Detecting Spatial Patterns of Sea Otters in Morro Bay, CA

Sea otters (Enhydra lutris) are a truly unique species. They are a member of the mustelid family and are the smallest marine mammal, weighing 35-100 pounds depending on the subspecies ¹ . Unlike other marine mammals, they lack blubber and depend on their dense fur coat and high metabolism to keep warm ² . Sea otters are carnivores, eating clams, urchins, and other benthic invertebrates. Their current range spans parts of Russia, Alaska, Washington state, and California ³ .

The species is listed as  Endangered  by the IUCN, and some U.S. stocks are listed as  Threatened  under the Endangered Species Act.

There are three subspecies of sea otters, but this project will focus on one of them: the southern sea otter (Enhydra lutris nereis) that lives strictly in California.

Sea otters were intensely hunted starting in the late 1700s, when Russian explorers discovered new islands in the North Pacific teeming with the species ⁴ . With the densest fur on the planet ⁵ , sea otters were highly sought after by Russian and American hunters. Between 1784 and 1911, the “Great Hunt” reduced the sea otter population from an estimated 300,000 to 2,000, forcing the species on the brink of extinction ^{1, 4} .

Realizing how dire the situation was, countries came together to establish the International Fur Seal Treaty in 1911, which granted much-needed protections to sea otters ⁶ . The treaty restricted their hunting, giving the remaining eight sea otter colonies a chance to survive. California’s remaining colony, located near Big Sur, had as few as 50 sea otters ⁷ . U.S. federal and state protections issued over the next few decades like the Marine Mammal Protection Act and Endangered Species Act would help grow the number and range of California’s otters ⁸ .

Today, 3,000 sea otters live between Half Moon Bay and Santa Barbara in California—a fraction of what the species once numbered (about 9,000-12,000) ⁹ .

Like in most parts of the California coast, Morro Bay experienced a sea otter extirpation during the “Great Hunt.” It was not until the 1970s when curious male sea otters ventured to Morro Bay ¹⁰ , but a permanent population did not take hold there until 2005. In the last decade, the number of sea otters inhabiting the estuary has greatly increased, and now, there is a stable population of about 40-50 otters in Morro Bay ^{9, 11} .

Researchers, though, don’t know how these sea otters are using habitats in Morro Bay. Questions about species distribution in the estuary and how that may be influenced by environmental considerations have yet to be addressed by the scientific community ^{12, 13} . This project attempts to fill some of these information gaps by answering the following research question:  

Are there spatial patterns in Morro Bay sea otter habitat use based on season?

In my project, I will map sea otter location data by season to see if species distribution is impacted by this key environmental factor. The goal of this investigation is to detect spatial patterns in sea otter habitat use in Morro Bay, which could provide insight relevant to sea otter conservation and estuarine management.

The geographic focus of this project is Morro Bay, a 10.3-mi ²  estuary. This small natural embayment is located near the center of the sea otter range with adjacent populations considered to be currently at or near carrying capacity. Currently, between 40 and 50 adult sea otters are regularly found within Morro Bay, with the majority occurring between the harbor entrance and the municipal boat launch at Tidelands Park.

Morro Bay includes jetty enclosed fore-bay, harbor, and estuarine back-bay portions. It is designated as a  State Marine Recreational Management Area , with a portion of the back bay designated as a  State Marine Reserve .

Morro Bay is situated within the larger Estero Bay. It has calmer waters, abundant vegetation, and more wildlife compared to Estero because of the protection provided by an extensive sandspit. This allows Morro Bay to serve as a nursery for fish, invertebrates, and an increasing number of sea otters ¹⁴ . Year-round water surface temperatures are also relatively mild and stable in the estuary (11°C), making it a safe haven from the cold, harsh open ocean. Morro Bay is comprised of brackish water, which is a mixture of saltwater from the Pacific Ocean and freshwater runoff from inland watersheds ¹⁵ .

This bay is comprised of harbor, estuarine, and rocky habitats, including a few small, seasonally persistent patches of large brown kelp (Macrocystis pyrifera) and eelgrass (Zostera marina). It is also characterized by extensive maritime infrastructure like piers, docks, and barriers, among other manmade structures. This contributes to the presence of heavy human traffic (recreational and commercial) in all but the most inaccessible saltmarsh regions. Areas surrounding Morro Bay were first settled in the early 1800s, and since then, human development has increased, putting anthropogenic stress on the estuary ¹⁵ .

To make it easier for surveyors to identify where sea otters are located, Morro Bay is broken up into 7 sections. From the harbor entrance to the back bay, they are: MBHM, MBCT, MBTP, MBSS, EMBA, TIDL, and MBBB.

1. MBHM (Morro Bay Harbor Mouth): Includes the area between the two jetties at the harbor entrance.
2. MBCT (Morro Bay Coleman Target): From Target Rock parking lot (west) to Coleman Beach (east).
3. MBTP (Morro Bay T-Piers): From North T-Pier to South T-Pier.
4. MBSS (Morro Bay Sandspit): Nearshore zone along the major sandspit.
5. EMBA (Embarcadero): From STAX Bistro (north) to Estero Adventures (south).
6. TIDL (Tidelands Park): Nearshore zone along Tidelands Park.
7. MBBB (Morro Bay Back Bay): From Tidelands Park parking lot (north) to Shark Inlet (south).

This project concentrates on two of these sections to ease analysis: MBCT and MBTP.

Two Landsat 8 images of the study site were found and downloaded from USGS’s  EarthExplorer , a data portal containing satellite and aerial imagery. One image was captured on February 24, 2020 (to represent winter) and the second on August 18, 2020 (to represent summer). Search criteria for these images included “0-10% cloud cover” to ensure the scenes were not obscured by clouds.

Landsat 8 images were used because they offer great coverage of Morro Bay and have a resolution of 30 m, which is fine enough to detect changes on the estuary’s surface and the surrounding land. They have previously been processed by USGS to remove distortion from solar or atmospheric effects, which eases analysis for my project. They are also freely available, saving me from having to acquire images through more expensive means like drones.

Once added to a project in ArcGIS Pro, both images were clipped using the “Clip Raster” tool to focus on the area of interest.

Sea otter surveys are conducted nearly monthly in Morro Bay by  Sea Otter Savvy , a non-profit organization based in Moss Landing, CA. Surveys are conducted either on shore using telescopes or on boats using binoculars, and take approximately 2 hours to complete. Data collected during these sessions include information on sea otters and environmental factors like tide height. Below is an example of the column headers for the CSV file produced after each survey. In total, 101 surveys have been conducted between 2015 to 2022 for a total of 1,651 sea otter observations.

The most relevant data collected during these surveys for this project are: Date, X, Y, Ind. X and Y are the longitude and latitude values of the otters observed, respectively. Ind represents the number of adult sea otters counted. These data were used to map the location of adult sea otters in the summer and winter to detect any seasonal spatial patterns in distribution.

Performing a classification on your remotely sensed data, whether supervised or unsupervised, allows you to assign land cover types for your image and run quantitative assessments. Supervised classifications tend to require more time and effort, but are generally more accurate because they have greater user control through the selection of training sites ¹⁶ . To ensure the classes are correct, I conducted supervised classifications. 

They were completed on both Landsat 8 images to discriminate against different types of water/land surface cover. A pixel-based supervised classification was performed in ArcGIS Pro using the Image Classification Wizard. Each scene was broken down into four main classes: water, sand/mud/rock, vegetation, and infrastructure. 15 polygons were carefully created for each class to properly train the Image Classification Wizard, with approximately an equal number of pixels for each class. The training method was set to “Maximum Likelihood” and the classes were finally merged to produce reclassified supervised images.

Creating the supervised classification images is necessary to detect changes between the two and see if there are water/land cover differences by season. This is important to consider because seasonal changes in estuarine habitats may be influencing which part of Morro Bay sea otters prefer to use.

ArcGIS Pro’s “Compute Change Raster” tool was used to compute and map the differences between water/land cover in the two supervised classification images. These changes in water/land cover are represented as different colors in a new raster layer.

Sea otter location data from Sea Otter Savvy’s surveys were read and plotted in ArcGIS using the “XY Table to Point” tool. Once all observations were plotted, the “Merge” tool was used to merge the observations by season into its own layer for more convenient analysis.

Heat maps of adult sea otters were then produced for summer and winter by changing the Symbol type under Symbology to “Heat Map.” In these maps, warmer areas represent a higher density of sea otters and lighter areas show a lower sea otter density. This is an appropriate map to use for easily visualizing preferences in sea otter habitat use ^{12, 17} .

Bar graphs were used to display the average number of adult sea otters present during each season at MBCT vs. MBTP. These graphs are ideal for showing comparisons between groups ¹⁸  and were created in R.

Due to the non-parametric nature of the sea otter data, the Mann-Whitney U test was used to determine if the number of sea otters using each site significantly differed between summer and winter. 

This statistical test is appropriate because it compares two independent groups (e.g., adult sea otters in the summer vs. winter) that do not follow a normal distribution ¹⁸ .

A bar graph was created and a Mann-Whitney U test was run in R to compare the number of sea otters in two habitat types in MBCT: eelgrass/kelp and adjacent open water.

Again, this is the correct statistical test to use because we are interested in seeing if there is a significant difference in sea otter counts between the two habitat types ¹⁸ . Sea otter counts are neither in paired groups nor are they normally distributed.

Upon visual inspection, supervised classification worked accurately to bin Morro Bay’s water/land cover into various classes. The dominant class in the estuary is water, followed by mud/sand/rock, vegetation, and infrastructure. Infrastructure surrounds the perimeter of the bay, while the other three classes are interspersed. 

Looking specifically at our sites of interest, we see water/land cover differences between MBCT and MBTP. MBCT has proportionally more vegetation and mud/sand/rock than MBTP, while MBTP has more infrastructure.

Comparing the two images reveals noticeable differences in aquatic vegetation cover in both MBCT and MBTP. From summer to winter, the vegetation decreased while water and mud/sand/rock increased in the two areas. It is also worth noting that the remaining class—infrastructure—had negligible seasonal change at either MBCT or MBTP. The summary table to the right breaks down the percent change in land cover classes from summer to winter 2020.

Mapping the abundance of adult sea otters at MBCT and MBTP through heat maps shows significant differences by season. In the summer, sea otter density is highest in the northwest corner of MBCT. Patches of moderate to low otter density can be found in other parts of MBCT, but there is a single hotspot of moderate density near the South T-Pier in MBTP. In the winter, the highest density of sea otters occurs in the top right corner of MBTP. Few pockets of low otter density are also present in MBCT.

Plotting the average number of adult sea otters at MBCT and MBTP confirmed the seasonal differences. At MBCT, the average number of otters present in the summer is 5 and in the winter is 2. At MBTP, the average number of otters present in the summer is 2 and in the winter is 4.

Results of the Mann-Whitney test for each site indicated that these seasonal differences in average otter numbers is statistically significant (p = 0.002 for MBCT, p < 0.0001 for MBTP). We can reject the null hypothesis that states no meaningful change in the means.

Seasonality, then, may be an important environmental factor influencing sea otter habitat use in the estuary.

.

Looking closely at the heat map for MBCT during the summer, sea otter density appears to be highest near but not necessarily in the aquatic vegetation. There is little overlap between the main hotspot of sea otters and the eelgrass/kelp bed. However, other patches of moderate sea otter density surrounding the main hotspot do intersect with vegetation. It is difficult to determine from this map alone which habitat type sea otters favor.

The box-and-whisker plot and Mann-Whitney U test provided clearer results. The boxplots revealed a distinct preference for vegetation over the adjacent water, with a median of 11 otters in the eelgrass/kelp bed and 1 otter in the water. The output for the Mann-Whitney U test verified this finding, with a p-value of 0.0001. This indicates that there is a statistically significant difference in sea otter habitat use between vegetation and water in MBCT during the summer.

MBCT has an eelgrass and kelp bed that grows along its north perimeter during the spring and summer months. Like terrestrial plants, these aquatic ones exhibit seasonality and die back when solar radiation decreases during fall and winter ^{20, 21} . This explains the shrink in aquatic vegetation seen in the raster change map.

MBTP, on the other hand, has minimal eelgrass or kelp, which was confirmed in the supervised classification images. It is dominated instead by maritime infrastructure like piers and docks. This site has two T-piers that serve as launching points for recreational and commercial boats ²² , including  Morro Bay Oyster Company . Many of these manmade structures are composed of concrete or wood and are permanent, so they are not expected to change from season to season. Thus, it is reasonable that the percent change in infrastructure between seasons is an extremely small value.

My results indicate that sea otter habitat use is seasonal, with more otters occupying MBCT during the summer and MBTP during the winter. Why might this be the case?

It is thought that sea otter habitat use and aquatic vegetation presence are linked and that the more eelgrass/kelp cover there is, the greater the sea otter abundance.

Some of the earliest studies on sea otters point out the species’ habit of resting in vegetation ^{23, 24} . Eelgrass and kelp beds are “the preferred resting habitat” for sea otters because they help anchor them in place ²⁵ . Despite being marine mammals, sea otters are poor divers and can only live in nearshore areas where the depth does not exceed 60 m ²⁶ . To avoid drifting out into the deep ocean, otters will tuck themselves into aquatic vegetation near the shore ¹¹ .

Recall that MBCT has more aquatic vegetation than MBTP, while the latter has more maritime infrastructure. MBCT’s vegetative cover likely attracts sea otters to that location during the spring and summer, when eelgrass/kelp cover is maximal. However, sea otters lose interest in MBCT when the vegetation dies back during the fall and winter. Exposed to winter storms with little protection from the shrinking eelgrass/kelp canopy ²⁷ , otters shift their habitat preference to MBTP at this time. MBTP’s manmade structures can offer otters some shelter from the cold weather and inclement storms. Sea otters are known to change habitats in the event of a storm ^{28, 29} , so this is a plausible explanation for the shift in habitat use from MBCT to the more protected MBTP.

Sea otters may be seeking additional protection during the winter for another reason: pups. Most of Morro Bay’s otters are female and pupping season reaches its peak in January ²⁷ , so there may be a mass movement of sea otter moms to the sheltered T-Pier in the winter. By summer, both mom and now independent pup no longer require the safety provided by the maritime infrastructure and venture out of MBTP.

Sea otters in Morro Bay exhibit spatial patterns in habitat use by season. They use the estuary differently depending on the time of year, with a statistically significant preference for MBCT in the summer and MBTP in the winter. We can also conclude that sea otters have a bias for habitat with vegetation over water in MBCT during the summer.

Together, these results demonstrate that sea otters are not inhabiting all parts of the bay equally. There are coarse- and fine-scale trends in sea otter habitat use detected through this project that were not able to be picked up by prior studies or by the annual USGS census. Though some explanations have been put forth about the seasonal dynamics, future research is necessary to confirm them and further build on the research presented here.  

These findings add to what little we know about estuarine habitat use by sea otters. Currently, only one other case study exists of sea otters actively living in a California estuary: Elkhorn Slough. Otters living in this water body have been researched extensively, but broadening the portfolio of case studies to include Morro Bay would give biologists a better understanding of how important estuaries are to the endangered species. In fact, knowing more about the relationship between sea otters and estuarine habitats could shape future conservation strategies and targets. 

 Sea otter reintroduction  efforts currently underway for Oregon and northern California are considering estuaries as potential release sites. This project and others like it could inform officials if estuaries make promising areas for reintroduced sea otter populations to recolonize.

This project also holds importance for estuarine habitat management. Based on the results, it is recommended that eelgrass/kelp restoration in Morro Bay be a high priority. Increasing aquatic vegetation cover would foreseeably lead to a healthier ecosystem capable of supporting sea otters and other nearshore creatures.

Additional research should be pursued to build on this project’s findings. One such study could involve detecting spatial patterns in sea otter habitat use based on other environmental factors. Tide height could influence species distribution in Morro Bay, as it has in Elkhorn Slough ¹² , and deserves a closer look. 

Another, more ambitious study could examine spatial patterns in habitat use based on sex. Male sea otters tend to travel more and may use the estuary more extensively than females, which have higher site fidelity ³⁰ . To see how these behavioral differences impact species distribution, a sample drawn from Morro Bay’s sea otters could be captured and tagged with colorful flipper markers. This would allow researchers to keep track of individuals, record their movements, and eventually map out any sex-based spatial trends.