Soil Ecology and Research Group last update December 17, 2001

Murphy Canyon Riparian Restoration
on Marine Corps Air Station Miramar,
San Diego, California
Draft Second Annual Report
1 July 2000-30 June 2001

Executive summary
Construction projects associated with the realignment of NAS Miramar filled a total of 0.295 acres of wetlands. As mitigation for this impact, the Department of the Navy signed a Cooperative Letter of Agreement with the Soil Ecology and Restoration Group (SERG) at United States International University (now Alliant International University) in June 1999 to restore a 1.03-acre area of riparian habitat on the eastern section of Marine Corps Air Station Miramar.

During the first year, a channel was graded and the site was planted in spring 2000 with 221 seedlings and cuttings of riparian and coastal sage scrub (CSS) species using propagation material collected on the Station. During the 2000 and 2001 winter storms, water flowed from the main creek bed into the channel created on the site. Seedling survival counted in August 2000 was only 38 percent. Therefore, 137 replacement seedlings were planted in January and April 2001. In order to ensure the survival of these new seedlings, they were watered much more frequently (every two weeks) over the summer 2001 than the original transplants had been. Percent vegetation cover and native shrub density was measured in May 2000 and 2001. Percent cover fell below the required success criteria both years, but it more than doubled from the first to the second year. Low cover is most likely due to the high weed cover (24.5%) measured this year. Weed control in 2000 was hampered by the delay in receiving proper authorization to spray on the Station. This year, herbicide spraying was allowed, but, due to scheduling conflicts with the licensed applicator, only two site visits were possible. Regardless, even limited herbicide application should reduce weed cover next year and more frequent applications during the spring 2002 should reduce it even further. Reduced weed cover should lead to increased seedling survival and percent cover of native species. Native species density was higher than that required in both 2000 and 2001.

Soil analyses conducted this year show that the restoration site soils are very similar to undisturbed site soils with the exception of nitrate levels. Elevated nitrate levels are likely due to the dominance of exotic annuals on site. These annuals are broken down during each dry season by soil microorganisms that deposit nitrates into the soil. By significantly reducing the number of exotic annuals through frequent herbicide applications, nitrates should return to levels more similar to those at the undisturbed site.

The site was surveyed for birds, small mammals, and reptiles in August 1999 and May 2001. Results of the bird surveys are difficult to interpret since 2001 marked the first survey conducted during the breeding season. Reptile and amphibian surveys showed slight improvement while mammal surveys showed no notable change. It is still too early to draw any major conclusions about the health of these populations at the restoration site.

Soil analyses and vegetation and faunal surveys will continue through June 2004. Seedlings will be irrigated twice per month through the summer 2001 and once per month beginning in the summer 2002. Seedling survival will be counted again in January 2002. If survival is below the required 80% for that year, replacement seedlings will be planted in the spring 2002 and will be watered twice per month during the dry season of that year.

Introduction
Marine Corps Air Station Miramar (MCAS), hereinafter referred to as the "Station", is located on 23,116 acres of land in San Diego County, California four miles east of the Pacific Ocean and thirteen miles north of downtown San Diego. Construction projects associated with the realignment of NAS Miramar filled 0.141 acres of vernal marsh, 0.009 acres of coastal and valley freshwater marsh and 0.066 acres of disturbed riparian forest for a total of 0.295 acres of wetlands. The Department of the Navy, in accordance with the Department of the Army permit (95-20158-ES) on the Realignment of Naval Air Station Miramar to Marine Corps Air Station Miramar, is required to mitigate impacts to non-vernal pool wetlands at a 2:1 replacement ratio by restoring or creating riparian habitat. The Department of the Navy signed a Cooperative Letter of Agreement with the Soil Ecology and Restoration Group (SERG) at United States International University (now Alliant International University) in June 1999 to restore a 1.03-acre area of riparian habitat on the eastern section of the Station.

During the first year of the project, a channel was graded and the site was planted with both seedlings and cuttings of riparian and CSS species using propagation material collected on the Station. The site was surveyed for birds, small mammals, and reptiles in August 1999 to use as a baseline for progress of the restoration site. A total of 221 riparian and CSS seedlings were planted on the site in Spring 2000. Vegetation cover was surveyed in June 2000. Percent cover and seedling survival on the site was lower than expected; however, native plant density was higher. Weeds such as mustard (Brassica nigra) and exotic grasses were cut down in June as spraying with herbicide was delayed due to permitting problems. During the winter storms, water flowed from the main creek bed into the channel created on the site.

In this, the second year of this project, 137 replacement seedlings were planted to make up for the low survival measured in August 2000. Maintenance and monitoring activities continued (Appendices 1 and 2) and included watering, weed control, soil analyses, percent cover and density of native plant species, and surveys of birds, mammals, reptiles, and amphibians. These activities will continue through June 2004.

Site Description
The 1.03-acre, flat site is located in Murphy Canyon along the eastern edge of a riverbed (Figure 1). The site is bordered to the east by a dirt road and to the north and south by riparian habitat. Vegetation on the site in August 1999 was primarily exotic species such as slender wild oat (Avena barbata), tocalote (Centaurea melitensis), filaree (Erodium sp.), and fennel (Foeniculum vulgare). Native vegetation on the site and surrounding area in August 1999 consisted of arroyo willow (Salix lasiolepis), poison oak (Toxicodendron diversilobum), fuchsia-flowered gooseberry (Ribes speciosum), laurel sumac (Malosma laurina), flat top buckwheat (Eriogonum fasciculatum), and mule fat (Baccharis salicifolia). The east and west edges of the canyon are sloped approximately 45 degrees and covered with CSS and chaparral. Access to the site is via a dirt road from the Camp Elliot area of the Station. Possible cause of disturbance, found by reviewing historical aerial photographs, USGS maps, and the history of the area, is most likely Camp Elliot activity, although creation of the dirt road may have added to the impact. The undisturbed reference site for this project is located approximately 2.7 kilometers northeast of the restoration site in Camp Elliot. The reference site is riparian woodland, bordered to the south by "H" Avenue. Vegetation on the reference site is dominated by California sycamore (Platanus racemosa) and Arroyo willow (Salix laevigata).

Figure 1. Location of riparian restoration site on Marine Corps Air Station Miramar.

Methods
Replacement planting
Results of seedling survival monitoring conducted in August 2000 indicated the need to plant additional seedlings. Replacement transplants were installed on two dates, 29 January 2001 and 9 April 2001. Table 1 lists the number of seedlings planted by species.

Table 1
Replacement seedlings planted in January and April 2001.

Figure 2. Willow cuttings taken in November 2000.

Maintenance
Hand watering of seedlings planted in spring 2000 began in May 2000 and continued two times per month through September 2000 via a 180-gallon water tank in the bed of a pickup truck. Each seedling received approximately ¾ gallon per watering. Significant winter rain events began in October 2000, so transplants were not watered again until June 2001 when they were watered one time. In January and April 2001, additional seedlings were transplanted to the site. Since they were new, these seedlings were watered periodically throughout the spring when rains were not frequent enough to keep the soil moist. These "new" seedlings were watered once in February, once in April, once in May, and twice in June.

Weeds on site consisted primarily of Brassica nigra, Avena fatua, Erodium spp., and Centaurium melitensis. In March 2001, weeds on almost the entire site were sprayed with Round Up Pro herbicide (40 ounces of glyphosate salt) by a licensed applicator (Figure 3). The remainder of the site and any weeds missed during the first application were sprayed in April 2001 (19 ounces of glyphosate salt). Due to scheduling conflicts, the herbicide applicator was not able to return to the site this spring, so weeds were cut with a gas-powered weed trimmer twice in April and once in May 2001.

Figure 3. Application of Round Up Pro herbicide to exotic species in March 2001.

Treepees and Tubex were removed from seedlings that had outgrown the protection devices. Carsonite boundary signs were posted along the edge of the access road adjacent to the northern section of the site where a vehicle drove off-road, creating tire ruts and killing numerous seedlings (Figure 4).

Figure 4. Damage to the restoration site in March 2001.

Monitoring
Vegetation
Ground water on the site was monitored by measuring water levels in two wells. The well on the eastern half of the site is placed among primarily CSS seedlings and is located between the road and the constructed channel. The well on the western half of the site is surrounded by primarily riparian vegetation and is located between the constructed channel and the existing channel. Wells were measured once in once in January, March, and May 2001.

Percent survival of transplanted seedlings was measured in late August 2000. In addition, vegetation percent cover and density were monitored on 24 May 2001 using a modified California Native Plant Society (CNPS) method (Sawyer and Keeler-Wolfe, 1995). Percent cover was estimated for perennial shrubs on the site using the same four 25-meter transects that were used last year, with measurements taken every 0.5 meters for a total of 50 points per transect. Three of the transects run perpendicularly to the riverbed and the road and one is parallel to the road in the CSS zone. Plant density for each species of perennial shrubs was measured by counting numbers of plants within 2.5 m on each side of the transect for a total area of 125 m2 for all three transects in the riparian area. The transect in the CSS area parallel to the road was narrowed to 1 m on each side (50 m2 total area) due to the close proximity of the dirt road. Project progress was photographed using 35 mm film from selected photo points.

Six soil samples, 10-12 centimeter (cm) deep, were collected on both the site and in the undisturbed riparian area on the Station in May 2001. Samples were analyzed for percent organic matter, available phosphorus, nitrate, total nitrogen, and pH by A & L Western Agricultural Laboratories.

Birds, reptiles, and small mammals were monitored during May and June 2001.

Bird Survey
Six bird surveys were conducted, three between the hours of 0520 and 0900 and three between 1645 and 2110 on 26 May and 1, 3, and 4 Jun 2001. The range of times for the surveys is due to a difference in start time. Point counts lasted 2.04 hours on average. Birds were censused during 5-minute point counts at 18 points within and surrounding the restoration site. Each point count was conducted from the center of a 45 meter squared plot. Four of these plots covered the area within the 180-meter long restoration site. The additional fourteen plots surrounded the restoration site (Figure 5).

Figure 5. The 18-square survey grid overlain on the restoration site.

· Diurnal surveys
Birds were identified by sight and sound during the 5-minute point counts and their approximate location marked on the grid of census plots overlaid on the restoration site and its surroundings (Figure 5). For each survey, counts began at a different square and the direction of movement between points varied. Birds observed outside the grid were recorded but were not used to calculate species density or diversity.

· Nocturnal surveys
Owls and goatsuckers were censused on the night of 03-04 June. Vocalizations of a Lesser Nighthawk (Chordeiles acutipennis), Common Poorwill (Phalaenoptilus nutallii), Western Screech-owl (Otus kennicottii), Barn Owl (Tyto alba), and Great Horned Owl (Bubo virginianus) were played at two locations along the dirt road East of the restoration site. Times of playback were 2120 on 03 June and 0230 and 0455 on 04 June. Each recording was played one time and paused for approximately two minutes to listen and watch for a response before playing the vocalization of the next species (at each location along the dirt road).

· Confirmed breeders
For species observed in the survey grid, Phil Unitt's guidelines (1997) for the San Diego County Bird Atlas project were used to determine whether or not the individuals were breeding.

· Species density, diversity, and equitability
Species density and diversity were calculated in the restoration site (0.4 ha), surrounding the restoration site (3.2 ha), and within the 18-square (3.6 ha) grid. Species density was calculated as the number of individuals found per hectare per survey. Species diversity was measured using the Shannon-Wiener diversity index (H) and species evenness was calculated with the Shannon equitability index (J). Calculation of these indices follows these formulae:

where S is the number of total species and pi is the proportion of individuals of species i.

Reptile and Amphibian Survey
The previous year's reptile and amphibian surveys were accomplished by checking cover boards placed in the restoration area. This method, while commonly used for "herp" surveys, was relatively unproductive resulting in the documentation of only a single lizard species, the Western fence lizard Sceloporis occidentalis. This year's surveys were accomplished by conducting time constrained area searches of a stretch of Murphy Canyon that included the restoration area. The survey reach stretched from the bridge, located approximately 300 meters upstream from the restoration area, to approximately 50 meters downstream of the restoration area. The surveys were conducted by walking slowly up the canyon and back while visually searching for reptiles and amphibians. All pre-existing cover boards located in the restoration area were checked during the surveys. The surveys were conducted for three hours each on four different dates beginning in the early spring. The survey dates and times were March 15, 2001: 1000-1300 hrs, April 9, 2001: 1000-1300 hrs, June 7, 2001: 1200-1500 hrs, and June 22, 2001: 1500-1800 hrs. Temperatures ranged from 70 to 90 F. The survey times and dates reflect an attempt to document as many reptile and amphibian species as possible based on their different ecological requirements (ex. surveying for amphibians in the spring when water is present in the creek). Surveying a stretch of Murphy Canyon that is greater in size than the restoration area itself resulted in locating reptiles and amphibians that could potentially occur in the restoration area. It was also possible to locate features of the canyon that are attractive to reptiles and amphibians and could serve as potential population sources (ex. pool near bridge where frogs breed or rock pile where snakes might occur).

Mammal Survey
Live trapping was utilized to assess diversity and abundance of small mammal species in a riparian woodland located in Murphy Canyon on MCAS Miramar, San Diego County, California. Two roughly parallel trapping transects were set out in the riparian habitat. Twenty-five Sherman live-traps were set on each transect with a 10 meter interval between trap stations. Traps were baited with rolled oats and set late in the afternoon, checked shortly after sunrise the next morning, and then closed for the day to avoid trap mortality for any diurnal species that may have been present. This protocol was followed for three consecutive nights (May 31 - June 2, 2001).

Captured animals were identified as to species and sex. Upon their initial capture, each animal was marked in the ear by a spot of black ink from an indelible marking pen; these marks were used to distinguish animals captured for the first time from those previously captured ("recaptures"). All captured animals were released at the point of their capture.

Species diversity for each habitat type was quantified using the Shannon-Wiener function:

where H is the measure of species diversity and pi is the proportion of the total sample captured along a transect belonging to the ith species (Krebs, 1989). Abundances of species on plots were estimated as the number of individuals captured per 100 trap-nights of trapping effort.

Results
Vegetation
Table 2 summarizes the ground water monitoring data from the two wells on site. In January, despite numerous rain events in prior months, there was no evidence of water flow down the constructed stream channel and the wells were dry. In March, however, several pools of standing water were present (Figure 6) and the wells held a significant amount of water. In April, it appeared that the constructed channel had flooded again. By May, the channel was dry and there was no standing water left in the wells.

Table 2
Well depth on the restoration site.

Figure 6. The constructed stream channel holding water in March 2001.

Percent survival of seedlings counted in late August 2000 (Table 3) was only 38 percent. This fell far below the success criteria of 98 percent for the first year after planting.

Table 3
Percent survival of transplants at the beginning and end of summer 2000.

The required success criteria for 2001 is 20% cover of native species. This year's cover (Table 4) was estimated at 9.5%, falling far below the requirement, but more than doubling the cover in 2000. Exotic species cover was 24.5%, also more than double the previous year's cover. All of the exotics counted in the percent cover estimate were treated with herbicide this year. By May, two months after treatment, all weeds were dead and dry while native species were still green and growing (Figure 7). Required species density for 2001 is 688 plants/Hectare. This year density was 1930 plants/Hectare, much higher than required (Table 5).

Table 4
Percent cover of native species in May 2001.

Figure 7. The restoration site two months after herbicide application.

Table 5
Density of native species in May 2001.

Results of the soil analyses (Figure 8) conducted for the restoration and undisturbed sites were similar this year and stable over the last two years, with one exception. Soil nitrate levels this year were much higher at the restoration site than at the undisturbed site, changing dramatically from May 2000 when levels were nearly identical.

Figure 8. Results of soil analyses since 1999.
(click for larger image)

Bird Survey
· Diurnal surveys
Twenty-five avian species were identified within the census grid overlaying the restoration site (Table 6). Some hummingbirds could be identified only to subfamily. Three additional species were identified only outside of the census grid and three species identified as fly-overs could only be identified to family or genus (Table 7). The unidentified orioles were likely Hooded Orioles (Icterus cucullatus) since Bullock's Orioles (Icterus bullockii) are less likely to be found in such dry habitat. Species density by area (individuals per hectare per survey) within the restoration site, inside the grid outside of the restoration area, and within the entire census grid is reported in Table 8. Number of species, density of individuals of all species, the Shannon diversity index (SDI), and the Shannon equitability index (SJI) for the census grid are reported in Table 9 along with pre-restoration results from the non-breeding season in 1999. Raw survey data and totals are reported in Appendices 3-9.

Table 6
Avian species survey totals.
Common Name Scientific Name Restoration Site Surrounding site Total in Grid

Table 7
Avian species identified only outside of the grid or only as fly-overs.

Table 8
Avian species density (individuals per hectare per survey).

Table 9
Avian species number, density, diversity (SDI), and equitability (SJI).

· Nocturnal surveys
No species were identified within the grid during the nocturnal surveys.

· Confirmed Breeders
Breeding of seven species was confirmed within the grid. Table 10 lists the species that were confirmed as breeders, how breeding was confirmed, the date of confirmation, and in what plot number of the grid the confirmation was made. Though not confirmed, it is believed that Blue Grosbeaks (Guiraca caerulea) and California Gnatcatchers (Polioptila californica) were also breeding in the restoration area because of activity observed by these species outside the restoration area.

Table 10
Avian species confirmed as breeders within the survey grid.

Reptile and Amphibian Survey
During the surveys, four reptile and amphibian species were observed (Table 11). They are the Pacific tree frog (Hyla regilla), the Western fence lizard, the Orange throated whiptail (Cnemidophorus hyperythrus), and the two-striped garter snake (Thamnophis hamondii). A rattle belonging to an unidentified rattlesnake (Crotalus sp.) was also found on site. Interestingly, no exotic reptiles, amphibians, or invertebrates were observed in the creek or any of the pools found on site.

Table 11
Reptile and amphibian species observed in and around the restoration site.

The observations are summarized by date below:
· March 15, 2001: 1000-1300 hours
Water is present in several large pools in Murphy Canyon Creek. Pool #1 (located at base of old bridge) was approximately 10 feet in diameter and 4.5 feet deep. Numerous Pacific tree frog larvae were present in the pool and no exotic species were observed. Pool #2 (located at edge of fence approximately 50m downstream of pool #1) was 30feet long, 4 feet wide, and 1 foot deep. Numerous Pacific tree frog larvae were also present in this pool and no exotic species were observed. Pool #3 (located adjacent to restoration area at bend in creek) was 8 feet long, 3 feet wide, and 3 feet deep. No reptiles, amphibians, or exotics were observed here. There were several small pools located in the restoration area itself but no reptiles, amphibians, or exotics were observed in any of them. A single Western fence lizard was observed basking approximately 200 meters up canyon of the restoration area.
Observed herps: Pacific tree frog (larvae), Western fence lizard.

· April 9, 2001: 1000-1300 hours
Pool #1 was of similar dimensions as during the previous visit. Numerous Pacific tree frog larvae were still present, exotics absent. A relatively large (approx. 70 cm long) two-striped garter snake was observed at the edge of this pool. Pool #2 also had numerous Pacific tree frog larvae present, exotics absent. Another two-striped garter snake was observed adjacent to the restoration area at the edge of pool #3. Pools in the restoration area were dry or almost dry.
Observed herps: 2 two-striped garter snakes, Pacific tree frog larvae.

· June 7, 2001: 1200-1500 hours
Pool #1 was reduced in size by at least 70%. Large numbers of bees were swarming near the water making it difficult to make observations. Pools 2 and 3 were dry. One Western fence lizard was observed in a willow (Salix sp.) adjacent to restoration area. Three more Western fence lizards and three orange-throated whiptails were observed up canyon within 200 meters of the restoration area. A rattle belonging to an unidentified rattlesnake was found in a rock pile located up canyon from restoration area.
Observed herps: 4 Western fence lizards, 3 orange-throated whiptails, 1 rattlesnake rattle.

· June 22, 2001:1500-1800 hours
Pool #1 still contains a small amount of water. No reptiles, amphibians, or exotics were observed in it. A single Western fence lizard was observed near the old bridge.
Observed herps: 1 Western fence lizard.

Mammal Survey
The live-trapping survey consisted of 150 trap-nights of trapping effort. During the survey, 8 different individuals were captured a total of 9 times for a trapping success rate of 6%. A total of 5 species of small mammals were captured during the study. None of the species captured has been designated as "endangered" or "threatened" by either federal or state agencies.

Table 12 presents a summary of results of the live-trapping survey. Results from a similar survey conducted at the same site in 1999 are also presented in Table 12 for comparative purposes. A total of 100 traps were deployed for 3 nights in 1999, so the trapping effort for that survey was twice that of the 2001 survey. The 1999 and 2001 surveys were very similar for most parameters and both revealed a typical riparian woodland rodent community. Species diversity in 2001 (H = 1.560) was slightly lower than diversity in 1999 ( H = 1.657).

Cactus mice (Peromyscus eremicus), California mice (Peromyscus californicus), San Diego pocket mice (Chaetodipus fallax), desert woodrats (Neotoma lepida), and deer mice (Peromyscus maniculatus) were captured in both surveys and are common inhabitants of riparian woodlands in southern California. Cactus mice were the most abundant species captured in 1999, but in 2001, captures were nearly equal across species (Table 12). Overall rodent abundance was slightly lower in 2001 (5.3 individuals/100 trap-nights) than in 1999 (7.3 individuals/100 trap-nights) as estimated from live trapping.

Table 12
Mammal species number, diversity (H), and relative abundance in 1999 and 2001.

Discussion
Vegetation
The high mortality rate of seedlings planted during spring 2000 was likely caused by the extremely dry rainy season during the La Niña event of 1999-2000. Riparian species had a much higher mortality than the CSS species. This follows because riparian species require more water than CSS species and, for reasons unknown, the western portion of the site where most of the riparian plants were installed had a lower water table than the eastern, CSS seedling dominated side of the site (Table 2). Since survival was so low, 137 replacement seedlings, primarily riparian species, were planted during spring 2001. These seedlings are being watered much more frequently (twice each month throughout the summer) than the original transplanted seedlings to ensure their survival. In March 2001, a vehicle drove through the extreme northern portion of the site, creating deep ruts in the wet soil and burying several transplants. The number of plants lost is unknown as they were buried in the mud, but next year's survival count will help determine this.

The presence of water in the constructed channel demonstrates its successful creation. It therefore does not appear that recontouring is necessary at this time.

The primary reasons for low percent cover of native species this year were high transplant mortality and high percent weed cover. Replacement seedlings planted this year and more frequent watering should improve cover for next year. In addition, the herbicide spraying done this year that could not be completed in 2000 will dramatically reduce the number of weeds that germinate next year. Native species density was far above the required success criterion, so no remedial measures are necessary to improve density.

Soil conditions at the restoration and undisturbed sites are very similar with respect to percent organic matter, available phosphorus, total nitrogen, and pH, so no remedial measures are necessary at this time. The high nitrate levels at the restoration site are likely due to the high density of exotic annuals. Since herbicide spraying was not permitted in time for the 2000 season, the site became dominated by exotic annuals. During the dry season, these annuals died and were broken down into nitrates by soil microorganisms. Since the number of annuals at the site is so much higher than at the undisturbed site, there are more nitrates being deposited into the soil. In addition, percent cover of shrubs at the undisturbed site is much greater than at the restoration site, so the uptake of nitrates by plants from the soil is much higher. This also contributes to the high nitrate levels at the restoration site. Nitrate levels at the restoration site should decrease over the next several years as the seed bank of exotic annuals is depleted by more regular herbicide spraying and as native shrubs grow larger and absorb more nitrates from the soil.

Bird Survey
Point counts were conducted in the area bordering the restoration site because, due to the small size of the site, conducting point counts only within the restoration site would underestimate the number of birds that use it. Additionally, any changes in the avifauna surrounding the site will provide a broader picture of any impacts of the restoration effort. One problem of using a small plot size for the point count is the potential for overestimation. This possibility was minimized by not counting individuals that were located near areas in which the same species was recently recorded.

This year's bird survey represents the first census of birds at the restoration site during the breeding season. Therefore, it is difficult to comment on these surveys without a similar study to which results can be compared. The most appropriate comment is that the species composition and densities within species are expected for the type of habitat in and around the restoration site. The diversity and equitability indices all increased over those calculated for the surveys conducted at the site in September of 1999. This change was expected as Neotropical migrants that are summer breeders were not present during the September surveys. The diversity indices are somewhat low, but may be reasonable for a small plot such as this site. The equitability indices, however, are relatively high, as all are near 0.90. The highest equitability score possible is 1 (Begon et al. 1996). The equitability indices would have been even higher were it not for the high numbers of Houses Finches (Carpodacus mexicanus), Spotted Towhees (Pipilo maculatus), and Wrentits (Chamaea fasciata). All of these species, however, are native to the region and are expected to be present in relatively high densities. Thus, the diversity and equitability indices calculated for these surveys indicate that the avian community within the survey grid is fairly diverse and is not dominated by a single species.

Reptile and Amphibian Survey
With only two reptile and amphibian species (Western fence lizard, two-striped garter snake) observed in the restoration area itself, it appears that there is relatively low reptile and amphibian diversity and abundance. However, surveys of the areas adjacent to the restoration site revealed the presence of at least three more species (Pacific tree frog, orange-throated whiptail, and rattlesnake species) that could potentially occur on site. Also, certain habitat features that exist in Murphy Canyon could act as population sources of reptiles and amphibians that might utilize the restoration area. These features include the series of pools, including the large pool (#1) located at the base of the old bridge, the rock and debris pile located near the old bridge, and the rock and debris pile located approximately 100 meters up canyon from the restoration area. As restoration of the riparian area continues, this area will likely become more appealing to reptiles and amphibians, especially those associated with riparian habitats, such as the Pacific tree frog and the two-striped garter snake.

Mammal Survey
The small decrease in species diversity this year can be attributed to the fact that the California vole (Microtus californicus) was captured in 1999 but not in 2001. Vole species are noted for undergoing cyclic changes in abundance and their absence from the 2001 survey may indicate that populations of M. californicus were in the low phase of a cycle. It should also be noted that M. californicus is, like most voles, primarily a grassland inhabitant, and riparian woodland may be a marginal habitat for them.

The difference in rodent abundance between the two years' surveys is not significant and undoubtedly reflects normal year-to-year variation in rodent abundance.

Conclusions and Recommendations
Vegetation
Water levels in the wells and water flow through the constructed stream channel will continue to be monitored once or twice each month during the 2001/2002 rainy season.

Seedling survival will be counted next in January 2002. If results fall below 80% as required by the success criteria, additional seedlings will be planted during the winter/spring 2002. Irrigation in 2001 will continue twice a month for the seedlings planted this past spring until winter rains begin. During summer 2002, irrigation will continue to be provided to these seedlings (once per month). Planting, if necessary, will be concentrated in the riparian section of the site and, after repairs to the tire ruts, in the section that suffered vehicle damage.

Exotic plant species will be sprayed with herbicide again in 2002, beginning when the first weed seedlings reach four inches tall. A more regular schedule of herbicide application is anticipated next year and through 2004 because more licensed applicators are now available. Weeds will be sprayed throughout the winter, spring, and summer until the dry season when seedlings have stopped germinating. This regimen will continue each year until the end of the project, thereby greatly reducing the existing seed bank and improving the likelihood of survival of native transplants that, once they have grown large enough, will outcompete the exotic species without the help of herbicide applications. Reducing the number of exotics at the restoration site should encourage higher native species cover and help reduce the elevated nitrate levels.

Bird Survey
Specific predictions regarding the potential changes in the avian community during the breeding season as a result of the restoration project are not feasible at this time. However, it is safe to make the general prediction that the densities of birds typical of riparian habitat will most likely increase with the age of the restored habitat. For example, the presence of the young male Blue Grosbeak and his mate may mark a dispersal event from a larger population in more suitable habitat. The addition of more large willows to the site may attract more Blue Grosbeaks. This survey will be repeated in Spring 2002.

Reptile and Amphibian Survey
Since sampling methods during this year's reptile and amphibian survey differed from those of the 1999 survey, it is impossible to examine changes in species density since 1999. However, in 1999, only one species was found on the site while in 2001, two species were found. This is a definite improvement, especially since the two-striped garter snake, which was not seen previously, is associated with riparian habitats. This survey will continue next year in Spring 2002.

Mammal Survey
Mammal species diversity and abundance were not notably different this year from the 1999 results. An increase in diversity and abundance are a sign of recovery of the restoration site, but it is still too early to notice a significant change. The survey will next be repeated during Spring 2002.

First Annual Report (October 28, 2000)

Third Annual Report (Feburuary 17, 2002)