Restoration of Bedrock Vernal Pools for the Threatened Vernal Pool Fairy Shrimp in Central California

Vernal pool creation and restoration has been an active industry in California for more than 40 years (Black and Zedler 1998, DeWeese 1998). The success of these creation and restoration activities has been less than consistent, with results that may not meet the goals of the restoration activity (Calhoun et al. 2014, Schlatter et al. 2016). For example, creation and restoration projects must mimic surrounding conditions, and more importantly, meet the natural history parameters of the target species for which the activity was designed (Eng et al. 1990, Vanschoenwinkel et al. 2009, Wortley et al. 2013).

Fairy shrimp, which often or exclusively occur in vernal pools, are a taxon for which many species are in steep decline, particularly in California (Eriksen and Belk 1999). Successful restoration of vernal pools is dependent on the target of the restoration (Bakker et al. 2000). We engaged in a restoration project that included the removal of sediment, topsoil and associated vegetation from historic vernal pools that lay within rock outcrops in a vernal pool complex. The goal of our restoration activity was to increase the number of pools and the volume of wetted vernal pool habitat for potential colonization by the federally threatened Branchinecta lynchi (vernal pool fairy shrimp). We report herein on the methods and results of those activities.

We surveyed 29 vernal pools in the Kellogg Creek Vernal Pool Complex from 1998 to 2016 to determine the annual presence of B. lynchi and other sympatric species. The site was comprised of sandstone bluffs in the upper Kellogg Creek watershed, Contra Costa County, California. These bluffs included various depressions in the rock that were likely formed by weathering (Way 1978) and ranged from 0.1–1.1 m in depth and 0.5–195 m² in surface area. Many pools were completely open and devoid of vegetation; however, approximately 50% of the pools included some level of soil intrusion that supported non-native annual grasses (Figure 1).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1. Initial soils removal in a large bedrock vernal pool (#NP-10), using hand shovel, Contra Costa County, CA, December 2015. Image credit Jeff Alvarez.

Following winter rains and pool inundation—typically late December to early January—surveys at these 29 pools were conducted bimonthly. Pools were considered inundated when they were > 3 cm deep (maximum depth; following USFWS 2017). Each survey included a single biologist conducting visual encounter surveys (i.e., walking or viewing the water column for two minutes per 2 m² of estimated surface area, while an additional biologist took two sweeps with a fine mesh net per 1 m² of approximate surface area. Surveys were conducted twice monthly until fairy shrimp were detected or pools dried. Additionally, physical characteristics of each pool (e.g., depth, surface area, temperature, etc.) were also collected (Table 1). Fairy shrimp samples collected were assessed based on visual estimates of animals captured and resulted in a determination of 1s (≤ 10 animals present); 10s (10–100 animals); 100s (100–1000); or 1000s (≥ 1000 animals present) of shrimp present per pool. Due to a history of presence of Linderiella occidentalis (California fairy shrimp), and adjacent populations of B. longianntenae (long-horned fairy shrimp) and B. mackini (alkali fairy shrimp), voucher specimens were collected each year to confirm the identity of species.

In November 2015 we visually assessed the vernal pool complex and determined that at least 12 sites appeared to have historic vernal pools that were entirely filled by soil and non-native annual grasses (Figure 1). We elected to restore these vernal pools by removing excess soil and allowing the sites to naturally hydrate from seasonal rains. These new pools would then be added to the annual monitoring efforts of the existing 29-pool complex.

Beginning 22 Dec 2015 and continuing until 15 Jan 2016, we hand-excavated each historic (i.e., potential) vernal pool until all soils were removed (Figures 1, 2 and 3). We removed soil and vegetation by hand shoveling and placed it in a wheelbarrow, which was then moved to an off-site location where it was deposited away from vernal pools (i.e., no chance of refilling the vernal pool). A fine layer (approximately 0.5–1 cm) of soil was left on the bottom of each pool as a result of using only shovels to remove soil. No pools were inoculated with soils from other pools, and all pools were left to inundate naturally from seasonal rain. Following restoration and inundation, each pool was monitored for vernal pool invertebrates.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2. Continued hand removal of soils for a large vernal bedrock vernal pool (#NP-10), Contra Costa County, CA, December 2015. Image credit Jeff Alvarez.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3. Six weeks following complete removal of all soils, this (#NP-10) and all other restored vernal pools were inundated by rainfall, and fairy shrimp were noted swimming within the pool, Contra Costa County, CA, January 2016. Image credit Jeff Alvarez.

Surveys conducted between 2013–2020 indicated that among the 29 original pools, there was a high level of interannual variability in the percentage of rock pools inundated (range: 39–95%). Among those same pools, and during the same time period, the percentage of pools that had B. lynchi present ranged from 13–69%, also indicating a high level of interannual variability in fairy shrimp presence.

Our work included restoration of 12 additional pools in 2015, accounting for an increase of approximately 72 m² of rock pool habitat. Newly excavated pools ranged from 1–30 m² and ranged in depth from 0.1–0.95 m. Following restoration and seasonal rain events, every pool excavated showed signs of inundation (25.4 cm [10 in.] the first year following restoration, with a single pool becoming inundated two years following restoration, and a single pool becoming inundated only after four years following restoration activities). Among the 12 pools that were restored, nine were observed having B. lynchi (i.e., 1s to 100s of individuals). Two years following the restoration of pools, an additional pool included B. lynchi. A single 2 m² restored rock pool (RP-09) was reported to have L. occidentalis in 2018, a species not seen in this vernal pool complex for the previous 20 years (Contra Costa Water District, unpublished data). In addition to fairy shrimp, the natural and restored pools included a variety of vernal pool invertebrates (e.g., representatives of orders Ostracada, Copepoda, Cladocera, Hemiptera, Coleoptera and Tricladida) as well as Pseudacris regilla (Pacific chorus frog) (Table 1).

We found that careful investigation can reveal cryptic, historic vernal pool depressions that may be obscured by sediments eroded from upland soils surrounding the pool. In our work, some of these potential vernal pools were overgrown with non-native annual grasses and required substantial excavation (Figures 1, 2, and 3.). Unearthing potential vernal pools in this complex significantly increased vernal pool habitat: the restoration activities increased the surface area of vernal pools in this complex by 34% and a maximum number of pools occupied by fairy shrimp by 11 (55% increase). We speculated that many of the pools were colonized by wind-blown cysts, which is common for B. lynchi (Ericksen and Belk 1999), or by surface runoff between pools.

Kistner et al. (1995) suggested that documenting a population’s viability was preferred over species presence as a method to determine successful creation or restoration. At this vernal pool complex, we noted a strong tendency toward interannual variability in the number of pools inundated and the number of pools supporting fairy shrimp (Eriksen and Belk 1999). This population has persisted since prior to the initiation of systematic annual surveys in 1998. Additionally, B. lynchi has been present annually within up to nine of the restored pools, and 10 of 12 total restored pools since their restoration.

We defined success for the proposed project as 1) an increase in inundated area for rock outcrop vernal pools within the complex, and 2) colonization of one or more pools by B. lynchi. Establishing target success criteria can guide or direct the methods and monitoring required to meet those targets (Bakker et al. 2000). Therefore, our project was successful.

In our case, although the amount of human labor to move an estimated 35 m³ of soil was high (approximately 66 person-hours), follow-up surveys require only a single permitted biologist to check pools at intervals that will increase the likelihood of detection (generally bimonthly, following inundation). We considered the time expended (1.1 hrs per m² of surface added) minimal when compared to the number of colonized restored pools (N = 10), and continued persistence of the species. The total cost of restoration of 12 pools was approximately $7,200 (in 2015) for both labor and tools.

We made no plan to inoculate pools with soils from occupied sites. Rather, we elected to allow for natural cyst dispersal through wind, intra-complex pool spillover, or from passive recruitment from within the existing soils that remained. We speculated that some cysts were likely banked in much of the upland areas surrounding the vernal pools. Furthermore, disturbing soils during excavation would likely leave some number of cysts within the restored pool depression following excavation. In this effort, 10 pools were colonized by one or more species of fairy shrimp, nine pools within the first wet season. Based on these results, we suggest that inoculation from pools that are within a complex is an unnecessary disturbance to existing sites.

Although the success rate related to the creation of vernal pools has been reported as variable, and even controversial (Sutter and Francisco 1996, Calhoun et al. 2014, Schlatter et al. 2016), we feel that restoration of rock outcrop pool habitat, particularly within a complex of similar habitat, may be valuable. We do acknowledge that monitoring and vigilant management (e.g., continued removal of soil or installation of erosion control measures) may be required to sustain these pools.

We also note that at this site, 14 of the original 29 pools that were monitored had soil encroachment that occupied as much as 50% of the pool area. We suggest a plan be developed for soil excavation in these pools. Based on our successful restoration efforts reported here, soil removal may increase pool hydroperiod and improve conditions for some longer-lived species (Brendonck 1996, Eriksen and Belk 1999, Philippi et al. 2001). Maintaining the greatest pool volume and the largest number of pools in a vernal pool complex will likely enhance habitat over time and should be considered when management decisions are made for maintaining a site similar to ours (Brooks and Hayashi 2002, Dittes et al. 2007).