be extirpated within the next fifty years (Benson et al. 2016).  Results of a study by Ernest et al. (2014) provide a similarly bleak outlook for the mountain lion population in the Santa Ana Mountains.  It has been shown, however, that the introduction of a single male lion from an outside population can greatly enhance the gene pool.  According to Riley et al. (2014), in February 2009, mountain lion P-12 was able to successfully cross U.S. 101 from the north and establish a home range in the Santa Monica Mountains.  Once in SMM, he was able to mate with a local female twice, which gave the local population a much-needed boost in potential genetic diversity.  These findings appear to support claims made by the “one-migrant-per-generation” rule discussed by Mills and Allendorf (1996).  For examples of inbreeding within the study area, see Appendix H.

            2.4 Fragmentation Mitigation Measures

Habitat fragmentation negatively affects local mountain lion populations.  However, there are solutions to the problem that have varying levels of effectiveness.

2.4.1 Warning Signs

The most widespread vehicle collision mitigation measure would likely be the standard black on yellow warning sign that can often be seen in rural areas (Beckmann et al. 2010, 57).  The problem with permanent warning signs is that they are most often ignored by drivers, resulting in extremely low levels of effectiveness.  An improvement over the standard warning sign is the dynamic, or variable message, sign (DMS/VMS) which may utilize brightly colored flags, flashing lights, or can be turned on/off depending on seasonal variability of animal migration.  A study by Hardy et al. (2006) reported that average vehicle speeds were in fact significantly lower in the presence of these enhanced warnings over what was reported in the presence of standard signs.  A further improvement on enhanced sign designs are systems that incorporate animal detection sensors.  These sensors will trigger a sign to turn on when large animals are detected within a certain range.  The advantage of this type of system is that the signs are only visible if an animal is actually approaching the roadway, which should have the effect of reducing driver complacency.  The drawback to this system, however, is that it needs to be reliable so as to not erode driver confidence in its efficacy (Hardy et al. 2006).

2.4.2 Bridges and Tunnels

Although warning signs may provide some level of wildlife-vehicle collision mitigation, they still do not address the issue of fragmentation.  The animals are at high risk of collision mortality in spite of the fact that there are signs warning the road-going public of their presence.  The only truly effective means of reconnecting the patchwork of habitat (outside of elimination of the roadway altogether) is by constructing overpass/underpass corridors that eliminate the need for wildlife to traverse the roadway surface entirely.  Bridges and tunnels are the most effective tools to achieve this goal.  There are drawbacks however.  While there is minimal monetary cost involved in the installation of warning signs along the roadway, the construction cost of a bridge or tunnel can be prohibitively expensive.  In spite of this, fundraising is currently underway for a planned wildlife crossing bridge that will span U.S. 101 just west of the San Fernando Valley (northwest of downtown Los Angeles).  Although they do incur high cost for implementation, corridors that employ bridges and tunnels (including preexisting drainage culverts) have proven to be useful tools for allowing safe wildlife migration from one side of roadway barriers to the other.  A study by Ng et al. (2004) used remotely triggered cameras stationed at fifteen different passage locations to capture images of multiple wildlife species utilizing these structures to move about.  There is similar evidence of multiple types of roadway crossing use demonstrated in Banff National Park in Alberta, Canada.  Remote-cameras have captured images of mountain lions, elk, and gray wolves all using overpasses and underpasses of various types along the Trans-Canada Highway (Beckmann et al. 2010, 171).  There is no single solution to habitat fragmentation.  However, a holistic approach to habitat linkage using an appropriate combination of fencing, signage and overpasses/underpasses can be used to promote and enhance habitat connectivity.

3. Project

            3.1 Project Aims

3.1.1 Identify Habitat Characteristics and Potential Linkage Corridors            

This project aimed to accomplish two objectives:  The first was to identify the landscape characteristics of mountain lion home ranges within portions of Ventura and Los Angeles Counties and compare the prevalence of these traits (or lack thereof) to what is found in the study area overall.  The second was to determine potential linkage corridors that might serve to reconnect the existing patchwork of habitat in the area.  This second objective included the discovery of suitable locations for potential future roadway crossing points.  These crossings are crucial as they will aid in promoting unimpeded movement of the region’s mountain lion population by reducing vehicle collision mortality and intraspecific strife.  Less restricted movement will also result in higher rates of successful dispersal leading to an accompanying increase in gene pool diversification – all of which will result in an increase in species persistence.

4. Methods

            4.1 Study Area

4.1.1 Northwest Los Angeles Region

The northwest Los Angeles region is unique amongst major urban centers in that it is situated in the middle of multiple mountain ranges.  The two major range systems in Southern California are the east-west trending Transverse Ranges and the northwest-southeast trending Peninsular Ranges.  The Peninsular Ranges include the Santa Ana Mountains and the Temescal Mountains southeast of Los Angeles.  This project, however, focused on ranges found in Ventura and Los Angeles Counties (Figure 2).  These counties not only contain larger ranges, such as the San Gabriel, Santa Susana, Santa Monica, and Topatopa Mountains, but also smaller ranges, such as the Verdugo Hills and the Simi Hills.  Sharing large portions of land area with these ranges are protected areas such as Los Padres and Angeles National Forests as well as the much smaller Griffith Park (as of this writing, home to a lone male mountain lion, P-22).  Prior to the urban development of the San Fernando, Santa Clarita, and Simi Valleys, wildlife had safe and unimpeded access to all of the aforementioned habitat.  Now, the region is fragmented, making it much more difficult to traverse.  Not only is it more difficult for the animals to disperse from one region to another, but, due to the incurred edge effect as well as pavement replacing natural terrain, the usable area of habitat has also dramatically decreased.

            4.2 Data and Sources

4.2.1 GIS Feature Data

For this project, various types of GIS data were used to not only visualize the study area but also to perform the analyses that were required to determine suitable habitat and linkage corridors.  State and county boundary data were used to contextualize the study area and was sourced from California state government’s website (cal.gov).  California protected areas vector data was used to delineate natural landscapes that fall under government jurisdiction. These areas have very minimal to zero levels of development and, as such, are likely to have high potential for suitable habitat.  This data was sourced from the United States Geological Survey (USGS).  A digital elevation model (DEM) was also sourced from USGS and was used to determine and visualize elevation, perform analyses, and also to derive a slope raster dataset. The slope raster was used not only to determine and visualize slope characteristics in the study area but also to perform analyses.  The National Land Cover Database (NLCD) was sourced from the Multi-Resolution Land Characteristics (MRLC) Consortium and was used to determine and visualize the various land cover types within the study area and to perform analyses.  Road network data was sourced from Environmental Systems Research Institute (Esri).  This data was used to visualize the roadways found in the study area. Also, since vehicle collision mortality is a primary concern, this data was used to maximize the distance of potential habitat linkages to roadways in the habitat suitability analysis.  Stream vector data was used to visualize the study area’s flowing water as well as the associated riparian zones.  As mountain lions show preference for riparian zones, if they are available, streams were used as a minimization criterion in the suitability analysis.  A water bodies vector dataset was used to visualize natural, enclosed water bodies such as lakes, ponds, etc.  More importantly, since water bodies should be considered as restricted areas in the least-cost path analysis, this data was used to nullify water bodies found in the NLCD raster dataset.  The streams and water bodies vector data were both sourced from Esri.

4.2.2 Existing Map Data

Past and current mountain lion home range data was extracted from existing maps that were created by and sourced from the National Park Service (Appendix A). These maps were georeferenced in Esri ArcMap from which a vector feature dataset was then created to delineate individual mountain lion home ranges.  Through area tabulation analyses, the landscape characteristics of these ranges were derived. Not only did this area analysis reveal habitat preferences of the local mountain lion population but also informed criteria weighting decisions which were needed to derive the suitability surface (cost raster).

            4.3 Analyses

4.3.1 Habitat Suitability Analysis

The first phase of the habitat suitability analysis was to determine landscape preferences of individual mountain lions that currently occupy or have occupied the region in the recent past.  Once the existing home range maps were spatially referenced, the ranges were digitized as polygon features.  After the ranges were digitized, a slope raster was derived from the DEM. Using the Euclidean Distance tool, the streams and road network datasets were used to derive continuous surface distance rasters.  Elevation, slope, distance from roads, and distance from streams were all classified to ten natural breaks (Jenks) classes.  Using ArcMap’s Tabulate Area tool, the newly created home range feature classes were used as input feature zone data to derive statistical information regarding mountain lion preference for land cover as well as the ten classes of elevation and slope.

The second phase was to determine suitable habitat across the entire study area.  Based on the statistical data for individual territory preferences derived in the first phase, the land cover (Figure 3a, Figure 3b), elevation (Figure 4a, Figure 4b), and slope (Figure 5a, Figure 5b) surfaces were reclassified as cost rasters with values between one and ten.  A value of one was assigned to indicate highest preference and a value of ten was assigned to lowest preference.  Distance from roads (Figure 6a, Figure 6b) and streams (Figure 7a, Figure 7b) surfaces were reclassified in a linear fashion from one to ten with streams being considered as a minimization criterion (close distance = low cost) and roads being considered as a maximization criterion (close distance = high cost).  Prior to NLCD reclassification, the water bodies data was used to generate a new NLCD dataset with water bodies being nullified in the raster.  This prohibited potential corridors from crossing water bodies when performing the least cost path analysis.

Once these rasters were created, the Weighted Sum tool was used to generate a final cost roster (Figure 8). 

Weighting values for the final cost raster were adapted from similar research conducted by LaRue (2007) with slight modification due to small differences in criteria used.  In LaRue’s study, human density was used as a criterion.  Since human density strongly correlates with development, it was not included as a criterion for this study.  However, elevation, which was not included in LaRue’s study, was included.  Since the literature does not demonstrate any significant difference with regards to mountain lion preference or aversion to different elevation ranges compared to slope, they were given the same weight (Table 1).

Criterion	Weight
Land Cover	0.41841
Distance from Roads	0.19665
Slope	0.13808
Elevation	0.13808
Distance from Streams	0.10879

4.3.2 Least-Cost Path Analysis            

The first step of the least-cost path analysis was to determine termini for potential habitat linkage corridors.  For the Santa Monica Mountains, points were chosen from the western, central, and eastern regions of the range.  For the Santa Susana Mountains, points were chosen from the western and eastern regions.  Points were then chosen within Los Padres National Forest (one) and Angeles National Forest (two – one on either side of State Route 14).  Points were also chosen within Simi Hills, Verdugo Hills, and Griffith Park.  This made for a total of eleven initial termini.  Once the termini were determined, the least-cost path analysis was performed.

5. Results

            5.1 Home Range Composition

5.1.1 Land Cover Composition

With a mean of 68.27% of territory being classified as Shrub/Scrub, the results of area tabulation of land cover types present in the twenty-seven mountain lion home ranges in this project clearly show a high preference for this particular land cover (Figure 9).  On average, this gives a Shrub/Scrub a comparative use of 156.19% of the total amount of that land type in the entire study area (43.71%).  The second and third highest preference went to Developed – Open space (10.5%) and Grassland/Herbaceous (8.57%).  These two land cover types were closely correlated to the percentages found in the study area (9.33% for both) with comparative usages of 112.54% and 91.85% respectively.  Deciduous Forest had the lowest level of occurrence in the home ranges with a mean of 0.01%.  However, the reason for such a low level of occurrence for a land type that would normally be quite hospitable for mountain lions is simply because there is not much of it in the study area to begin with.  When compared to the total study area (.01%), home range use of Deciduous Forest correlates perfectly.  For the same reason, Evergreen Forest and Mixed Forest also both show very low levels of home range use (2.29% and 3.39% respectively).  Not surprisingly, when compared to occurrence in the study area (2.55%), Developed – High land cover had the lowest level of occurrence in the home ranges with a mean of 0.05%.  This equates to 1.96% home range use compared to the entire study area.  For details regarding individual home range land cover composition, see Appendix B.

5.1.2 Land Cover Reclassification

The fifteen different land cover types were clustered into four distinct categories with cost values of one, four, seven, and ten (Table 2).  The categories were chosen based on a consideration of natural mountain lion habitat preference (i.e. Shrub/Scrub) and avoidance of development (i.e. Developed – High).  These categories were given cost values of one and ten respectively.  Since a goal of this project was to determine travel corridors (as opposed to “permanent” home range potential), other factors considered that were not strongly tied to territory preference/avoidance included open space land cover types such as Developed – Open, pastures, crops, etc.  These land cover types are generally open enough for migration, yet have enough human presence to discourage long‑term residency and/or not enough ungulate prey to sustain a resident lion.  Due to the relative ease of passage through these types of land cover, this category was given a cost value of four.  The last category to be considered was land cover types that generally provided enough resistance to prevent easy passage through the landscape, yet was not impenetrable.  This included land cover types such as Woody Wetlands (where passage through water would prove difficult) or Developed – Low (where there exists enough human development to deter a lion from passing through the terrain).  Due to the relative difficulty of passage through these types of land cover, this category was given a cost value of seven.  Finally, due to being an impenetrable barrier to passage, the Open Water land cover type was nullified in the reclassification.

Land Cover	Cost Value
Shrub/Scrub	1
Mixed Forest	1
Evergreen Forest	1
Deciduous Forest	1
Developed – Open	4
Grassland/Herbaceous	4
Pasture/Hay	4
Cultivated Crops	4
Woody Wetlands	7
Emergent Herbaceous Wetlands	7
Developed – Low	7
Developed – Medium	10
Developed – High	10
Barren Land	10
Open Water	null

5.1.3 Elevation Composition

It is important to note that the home range elevation data shown here are only for radio‑collared mountain lions that inhabit (or have inhabited) an area immediately surrounding the San Fernando Valley within the study area.  As such, the data is somewhat limited.  For example, the study area has low occurrence of elevation ranges between 1164 m – 2248 m with extremely low comparative usages (home range vs. study area) of 2.35%, 0.28%, and 0.00% for the three highest elevation classes (Figure 10). However, as previous research has shown, mountain lions are highly adapted to virtually any elevation range upwards to 4000 m.  Thus, the results of this elevation analysis could lead to erroneously thinking that the animals show an aversion to these altitudes when, in actuality, they do not.  With regards to the lower elevations that predominately exist in the study area, most especially the three classes between 145 m – 645 m, mountain lions show a very high comparative use with values of 116.80%, 212.69%, and 205.95% respectively.  The range of 0 m – 145 m has a lower comparative use of 30.28% which is likely due to the fact that this elevation range is dominated by agriculture, open water (Pacific Ocean), or medium to high levels of development.  For details regarding individual home range elevation composition, see Appendix C.

5.1.4 Elevation Reclassification

As mentioned, mountain lions have adapted to thrive in a very wide range of elevations.  As such, cost values of one have been assigned to elevation classes between 145 m – 823 m.  A slight uptick in cost to a value of two was given to ranges of 0 m – 145 m and 823 m – 985 m due to lower usages of 30.28% and 33.59% (respectively) compared to study area availability.  With significantly lower comparative use, yet keeping in mind that mountain lions have no significant aversion to these elevation classes, the four ranges with the highest elevation were given cost values of three (Table 3).

Elevation	Cost Value
145 m – 308 m	1
308 m – 465 m	1
465 m – 645 m	1
645 m – 823 m	1
0 m – 145 m	2
823 m – 985 m	2
985 m – 1164 m	3
1164 m – 1382 m	3
1382 m – 1654 m	3
1654 m – 2248 m	3

5.1.5 Slope Composition

As was the case with the previously discussed elevation data, the slope figures revealed here must be considered with the understanding that the project’s home range data comes from ranges that exist in areas that are not conducive to high slope values (i.e. >60%).  In fact, ~94% of the study area consists of slope values <64%.  Because of the high abundance of these lower slope ranges, the third through sixth classes (which encompass slope values of 19.86% – 64.56%) contain very high comparative use values ranging from 137.15% up to 152.27% (Figure 11).  Note that the class with the highest comparative use (19.86% – 31.78%) is supported by Dickson and Beier’s (2002) findings.  Further, the class with the second highest prevalence in the study area (8.96% – 19.86%) is very closely correlated in terms of comparative use.  The class with the highest overall prevalence in the study area (0% – 8.96%) also has the lowest comparative use of all classes (31.91%).  As with what was found in the elevation analysis, this negative correlation can be attributed to agriculture, open water, or medium to high levels of development.  In spite of the low abundance of higher slope values in the range of 76.48% – 114.22%, the comparative use is sufficient to suggest low aversion to these slope values.  Finally, the class with the highest slope value (114.22% – 253.27%) computed a 61.54% comparative use, which results in the second lowest average use across all home ranges.  For details regarding individual home range slope composition, see Appendix D.

5.1.6 Slope Reclassification

The five classes that contain slope values from 19.86% up to 76.48% were given cost values of one due to their high levels of comparative use.  Although, the shallow terrain grade range of 0% – 8.96% would not necessarily be an impediment to mountain lion travel, there is a very low level of comparative use owed to the fact that there is much human development on flatter landscapes.  For this reason, the range was given a cost value of three.  The remaining four classes encompassing ranges of 8.96% – 19.86% (one class) and 76.48% – 253.27% (three classes) were all closely correlated with regard to home range use compared to study area availability.  As such, they were all given cost values of two (Table 4).

Slope	Cost Value
19.86% – 31.78%	1
31.78% – 42.71%	1
42.71% – 53.63%	1
53.63% – 64.56%	1
64.56% – 76.48%	1
8.96% – 19.86%	2
76.48% – 91.38%	2
91.38% – 114.22%	2
114.22% – 253.27%	2
0% – 8.96%	3

6. Discussion

            6.1 Summary

For the sake of clarity, here is how the terms path, corridor, and crossing point are used in the context of this project:

Path – Line drawn in the GIS (Esri ArcMap) as a result of a least-cost path analysis.

Corridor – Segment of a GIS-derived path that is located on the ground which extends ~5km or less in either direction of a crossing point.

Crossing point – Point on the ground where a potential roadway crossing can be located.

The results of this project’s analysis clearly demonstrate just how severely fragmented the landscape is within the study area.  Of the twenty-four initial start-point/end-point path combinations, many were either too developed along segments of the path, had an abundance of cropland that would not be conducive to travel, or were simply too far apart to be viable options.  The termini located in Verdugo Hills and Griffith Park were rejected for further analysis due to high development which almost completely isolates these areas from larger open spaces and, thus, makes them unlikely candidates for viable habitat corridors.  One particular path, from the eastern region of the Santa Monica Mountains to the section of Angeles National Forest on the eastern side of the study area, cuts diagonally through the heart of the San Fernando Valley.  This is clearly an unacceptable option for a habitat connectivity corridor.  As such, this terminus was also rejected for the study.  Another terminus, located in the western Santa Monica Mountains, marks the start point for a path to Los Padres National Forest but cuts virtually straight through the high level of cropland that is found along the western side of the study area (east of State Route 23) (Figure 12).  This terminus was also rejected.  The Simi Hills terminus was rejected simply because the resulting paths to the national forest areas exhibited line coincidence with the vast majority of the corresponding paths starting in the central region of the Santa Monica Mountains.

The results of analysis of the remaining six termini revealed seven potential linkage corridors (Figure 13).  The path that showed the highest potential for highway crossing points (and more generally, corridor locations) was one drawn from the central Santa Monica Mountains to the portion of Angeles National Forest that is situated southeast of State Route 14.  The first highway encountered by this path is U.S. 101, which will be addressed by the Liberty Canyon Wildlife crossing.  Once past U.S. 101, the path follows through the largely undeveloped and shrub/scrub bountiful Simi Hills, which is bounded to the north by State Route 118.  Mountain lions have been tracked crossing the highway near this point via the Corriganville Tunnel, a drainage culvert that passes under the roadway.  At this point, the path turns northeast towards the Interstate 5 / State Route 14 interchange.  In the vicinity of the Los Angeles Aqueduct Cascades and just northwest of Balboa Blvd., another potential crossing point can be found.  Just beyond this point, to the east, lies the boundary of Angeles National Forest.  This particular path is notable in that it is spatially concurrent with thirteen of the twenty-seven home ranges that were used to gather local mountain lion territory data.  This would reasonably suggest a high likelihood of usage of these crossing points/corridors.

Two paths originating in the Santa Susana Mountains show potential for enabling safe crossing of State Route 126.  The first potential crossing point is situated in the western end of the range between the cities of Fillmore and Buckhorn.  Although there does exist some level of crop cultivation in the vicinity of this point, it is fairly minimal compared to what is found further west and, to a lesser degree, to the east.  The second point, further east, is situated in close proximity to Camino Del Rio.  Although only three of the home ranges used in this project extend to the vicinity of the potential Fillmore/Buckhorn crossing point, six appear to be close enough to take advantage of a crossing in the area near Camino Del Rio.

Finally, the two remaining corridors that warrant further investigation include one connecting Los Padres National Forest to Angeles National Forest and another between the two noncontiguous sections of Angeles National Forest (divided by State Route 14).  For both of these areas, the vast majority of land cover consists of shrub/scrub with the remaining land consisting of grassland/herbaceous cover types.  For the former, a potential point crosses a section of Interstate 5 which forms the boundary between the protected and undeveloped confines of the two national forests.  For the latter, another point crosses State Route 14 in a similarly undeveloped region between the two separated sections of Angeles National Forest.  There is only one confirmed home range that is situated in the area of either of these potential crossing points.  However, it should be reiterated that the home ranges used in this project belong solely to (or have, in the past, belonged to) radio-collared mountain lions that are/were involved in the NPS mountain lion study.  It can be assumed that there are other lion subpopulations inhabiting these largely undeveloped regions.  Thus, with the goal of improving species survivability in the greater Los Angeles region, it would be prudent to consider the migration/dispersal needs of the uncollared animals that roam throughout these areas.

            6.2 Limitations

6.2.1 Data Limitations

The only data available for habitat preference analysis came from radio-collared mountain lions that inhabit (or inhabited) a somewhat limited topography in terms of elevation and slope.  The Santa Monica Mountains, Santa Susana Mountains, and Simi Hills all contain fairly similar topographic characteristics which resulted in some redundancy of data amongst the twenty-seven mountain lion home ranges that were analyzed.  Data from previous research has shown these animals to be quite tolerant of topographies that were not well represented in the acquired data.  Therefore, this previous information was used to more accurately inform ranking and weighting decisions.  If it were available, the project would have surely benefited from region-specific data derived from the area’s more mountainous terrain.

6.2.2 Time Constraints

Constraints on time prohibited deeper analysis of the acquired data, which could have served as the basis for a project-specific multi-criteria decision analysis.  This, along with a home range dataset that is more fully representative of the study area as a whole, would have resulted in higher ranking and weighting accuracy for the derivation of the final cost raster.  Regardless, since mountain lions are so highly adaptable, the Analytical Hierarchy Process data adapted from previous research provided a strong enough foundation to move forward with an initial exploration of the acquired data.

            6.3 Conclusions            

Rapid urbanization is having a dramatic effect on the environment and the consequences of human development can be seen in the shrinking levels of global biodiversity.  At the local scale, the fragmentation of habitat is clearly visible to those who choose to see it.  This fragmentation is having potentially irreversible effects on the viability of local species populations.  In the case of apex predators, such as the mountain lions of the Los Angeles region, extirpation of the species has the potential to greatly degrade the ecosystem.  It has been shown, in the case of Yellowstone National Park, that keystone species removal has devastating effects on the entire ecosystem.  There is no reason to think that the results would be any different if Southern California mountain lions were erased from the region’s landscape.  Fortunately, the solutions to these problems exist.  However, there is usually high cost associated to more complex solutions, such as bridges and tunnels, so funding is very often a barrier to effective habitat linkage.  There is hope however.  In her 2014 study concerning people’s attitudes towards wildlife in the Los Angeles area, geographer and wildlife advocate Claudia Hasenhüttl reports that the overall trend in wildlife value orientations is moving in favor of mutualism – an ecocentric view that emphasizes equality between humans and wildlife.  With increased support from all levels of society, governments and individuals alike, the challenge of habitat fragmentation is one that can be overcome.  Shrinking global biodiversity is a world problem that can be solved if people take action at the local level.  Reconnecting fragmented habitat is a step in the right direction towards solving this problem.

Bibliography

Beckmann, Jon P., Anthony P. Clevenger, Marcel P. Huijser, and Jodi A. Hilty. Safe Passages: Highways, Wildlife, and Habitat Connectivity. Washington, DC: Island Press, 2010.

Beier, Paul. Cougar: Ecology and Conservation. Edited by Maurice Hornocker and Sharon Negri. Chicago, Ill: University of Chicago Press, 2010.

Benson, John F., Peter J. Mahoney, Jeff A. Sikich, Laurel E. K. Serieys, John P. Pollinger, Holly B. Ernest, and Seth P. D. Riley. “Interactions Between Demography, Genetics, and Landscape Connectivity Increase Extinction Probability for a Small Population of Large Carnivores in a Major Metropolitan Area.” Proceedings of the Royal Society B: Biological Sciences 283, no. 1837 (2016): 20160957. doi:10.1098/rspb.2016.0957.

Beschta, Robert L., and William J. Ripple. “Large Predators and Trophic Cascades in Terrestrial Ecosystems of the Western United States.” Biological Conservation 142, no. 11 (2009): 2401-414. doi:10.1016/j.biocon.2009.06.015.

California Department of Fish & Wildlife. “California Condor.” Accessed October 29, 2019. https://www.wildlife.ca.gov/Conservation/Birds/California-Condor.

California Department of Fish & Wildlife. “Endangered and Threatened Animals List.” https://www.wildlife.ca.gov/Data/CNDDB/Plants-and-Animals (accessed September 19, 2019).

California Department of Fish & Wildlife. “Mountain Lions in California.” Accessed September  28, 2019. https://www.wildlife.ca.gov/Conservation/Mammals/Mountain-Lion.

Critical Ecosystem Partnership Fund. “California Floristic Province – Species.” Accessed September 28, 2019. https://www.cepf.net/our-work/biodiversity-hotspots/california-floristic-province/species.

Dewey, Tanya, and Anupama Shivaraju. “Puma Concolor (cougar).” Animal Diversity Web. 2003. Accessed December 11, 2018. https://animaldiversity.org/accounts/Puma_concolor/.

Dickson, Brett G., and Paul Beier. “Home-Range and Habitat Selection by Adult Cougars in Southern California.” The Journal of Wildlife Management 66, no. 4 (October 2002): 1235-245. doi:10.2307/3802956.

Environmental Science. “Biodiversity Hotspots ~ The California Floristic Province.” Accessed October 29, 2019. https://environmentalsciencesite.weebly.com/biodiversity-hotspots-the-california-floristic-provice.html.

Ernest, Holly B., T. Winston Vickers, Scott A. Morrison, Michael R. Buchalski, and Walter M. Boyce. “Fractured Genetic Connectivity Threatens a Southern California Puma (Puma concolor) Population.” PLoS ONE 9, no. 10 (2014). doi:10.1371/journal.pone.0107985.

Grigione, M. M., P. Beier, R. A. Hopkins, D. Neal, W. D. Padley, C. M. Schonewald, and M. L. Johnson. “Ecological and Allometric Determinants of Home-Range Size for Mountain Lions (Puma concolor).” Animal Conservation 5, no. 4 (2002): 317-24. doi:10.1017/s1367943002004079.

Hardy, A.R., J. Fuller, S.Lee, L.Stanley, and A. Al-Kaisy. 2006. Bozeman Pass Wildlife Channelization ITS Project final report. Western Transportation Institute, Montana State University, Bozeman.

Hasenhüttl, Claudia, “People’s Perception of Wildlife in Urban Parks: A Case Study in the Santa Monica Mountains and Griffith Park.” (Master’s thesis, California State University, Northridge, 2015).

LaRue, Michelle A., “Predicting Potential Habitat and Dispersal Corridors for Cougars in Midwestern North America.” (Master’s thesis, Minnesota State University, Mankato, 2007).

Logan, Kenneth A., and Linda L. Sweanor. Desert Puma: Evolutionary Ecology and Conservation of an Enduring Carnivore. Washington, DC: Island Press, 2001.

Mills, L. Scott, and Fred W. Allendorf. “The One-Migrant-per-Generation Rule in Conservation and Management.” Conservation Biology 10, no. 6 (1996): 1509-518. doi:10.1046/j.1523-1739.1996.10061509.x.

Moseley, William G. An Introduction to Human-environment Geography: Local Dynamics and Global Processes. Chichester: Wiley-Blackwell, 2014.

Mountain Lion Foundation. “California Proposition 117.” Accessed October 29, 2019. http://www.mountainlion.org/us/ca/Prop117/-ca-proposition117.php.

National Park Service. 2018. Lions in the Santa Monica Mountains? https://www.nps.gov/samo/learn/nature/pumapage.htm (accessed December 6, 2018).

National Park Service. 2019. Mountain Lion P-61 Killed Crossing 405 Freeway. https://www.nps.gov/samo/learn/news/mountain-lion-p-61-killed-crossing-405-freeway.htm (accessed September 19, 2019).

Ng, Sandra J., Jim W. Dole, Raymond M. Sauvajot, Seth P.D. Riley, and Thomas J. Valone. “Use of Highway Undercrossings by Wildlife in Southern California.” Biological Conservation 115, no. 3 (2004): 499-507. doi:10.1016/s0006-3207(03)00166-6.

Nicholson, M. C., R. T. Bowyer, and J. G. Kie. “Habitat Selection and Survival of Mule Deer: Tradeoffs Associated with Migration.” Journal of Mammalogy 78, no. 2 (1997): 483–504. https://doi.org/10.2307/1382900.

Riley, Seth P.D., Laurel E.K. Serieys, John P. Pollinger, Jeffrey A. Sikich, Lisa Dalbeck, Robert K. Wayne, and Holly B. Ernest. “Individual Behaviors Dominate the Dynamics of an Urban Mountain Lion Population Isolated by Roads.” Current Biology 24, no. 17 (2014): 1989-994. doi:10.1016/j.cub.2014.07.029.

Ruth, Toni K., and Kerry Murphy. Cougar: Ecology and Conservation. Edited by Maurice Hornocker and Sharon Negri. Chicago, Ill: University of Chicago Press, 2010. Vickers, T. Winston, Jessica N. Sanchez, Christine K. Johnson, Scott A. Morrison, Randy Botta, Trish Smith, Brian S. Cohen, Patrick R. Huber, Holly B. Ernest, and Walter M. Boyce. “Survival and Mortality of Pumas (Puma concolor) in a Fragmented, Urbanizing Landscape.” Plos One 10, no. 7 (2015). doi:10.1371/journal.pone.0131490.