Skip to main content

Sentenac Cienega wildlife and vegetation data

Cite this dataset

Wang, Yanghongfan; Cartagena, Cherish; Yuan, Rachel; Masjuan, Victoria (2021). Sentenac Cienega wildlife and vegetation data [Dataset]. Dryad.


Sentenac Cienega is a degraded wetland ecosystem in Southern California’s Anza-Borrego Desert State Park. Recently, researchers have taken an interest in the conservation and restoration of this Cienega, and it is in accordance with this goal that we performed our work. This project had four major components: a wildlife monitoring protocol to determine species richness and relative abundance, a greenhouse drought/salinity resilience trial, in-field plant transplantation trials, and a survey of vegetation composition throughout the entire area. The wildlife monitoring project results indicate that wildlife activity is highest in areas with easily accessible food, water, and routes for transportation. The greenhouse trial indicated that soil moisture levels had the greatest effect on the growth of Juncus mexicanus, while salinity levels had the greatest effect on the relative growth of Anemopsis californica. Mortality was very high in our field transplantation trial, with all plant species exceeding 90% mortality across all planting plots. This mortality was likely due to a combination of herbivory and drought effects. Vegetation composition mapping results show that the species richness of native species is higher than that of non-native species but two-thirds of the area is dominated by non-native species. There is a strong positive correlation between native and non-native richness and a strong negative correlation between native and non-native cover.


Eight total camera traps (Browning DarkOps Elite HD Model BTC-6HDE) were placed throughout Sentenac Cienega from November 2020 to April 2021 (Figure 4). Various areas to scout were chosen based on a map of known GPS tracking data locations of collared lions from a prior unrelated, unpublished study (Figure 5). This was done to heighten the chance of mountain lion image captures, as well as create a baseline for scout-worthy locations considering that wildlife had been seen in these areas at one time. A map of potential locations for cameras was created using ArcGIS Online. Scouting was done in November 2020 over a two-day period and consisted of finding signs of animal tracks, trails, beds, herbivory, scat, and carcasses. Areas with several signs of wildlife activity, as well as locations that appeared attractive to wildlife within varying microhabitats and vegetation zones, were ultimately chosen for camera installation. This was to ensure we would capture as many different species as possible to determine species richness, as well as areas with high animal activity frequency. 

We established eight cameras throughout the cienega: SC1 and SC2 in Riparian Wash- East, SC3 and SC4 in Willow Woodland, SC5 in Marsh to Upland Transition Zone, SC6 in Bridge Wash- West, SC7 in Bridge Wash- East, and SC8 in Remnant Mixed Riparian Zone (Figure 4). Due to possible extreme weather conditions, SC1, SC7, and SC8 were relocated in January 2021 near their original locations, but with different angles, slopes, and heights. Therefore, they were then renamed and treated as separate cameras. SC1, SC7, and SC8 captured images from November 2020 to January 2021. SC1 became SC9, SC7 became SC10, and SC8 became SC11. These three cameras captured images from January 24, 2021 to April 7, 2021. 

During camera installation, we attached cameras to either a metal post, tree, or fence that appeared stable enough to house a camera. All cameras were run continuously from the time they were installed until we stopped collection data or a camera was moved, which ranged from 68 days (SC7, SC8), 69 days (SC1), 74 days (SC 9, SC10, SC11), and 141 days (SC2, SC3, SC4, SC5, SC6). Cameras were attached between heights of 4 cm to 80 cm (depending on location, slope, and angle) and set to take three consecutive shots with 60-second delays in between camera triggers. 

During photo processing, we extracted data consisting of time and date, species, number of animals and species per photo, directionality, and camera location. These data we entered on several Google Sheets, which we compiled in R for further analysis. 

Within R, the total number of individuals for each species in a 30-minute were tallied for each camera location (SC1 to SC11) across all dates. From the master file compiled in R and excluding birds, humans, and domestic dogs, we determined the species richness at each camera location and across the entire cienega, the relative abundance index (RAI) of each species (measured as the total number of unique species divided by the duration the camera collected data) at each camera, and the RAI of overall activity at each camera (measured as the total number of days the camera captured activity over total active camera days). 

The subjects of our greenhouse drought and salinity tolerance experiment were two plant species native to Sentenac Cienega, Juncus mexicanus and Anemopsis californica. Three levels of water and salinity treatments were applied to the plants, combined for a total of nine different treatments for both species. 

After harvesting similarly-sized rhizomes of each species in the field, we transported them to the UCI greenhouse and potted them in one-gallon nursery containers filled with custom potting mix. This potting mix (2 parts pumice: 2 parts redwood compost: 2 parts peat moss: approximately 0.5 parts sand) has previously been used by UCI researchers to grow coastal sage scrub plants with good results. 

The water treatments had three levels: low, medium, and high. These treatments were designed to simulate field conditions under varying levels of precipitation. The low water treatment was designed to maintain potting mix moisture in line with the lowest soil moisture in the field (5-10%), the medium treatment was designed to maintain potting mix moisture near the average soil moisture in the field (20-25%), and the high water treatment was designed to maintain potting mix moisture in line with the highest soil moisture at Sentenac Cienega (45-50%).

The salinity treatments also had three levels: control, low, and high. The control treatment was intended as a base comparison for the other treatments. The low treatment was determined by averaging the conductivity levels found at the cienega, and the high treatment was determined by increasing the highest salinity level found at the cienega by one-third. Treatments were applied by initially irrigating each potted plant with one liter of treated DI water. The control salinity treatment was applied by irrigating plants with untreated DI water. The low salinity treatment was applied by irrigating plants with DI water salinized with table salt (NaCL) to a conductivity of 3-4 millisiemens/cm. The high salinity treatment was applied by irrigating plants with DI water salinized to a conductivity of 11.5-12.5 millisiemens/cm. All saline solutions were measured with an Oakton PCD650 calibrated to a solution of 10 millisiemens/cm. 

Prior to starting the greenhouse experiment, we conducted a soil water capacity experiment to determine the watering schedule. First, we took the weight of three empty pots, then added potting medium to the pots and took the weight of the potting medium plus the pots. Next, we watered the pots to excess and waited 30 min for the pots to drain until the three pots reached constant weight. After the pot weights stabilized, we took the total weight of each pot (potting mix + water + pot). We continued to take the total weight of each pot daily, recording the downward trend of the soil moisture, until the soil moisture reached the lowest level in the field (under 10%). Based on the results of this soil water capacity experiment, we determined that the high water treatment would be irrigated with 120 ml of water every four days beginning eight days from the start of the experiment; the medium water treatment would be irrigated with 120 ml of water every four days beginning 18 days from the start of the experiment; and the low water treatment would be irrigated with 120 ml of water every four days beginning 28 days after the start of the experiment. This irrigation schedule would keep the soil moisture levels constant and within the correct treatment range over the course of the experiment. 

After being allowed to acclimate to the greenhouse environment for a month, treatments were started at the end of January. Data was collected weekly until the experiment ended mid-April. Plant biomass was measured differently per species as appropriate. A. californica growth was estimated by measuring the number, length and width of green leaves, as well as noting any stolon growth/offsets produced by the plant. J. mexicanus growth was estimated by counting stems and measuring their height. 

To determine how salinity had changed over the course of the experiment, we analyzed the electrical conductivity (EC) of the potting medium using a pour-through technique (Nemali, 2018; Appendix C). This technique measures the EC of a leachate solution as a proxy for how much soluble salt is present in the growing medium. To perform the pour-through method, we grouped J. mexicanus plants by treatment, then irrigated thoroughly using tap water. After an hour, we poured 120 ml of DI water into each pot, collecting the leachate from each treatment group into one bucket. We then measured the EC of that leachate to obtain our result.

The field transplantation experiment was modeled after a similar experiment performed by UCI researchers in 2018, and made use of the same plots established for that experiment. Ten plots were laid out across the cienega, situated to capture the environmental gradient present onsite (Figure 6). We chose three native species, Anemopsis californica, Juncus mexicanus and Distichlis spicata as our target species. On the morning of transplantation day, we harvested rhizomes of each species from previously identified donor sites. In each plot, we planted four individuals per species, totalling 12 individuals per plot. In areas with hard, dry soils, soil was loosened to maximize root contact with soil and prevent “root boundedness.” We flagged all transplants and watered them well, with plots 1-4 receiving four gallons each of supplemental irrigation and plots 6-10 receiving two gallons each of supplemental irrigation. Rough terrain and a tight schedule prohibited us from irrigating all 10 plots with four gallons.  

After transplantation, we visited the site twice for data collection. We noted whether individuals survived, the number of any green leaves, and the maximum height and width of each surviving individual. 

We conducted our vegetation mapping using spatial information captured by Survey123 (2021 Esri 3.12.277) on cell phones. We walked across the site and when there was a change in the dominant species, we drew a polygon on the map. Within each polygon, we collected all the present species and their estimated percent cover. The name of each polygon was the dominant species code (a combination of the first three characters of the scientific name, e.g. the code for Anemopsis californica is ANECAL), number of the dominant species occurrences, and the date (e.g. BASHYS_1_022421). For the percent cover, we set five levels in increments of 20, from 1 to 100 (e.g. 0-20, 20-40, etc.)  We estimated the percent cover by the area of each polygon and the species distribution within the polygon. All other data was also recorded on standard paper datasheets. Data from the field survey was uploaded to ArcGIS Online at the end of each data collection day, checked for quality and consistency, and prepared for use in the mapping process. 

Mapping was done using ArcGIS (Esri ArcMap 10.7.1). Since the data were collected over multiple days, there are several different layers of the vegetation polygons. We downloaded the layers from ArcGIS Online and opened them in ArcMap. We used the merge tool to merge different layers into one layer. Then we used the reshape feature tool and edit vertices tool to make the boundaries of the polygons coincide. We used graduated symbols to show the native and non-native species richness. To show the distribution of native and non-native species, we used graduated colors to represent the species percent cover trend across the site. We also used different colors to show specific dominant native and non-native species of different alliances. Finally, we finished the maps by adding the appropriate legend, North arrow, scale bar, and title.

To further explore the relationship between native and non-native species richness and coverage, we used R (R Core Team, 2019) to analyze if there was any correlation between native and non-native species richness and percent cover. A Shapiro-Wilk test was done to see if the variables were normally distributed. We ran the Pearson's test because the four variables were not normally distributed. Finally, we visualized the data by creating a scatter plot and a corresponding trendline to show the correlation between native and non-native species richness and percent cover.

Usage notes

ReadMe file is uploaded as a .txt file named README_Capstone