In response to declining flycatcher numbers on Base, a conspecific playback study was initiated in 2018 to explore if this method could be used to attract breeding flycatchers to recolonize MCBCP. Conspecific attraction, the tendency for individuals of a species to settle near one another, has been used successfully as a tool for recolonizing restored Sierra Nevada meadows with Willow Flycatchers of the adastus and brewsteri subspecies (Schofield and others, 2018), but it has not been applied to any populations of the extimus subspecies. This report presents the results of Southwestern Willow Flycatcher response to conspecific playback.

In 2019, MCBCP began installing artificial seeps along the Santa Margarita River as part of a Conjunctive Use Project to enhance Willow Flycatcher habitat. The artificial seep pumps were designed to increase the amount of surface water present to enhance flycatcher breeding habitat. From 2020 through 2024, as a surrogate for absent flycatchers, we monitored Least Bell’s Vireo breeding activity associated with changes in vegetation and soil saturation at nest sites surrounding the artificial seeps (Lynn and Kus, 2024; Lynn and others, 2024a, 2024b, 2024c; B. Kus, U.S. Geological Survey, unpub. data, 2024). However, changes in nest sites between years introduced variation in vegetation and soil saturation conditions that could not be attributed directly to the artificial seeps. Therefore, starting in 2025, we established permanent sampling plots to monitor changes in vegetation and soil saturation associated with seep pumps over time. This report presents baseline analyses and results of the response of vegetation to artificial seeps starting in the sixth year of operation.

The primary purpose of this study, which began in 2000, was to document the status of Southwestern Willow Flycatchers at MCBCP in San Diego County, California. Specifically, our goals were to (1) determine the size and composition of the Willow Flycatcher population on Base; (2) document annual survival, fidelity, and movement of resident flycatchers; (3) document nesting activities; and (4) characterize habitat used by flycatchers. The second purpose of this study was to present results of management efforts to attract and retain Southwestern Willow Flycatchers at MCBCP. Specifically, those goals were to (5) evaluate the use of conspecific playback to attract breeding Southwestern Willow Flycatchers to historically occupied habitat to facilitate recolonization; and (6) document the vegetation structure, plant composition, and soil saturation within the areas affected by artificial seeps established to enhance flycatcher habitat compared to similar areas without artificial seeps.

Data collected from this study are critical to inform natural resource managers about the status of this endangered species at MCBCP and guide modification of land use and management practices as appropriate to support the species’ survival. This work was funded by and performed in cooperation with the Assistant Chief of Staff (AC/S), Environmental Security, Resources Management Division, Marine Corps Base Camp Pendleton, California. All activities were covered under 10(a)1(A) Recovery Permit #ESPER0004080-0.3 and a protocol approved by the Western Ecological Research Center Institutional Animal Care and Use Committee for compliance with the Animal Welfare Act. This report includes an annual update to surveys that have been performed since 2000 (Kus, 2001; Kus and Ferree, 2003; Kus and Kenwood, 2003, 2005, 2006a, b; Kenwood and Kus, 2007; Rourke and others, 2008; Howell and Kus, 2009a, b, 2010, 2011, 2012, 2013, 2014, 2015, 2016, 2017, 2024a, b, c, 2025b, c; and Howell and others, 2018, 2020). The study background, objectives, and methods originally were presented in Kus (2001) and are updated in this report.

From 2000 to 2018, all MCBCP’s major drainages and several minor ones that supported riparian habitat were surveyed annually for flycatchers. In 2019, a reduced monitoring plan was implemented in which annual surveys for flycatchers were done only in core survey areas, where breeding had historically been documented on Base. The remaining unoccupied riparian habitat was divided into five non-core survey groups, with each group to be surveyed on a rotational schedule every 5 years, beginning in 2020. In 2025, all of MCBCP’s historically occupied riparian habitat (core survey area; fig. 1) was surveyed for flycatchers. In 2020 and 2021, non-core survey areas A and C (respectively, fig. 1) were surveyed for flycatchers; however, because of funding limitations, none of the non-core survey groups were surveyed for flycatchers in 2022, 2023, 2024, or 2025. Protocol surveys were done three times between May 15 and July 31, 2025 (fig. 1; app. 1, figs. 1.1–1.6). Field work was completed by U.S. Geological Survey personnel: Lisa Allen, Annabelle Bernabe, Alexandra Houston, Scarlett Howell, Suellen Lynn, and Shannon Mendia. The specific areas surveyed were as follows:

Habitat was characterized by visual inspection within 50 m of each flycatcher location. Dominant native and non-native plants were recorded, and percent cover of non-native vegetation was estimated using cover categories of less than 5, 5–50, 51–95, and greater than 95 percent within the area used by each flycatcher detected. The overall habitat type at each location was specified according to the following categories:

Historic breeding territories at MCBCP were grouped into 14 plots; 7 of the plots received conspecific vocalization broadcasts designed to attract Southwestern Willow Flycatchers, and the remaining 7 served as a control group, receiving no vocalization broadcasts (fig. 2; Schofield and others, 2018). Of the seven conspecific playback plots, five were on the Santa Margarita River, one was at Lake O’Neill, and one was at Pilgrim Creek. Six of the control plots were on the Santa Margarita River, and the remaining control plot was on Pilgrim Creek. In the conspecific playback plots, an automated unit broadcasted a combination of Willow Flycatcher vocalizations (primary “fitz-bew” song and various calls, interspersed with silence) from 0100 to 0600, 0700 to 0900, and 2000 to 2100 Pacific Standard Time. Vocalizations were broadcast at a volume level mimicking the typical level of spontaneously singing Willow Flycatchers and could be heard by observers at a maximum of 80 m away from the broadcast unit. Automated broadcast units consisted of a FOXPRO NX4 wildlife caller (FOXPRO, Lewiston, Pennsylvania, USA) connected to a Favolcano CN101A digital programmable timer (Favolcano, Fujian, China) and an external power source (12-volt, 9 amp-hours, AH, battery). The automated broadcast units were operated from April 30 to August 1, 2025, spanning the time when northbound flycatchers would be searching for locations to settle and potentially encompassing southbound flycatchers that might settle in future breeding seasons. Control plots not receiving broadcast vocalizations were at least 200 m from the broadcast unit to eliminate influence from the broadcast.

Surveys were done every other week from May to July in the conspecific playback and control plots to determine if any Southwestern Willow Flycatchers had established a territory. Surveys in 2025 began on May 6 and concluded on July 23. Investigators surveyed within 50 m of all historic breeding territories that fell within the playback or control plot, following a slightly modified survey protocol developed to attract Willow Flycatchers of the adastus and brewsteri subspecies to restored Sierra Nevada meadows (Schofield and others, 2018). Upon initiation of the survey, investigators stood quietly for 3–5 min, listening for spontaneously singing or calling Willow Flycatchers. If flycatchers were not detected during the initial listening period, investigators broadcasted the Willow Flycatcher song for approximately 30 s and then looked and listened for approximately 2 min for a response. If no response was detected, investigators repeated the 30-s broadcast and 2-min listening period. If flycatchers were not detected after the second round of broadcasting/listening, the investigator moved to the next historic location within the survey plot and repeated the sequence with a 1-min, pre-broadcast listening period before beginning the 30-s playback. In plots with automated broadcast units, if the survey was done during the time the broadcast was scheduled, the units were turned off before beginning the survey.

Flycatchers observed during protocol surveys that were suspected to be resident birds (for example, observed in more than one survey period, pair vocalizations heard, evidence of nesting seen) were revisited within 3 days of the detection date. Resident birds were observed for evidence of nesting, and nests were located and monitored following standard protocol (Rourke and others, 1999). Nests were visited as infrequently as possible to minimize disturbance and reduce the chances of leading predators or cowbirds to nest sites. Typically, there were three to four visits per nest, spaced approximately 5–10 days apart, depending on the stage of the nest when initially detected. The first visit was timed to determine the number of eggs laid, the next to confirm hatching and age of young, and the last to band nestlings. After a nest became inactive, six possible nest fates were assigned based on the following parameters:

Nest site characteristics were recorded after the abandonment or fledging of nests. Measurements included nest height, host species, host height, distance from the nest to the edge of the host species, and distance from the nest to the edge of the clump of riparian vegetation (Rourke and others, 1999). Distance to edge of the clump was expressed as a negative number if the nest was not in a clump of riparian vegetation. For example, if the nest was in a field of poison hemlock (Conium maculatum) without any other riparian vegetation present, the distance to the nearest clump of riparian vegetation was measured and the value expressed as a negative number.

We attempted to capture and color band all resident flycatchers detected on MCBCP starting in 2000. Attempts were made each year to capture any unbanded adults within their territories using mist nets and band them with a numbered federal band on one leg and a solid or bi-colored metal band on the other. Any observed nestlings would be banded at 7–10 days of age with a silver, aluminum, federal-numbered band on the left leg. Returning adults previously banded as nestlings (natal) were target netted to determine their identity, and their original band was supplemented with one additional band to generate a unique color combination.

During surveys, we attempted to resight all Willow Flycatchers to determine whether they were banded, and if so, to confirm their identity by reading their unique color band combination or by recapturing birds with single federal bands. We used resighting data to determine the number of banded birds present in one year that returned the following year, and of the birds that returned, to determine site fidelity and movement. If banded flycatchers were identified, site fidelity and between-year and within-year movements were determined by measuring the distance between the center of a flycatcher’s breeding territory in 2024 (or last year detected) and the center of the same flycatcher’s breeding territory in 2025. Territory centers were estimated based on the area used by the flycatchers during surveys and nest monitoring. Adult flycatchers exhibited site fidelity if they returned to within 100 m of their last occupied territory, and natal flycatchers exhibited natal site fidelity if detected anywhere on Base.

In April 2019, MCBCP completed construction of a weir system designed to divert water from the Santa Margarita River to Lake O’Neill and several recharging ponds for the Conjunctive Use Project (P. McConnell, Vernadero Group, unpub. data, 2018). The purpose of the Conjunctive Use Project is to provide additional water for MCBCP and the Fallbrook Public Utility District (P. McConnell, Vernadero Group, unpub. data, 2018). Upstream groundwater withdrawal could affect flycatcher habitat adversely by reducing surface water, a critical component of flycatcher habitat (Sogge and others, 2010); therefore, the Biological Opinion issued by the U.S. Fish and Wildlife Service (2016) required MCBCP to initiate a habitat enhancement project to offset anticipated impacts to flycatcher habitat. We collaborated with AC/S Environmental Security to design an experiment testing the effectiveness of artificial seeps to enhance flycatcher habitat, and in January 2019, MCBCP began operating an artificial seep along the Santa Margarita River to compensate for groundwater withdrawal upstream. Two more artificial seeps were installed in early 2021 for a total of three artificial seeps. A low-volume (20–40 liters per minute), shallow groundwater irrigation pumping well was installed at each artificial seep location to draw water to the surface during daylight hours from approximately mid-April to the end of August each year, although maintenance issues prevented consistent operation on exact dates every year. The pumps were solar-powered and directed water to two–six outlet pipes at each site. The seeps were within three of the conspecific playback plots (fig. 2) along the Santa Margarita River. During conspecific playback surveys, we observed the habitat immediately surrounding the seeps and recorded if any Willow Flycatchers were using the area.

We focused on two of the three artificial seeps to examine the effect of the seep pumps on flycatcher habitat (fig. 3). Two outlet pipes at the Pump Road site and six outlet pipes at the Old Treatment Ponds site directed water to an area of approximately 1,500 square meters (m²). To adjust for even distribution of surface water at the Old Treatment Ponds site and to compensate for low water pressure, two outlet valves were closed for the full operating season in 2020 and 2024, one outlet valve was closed for the full operating season in 2021 and 2022, and one outlet valve was closed for the last month of the operating season in 2023. Water was pumped to the surface when there was sufficient sunlight for solar panels to generate power beginning in April and ending in August each year. Shallow pools created by the seep pumps were small (8–44 m²) and limited to the immediate vicinity of the outlet pipes. The pump at the Pump Road North Seep site was non-functional for the 2023 field season and the pump at the Old Treatment Ponds North Seep site was non-functional for the 2025 field season.

We established two types of study plots: Seep and Reference sites. In 2023, we temporarily reclassified one site as Intermediate to incorporate the potential consequences of an unanticipated disruption to the operation of the seep pump (fig. 3); however, we abandoned this designation when the second seep pump failed to operate in 2025, reestablishing parity between the Seep sites. Originally, two Seep sites were selected. The Seep sites surrounded and extended downstream from (1) the seep installed in 2019 northwest of the Old Treatment Ponds area and (2) the seep installed in 2021 in the Pump Road area. A Reference site was selected 0.5–0.8 km from each Seep site. Reference sites were on the same side of the Santa Margarita River as their corresponding Seep sites and encompassed similar vegetation as the corresponding Seep site. Although seep pumps at both Seep sites experienced unanticipated disruptions causing inconsistency in annual operation, we anticipated that both Seep sites would become wetter relative to the Reference sites as a result of the effects of the upstream water diversion and annual surface water augmentation near the seep pumps.

Transient flycatchers were detected in 14 percent (1/7) of the conspecific playback plots and 14 percent (1/7) of control plots with no playback (app. 3, fig. 3.1). Two transient flycatchers were detected within 20 m of a conspecific broadcast unit. Transient flycatchers were not observed near any operational artificial seep outputs.

Transient flycatcher detections occurred in habitat classified as riparian scrub and mixed willow riparian (table 2). Most flycatcher locations were dominated by one of three willow species: arroyo willow, red willow, or sandbar willow. Exotic vegetation was present in most flycatcher locations in 2025. Twenty-five percent (2/8) of flycatcher locations had habitat composed of 5–50 percent exotic vegetation, and 63 percent of locations (5/8) were dominated by exotic vegetation (percent cover of exotics greater than 50 percent). The most common exotic plant in the habitat used by flycatchers was poison hemlock, which was present in 75 percent of locations (6/8; table 2).

During surveys, observers incidentally observed large stands of dead and dying trees within many of the historic flycatcher breeding areas because of ongoing invasive shothole borer (Euwallacea spp.) infestation (Nobua-Behrmann and others, 2023). The tops of many willow trees were snapped off, which reduced overall canopy height and density. In addition to trees that were already dead, others showed evidence of active infestation (for example, frass and wet staining at entrance holes; Nobua-Behrmann and others, 2023).

No Southwestern Willow Flycatcher nesting was observed in 2025.

Most transient flycatchers were observed closely enough to confidently determine if they were banded (6/8). None of the closely observed flycatchers were banded.

In 2025, we identified 43 common plants at vegetation plots throughout all sites. The most common plant species was the woody red or arroyo willow, which was one of the top 3 species at 37 of the 48 sampling locations (77 percent). Other common woody plant species were Goodding’s black willow (26 locations; 54 percent), poison oak (Toxicodendron diversilobum) (24 locations; 50 percent), mule fat (23 locations; 48 percent), and blue elderberry (Sambucus nigra; 20 locations; 42 percent). The second most common plant species and the most common herbaceous species was stinging nettle (32 locations; 67 percent), followed by poison hemlock (29 locations; 60 percent). We determined that 13 plant species were common at 20 percent of locations or more. Of these 13 plant species, 3 were non-native: poison hemlock, black mustard (Brassica nigra), and perennial pepperweed (Lepidium latifolium).

Overall vegetation cover in 2025 was highest near the ground and decreased with increasing height (fig. 6). Woody vegetation made up most of the cover at all height categories, with cover ranging from 35 percent below 3 m to 8 percent above 6 m. Herbaceous cover (total and native) was concentrated near the ground with very little (less than 2 percent cover) above 3 m. Non-native vegetation was present at all height categories, tapering from 19 percent below 3 m to 1 percent or less above 6 m.

Soil saturation varied among sites, with highly saturated plots more common at the Seep sites than at the Reference sites in 2025 (figs. 7, 8). Most of the plots at the Old Treatment Ponds Seep site (85 percent) and Pump Road Seep site (100 percent) had greater than 50-percent soil saturation (fig. 7), whereas only 20 percent of plots at the Old Treatment Ponds Reference site and 36 percent of plots at the Pump Road Reference site had greater than 50-percent soil saturation (fig. 8). Soil was 100-percent saturated at more than 60 percent of plots at the Seep sites and less than 20 percent of plots at the Reference sites (figs. 7, 8). Plots were wet throughout the Pump Road Seep site. At the Old Treatment Ponds Seep site, wet plots were prevalent in the center of the site, with drier plots at the extreme northern end and an area southeast of the seep outlets. The plots in the Pump Road Reference site were mostly dry, with a few wet areas at the north end and at the southwest end of the site where water collects on a dirt road in wet years. Similarly, the Old Treatment Ponds Reference site was mostly dry, with a few wet plots at the eastern edge near a seasonal pond. Soil saturation was significantly higher at the Seep sites (90±15 percent) than at Reference sites (30±32 percent; t=8.29, P < 0.001). We did not find a linear association between soil saturation and distance from the nearest seep outlet at the Seep sites (ρ=0.20, P=0.38).

In 2025, we determined that the relationships of total cover and woody cover with soil saturation were not statistically significant; however, non-native vegetation above 3 m decreased with increasing soil saturation, and native herbaceous cover above 3 m, albeit low, increased with increasing soil saturation (table 3). Statistically, total herbaceous cover above 6 m also increased significantly with increasing soil saturation (ρ=0.27, P=0.07), although the amount of herbaceous vegetation at this height was small, and the relationship was not biologically significant. We did not find a significant correlation between soil saturation and average canopy height (ρ=−0.09, P=0.53).

In 2025, we did not find significant differences between Seep and Reference sites in total, woody, total herbaceous, or native herbaceous cover (fig. 6). However, non-native cover was higher at the Reference sites than at the Seep sites at all height categories, confirming the negative relationship described above between soil saturation and non-native vegetation. The average canopy height did not differ significantly between Seep and Reference sites (7.4±2.9 m, 8.1±2.9 m, respectively; t=0.86, degrees of freedom [df]=45.98, P=0.19). We also did not find a significant difference in maximum canopy height between Seep and Reference sites (9.5±3.7 m, 11.0±3.9 m, respectively; t=1.35, df=45.85, P=0.18).

In 2025, the average percent cover was low for all plant species that were considered important or noteworthy in the Southwestern Willow Flycatcher habitat model (Howell and others, 2018) and we did not find significant differences between Seep and Reference sites for any species at any height category (table 4). However, poison hemlock cover below 3 m significantly decreased with increasing soil saturation (table 5). We did not find significant correlations between stinging nettle cover and soil saturation or cover of Goodding’s black willow and soil saturation.