Soil Ecology and Research Group

last update February 3, 2000




The purpose of this project was to develop, test and implement coastal sage scrub plant community restoration techniques on open space adjacent to the Palos Verdes Navy Housing Site (San Pedro). This is being done to reduce erosion and improve ecological quality. This project improved and implemented methodologies that can be used to restore sensitive Navy lands on the Palos Verdes Peninsula. This involved all aspects of restoration including development of techniques, evaluating local species for use, nonnative species removal, designing experiments to evaluate effectiveness, implementing restoration and maintaining and monitoring the site.


The scope of this project encompasses the development and implementationof plant restoration techniques on open space adjacent to the Palos VerdesNavy Housing Site in San Pedro, California. Field work was commenced inOctober 1996 and continued through November 1999.

The site required both erosion control and plant restoration. Techniquesof slope stabilization and erosion control using native plant revegetationand other organic means are effective not only in reducing erosion problemsbut also provide aesthetic and biological benefits. The coastal sage scrubhabitat, which occurs naturally throughout the San Pedro site, definedthe revegetation plant palette, and the propagules needed for the revegetationeffort were collected locally. Both hydroseeding and planting of nurserygrown native plant stock was used to revegetate the site. Emulating thenative vegetation structure maximizes the habitat value of the revegetatedareas, supporting populations of California gnatcatcher (Polioptilacalifornica ssp. californica ), Palos Verdes Blue Butterfly (Glaucopsychelygdamus ssp. palosverdesensis ) and other sensitive species. The proximityof the site to several oil refineries and other heavy industry puts itat risk for high nitrogen deposition, which can create a profound shiftfrom native perennial plants to exotic annuals in the absence of activemanagement.

Study Area

The site is located in Los Angeles County at the Defense Fuel Support Point (DFSP) in San Pedro. The site is south west of the intersection of Palos Verdes Drive North and Gaffey Road, with access occurring from Gaffey Road. Revegetation and research took place on an open area adjacent to the softball fields at the Palos Verdes Naval Housing Site in San Pedro (Figure 1). The site is approximately 1.5 acres in size and is located directly north of the softball fields. The area is composed of sandy loam soil texture with a large south-facing slope and a smaller east-facing slope on its southwest side. The south-facing slope drops approximately 25 feet and the east facing slope about 6 feet. Slope inclinations are as high as 45o. The slopes appear to have been shaped when the softball fields were constructed and may have been mechanically compacted during development. Since its construction, the slope has developed several large erosion gullies which, after heavy rainfalls, wash soil onto the road surrounding the perimeter of the softball fields. Such erosion was, in part, caused by an inadequate amount and type of vegetation established on the slope after construction was completed. Vegetation at the site may be characterized as a "typical urban weed assemblage", consisting of several native species and exotics such as Hirschfeldia incana, Cynodon dactylon, Helianthus annuus, Conyza canadensis, Ricinus communis, Salsola tragus and brome grasses.

Figure 1.  Site before installation.  1) East Slope in foreground and South Slope in background 2) South Slope

Materials and Methods

Soil samples were collected on 3 November 1996 and 11 November 1997and analyzed to test and monitor several factors. Inorganic nitrogen levelsfrom both years were analyzed to indicate the severity of atmospheric nitrogendeposition. The proximity of the site to several oil refineries and otherheavy industry puts it at risk for high nitrogen deposition; a situationwhich can create a shift from native perennial plants to exotic annualsunless active management practices are in place. Exotic annuals are muchmore capable of using excess available nitrogen in the soil than nativeperennials, giving them a competitive edge in high nitrogen environments.Excessive nitrogen deposition may be remediated by application of carbon-richsurface mulches which should immobilize excess nitrogen and disadvantageexotic annuals.

Soil samples also underwent a standard suite of tests to ensure suitabilityfor native plant germination, growth and establishment. These tests includenitrogen, phosphorus, pH, electrical conductivity and texture. Soil seleniumwas included in these tests due to its potential significance in relationto Ocean Locoweed (Astragalus trichopodus var. lonchus) andthe Palos Verdes Blue Butterfly (Glaucopsyche lygdamus ssp. Palosverdesensis).

Additionally, the soil was tested for presence and density of soil microbiota(fungi and bacteria) using the europium staining method. This allowed usto determine whether steps needed to be taken to increase beneficial soilorganisms and also served as a method by which to quantify the successof the revegetation at restoring a self-sustaining soil biota.

Plant tissue of selected plants were also analyzed. The standard suiteof tissue nutrients (nitrogen, phosphorus and potassium) were performedon the dominant species. This served to assess the health of the plantsin the revegetation and also relates to the food quality for G. lygdamus.For A. trichopodus, tissue selenium was analyzed. High tissue seleniumimparts pest resistance to the A. trichopodus and may relate tothe unpalatability of G. lygdamus to its predators.

Slopes were initially prepared by filling erosion rills with straw bale dams and, using hand tools, generally contouring the slopes to reduce focused water runoff. The installed irrigation system was repaired and used to implement a grow-and-kill program of weed eradication. The existing seed bank was induced to germinate through irrigation and nonnative species were killed with a foliar Round-Up application. This procedure was periodically repeated several times over a three year time span, starting in October 1996 and ending in October 1999. The softball field slope was partially fenced (with a temporary plastic link fence) at the major foot traffic areas to minimize disturbance within the restoration area.

A surface erosion control experiment was set up in January 1998. Thisexperiment was designed to determine the most effective surface erosiontreatment to stabilize the soil surface and facilitate vegetation establishment.This experiment is identical to and was set up as a companion study toones occurring at Point Loma Fleet Industrial Supply Center (FISC) andFleet Combat Training Center, Pacific (FCTCP) Navy commands. In the DFSPexperiment, 14 treatments were placed on the upper portion of the southernslope. There were roughly two replicates per treatment, which include organicerosion control materials such as Curlex, ENCS-2 (straw mats), jute cloth,coir fence, cocoa mulch, pitting with cocoa mulch, vegetation plots (eachwith five native plants) and a control plot. Each treatment encompassedan area of 50 square feet, sectioned off from the surrounding soil on theslope by 1"x4" wooden boards placed upright in the soil. Each treatmentwas chevron shaped at the base, with the open end of the chevron facinguphill. An opening at the base was attached with a porous bag to capturesoil that washed down the slope of the enclosed treated area after everyrainfall. The bag captured the soil but allowed the water to pass through.The weights of the bags were then measured after each rainfall to comparethe different rates of erosion within each treatment.

Adjacent undisturbed vegetation was surveyed by species to fine tunethe plant palette and also to serve as a revegetation performance standard.Seed was collected in the DFSP area to ensure that plant material usedin the revegetation effort was of local genetic stock.

In May 1997, 167 Astragalus trichopodus and 24 Lotus scopariusnurserygrown container shrubs were planted. Half of the A. trichopodusand16 of the L. scoparius were inoculated with Rhizobia.TheA.trichopodus were grouped into four and one-half experimental blocks.Each block contained 24 plants, seven groups each with six plants. Thesix plants received the following treatments: inoculated and treated witha bark soil amendment; inoculated and treated with a sawdust soil amendment;inoculated with no soil amendments; uninoculated and treated with a barksoil amendment; uninoculated and treated with a sawdust soil amendment;and uninoculated with no soil amendments.

The southern slope was hydroseeded in January 1998 according to an experimentalplan designed by SERG personnel. The experiment was designed to explorethe effectiveness of several variations on standard hydroseeding and theirpotential for habitat restoration. The entire south facing slope was dividedinto fifteen 20 meter wide plots. These plots were divided into west, middleand east blocks, with each block containing five different hydroseeding/mulchtreatments (Figure 2). The hydroseeding/mulch treatments included: (1)standard single hydroseed application using psyllium as a binder and recycledpaper as a carrying agent; (2) punched straw installed prior to hydroseedapplication; (3) chipped straw containing psyllium as a binder blown onfollowing standard single hydroseed application; (4) redwood compost mixedwith psyllium and blown on following standard single hydroseed application;and (5) application of extra recycled paper hydromulch following hydroseedapplication. The seed mix consisted of coastal sage scrub species suitablefor growth on open slopes. Species included Eriogonum fasciculatum,Eriogonum cinereum, Artemisia californica, Encelia californica, Lotus scopariusandHeterothecagrandiflora. Percent cover and density for all species present in thehydroseeding experiment was recorded using seven randomly placed 100 x120 cm quadrats per treatment. Data was collected in June 1998, five monthsafter application and in March 1999, 14 months after application.

In February 1998, in an effort to explore alternative revegetation options,approximately 150 individual plants, including Artemisia californica,Lotus scoparius and Baccharis glutinosa seedlings, were removedfrom the softball overflow parking lot and transplanted on to the eastfacing slope.

In May 1998, due to insufficient seed germination from the hydroseedingexperiment, 544 nursery grown container species from Palos Verdes seedsources were installed in the western and middle third of the southernslope. Species include 130 Eriogonum fasciculatum, 141 Artemisiacalifornica, 118 Encelia californica, 106 Eriogonum cinereum,25Lotus scoparius, 13 Baccharis glutinosa, 8 Ericameria cooperiand 3 Isomeris arborea. Data was collected in March 1999 forall transplanted or container grown species.

In April 1999, due to low survival rates and slope disturbance fromboth humans and wild dogs, 255 nursery grown container species were plantedon the southern slope. Species include 112 Artemisia californica,62 Encelia farinosa, 47 Salvia mellifera, 30 Lotus scopariusand 4 Eriogonum cinereum.


Soil samples were collected on 3 November 1996 and 11 November 1997and analyzed. To compare nitrogen deposition results over time, NO3(ug/gm), NH4 (ug/gm) and total known nitrogen (mg/gm) were investigated(Table 1). There were significantly higher levels of NO3 (p=0.0003)and NH4 (p=0.047) found in the 1997 samplings. There was nosignificant difference (p>0.05) of total known nitrogen levels.

Table 1. Comparison of levels of nitrogen in soil between 1996 and 1997.


Nitrogen soil sample

NO3 (ug/gm)

NO3 Std. Error

NH4 (ug/gm)









NH4 Std. Error
Total Known Nitrogen (mg/gm)
TKN Std. Error


Soil samples were taken in November 1997 to investigate the soil propertiesin the three blocks of the hydroseeding experiment. One-way ANOVA was usedto analyze soil properties, including NO3 (ug/gm), NH4(ug/gm),total known nitrogen (mg/gm), phosphorus (ug/gm), pH, electrical conductivity(mhos), percent moisture and percent organic matter (Table 2). There wasno significant difference (p>0.05) between any of the blocks in terms ofNH4, pH, phosphorous, percent moisture or percent organic matter.There was significant difference in the following soil properties: NO3(p=0.02), with block one containing the highest amount at 23.940 ug/gmand block two containing the lowest amount at 7.367 ug/gm; electrical conductivity(p=0.004), with block two the highest at 0.148 mhos and block one the lowestat 0.070 mhos; and total known nitrogen (p=0.047), with block one containingthe highest amount at 0.293 mg/gm and block two containing the lowest amountat 0.153 mg/gm. Traces of selenium could not be detected in the soil samples.

Table 2. Soil sample results before application of Hydroseed.


Soil sample
Block one
Block two
Block three
NO3 (ug/gm)
NO3 Std. Error
NH4 (ug/gm)
NH4 Std. Error
TKN (mg/gm)
TKN Std. Error
P- (ug/gm)
P- Std. Error
pH Std. Error
EC (mhos)
EC Std. Error
% Moisture
% Moist. Std. Error
% Organic matter
% OM Std. Error


The 3 November 1996 soil samples were tested for presence and densityof soil microbiota (fungi and bacteria) using europium staining methods.Results are forthcoming.

Plant tissue samples from selected plants have not been analyzed atthe present time.

The surface erosion control experiment revealed that ground cover typeof erosion control devices were more effective at reducing the rate ofsoil erosion after periodic rainfall events (Table 3). The cocoa mulchcould not be analyzed adequately because the mulch material tended to clogup the mouth of the soil collection bag, reducing the amount of soil capturedin the bag. The least effective means of erosion control was the vegetationplots while the most effective methods appear to be curlex, jute and ENCS-2.

Table 3. Data from erosion control experiment


Pounds sediment per acre per 1.0" rainfall
Coir fences
Pitting and mulch


The slopes were hydroseeded in January 1998 according to the experimentaldesign. Data was taken in March 1999, 14 months after application of hydroseed(Figures 3, 4 and 5). Data was collected only from block three because,with the exception of the hydroseed/hydromulch and hydroseed/compost plotsin block two, blocks one and two had virtually no seedling growth. In blockthree, average percent cover of native species seedlings was six percentwhile the average exotic species percent cover was 8.2 percent (Figure6). The average native species density was 12.1 seedlings/1.1m2(Figure 7). There was a significant difference (p=0.008) in percent coverof native species, with the hydroseed/hydromulch treatment highest at 10.9percent and the punched straw/hydroseed lowest at 2.4 percent. There wasno significant difference (p>0.05) in density of native species among thetreatments. An overall list of native plant species with their respectivepercent cover and density is shown in Table 4. There was no significantdifference (p>0.05) of preferential growth among the treatments for exoticspecies. Exotic plant species seedlings found include Anagallis arvensis,Brassica nigra, Centaurea melitensis, Chenopodium murale, Chrysanthemumcoronarium, Cotula australis, Erodium cicutarium, Medicago polymorpha,Melilotus officinalis, Mesembryanthenum crystallinum, Raphanus sativus,Salsola tragus, Schismus barbatus and Taraxacum officinale.

Figure 3.  Hydroseeding experiment plots 1 & 2.  1) Standardsingle application of hydroseed and 2) Punched straw treatment.

Figure 4.  Hydroseeding experiments plots 3 & 4.  1) Plot3 blown straw and 2) Plot 4 corrpost.

Figure 5.  Hydroseeding experiment plots. Hydromulch

Table 4. List of native plant species seedlings found growing in hydroseedexperiment.


Mean % cover
Standard error
Standard error
Artemisia californica
Eriogonum fasciculatum
Heterotheca grandiflora
Ambrosia acanthicarpa
Lotus scoparius
Eriogonum cinereum
Baccharis glutinosa
Datura wrightii
Encelia californica


Survivorship of planted native species was monitored in March 1999.Monitoring of all planted native species proved to be problematic. Plantprotection devices and flags, used to mark the location of the plantednative species, were subject to vandalism or removal. As a result of theseactivities, native species planted later in the term of the contract werenot marked in an attempt to make them inconspicuous and therefore lessopen to vandalism. Monitoring was done for those planted species not markedby assuming that a successful plant should be 1-2 feet in size (Figures8 and 9).

In May 1997, 167 Astragalus trichopodus and 24 Lotus scopariusnurserygrown container shrubs had been planted. In March 1999, only two A.trichopodus species were alive. Many of these species, however, hadgone to seed before they died. The L. scoparius species used inthe inoculation experiment could not be detected. Approximately 96 of the150, or 64 percent, transplanted species onto the east-facing slope werealive. Of the 544 native plant species planted in May 1998, approximately300, or 55 percent, were still alive. In March 1999, average percent coverof native species, as a result of both planting and seeding, is roughly20-25 percent. The contract required 50 percent cover at the site in May1999. Two hundred and fifty-five more native species were planted in Aprilto raise the percent cover. In November 1999, a plant survey was takenusing the California Native Plant Society (CNPS) standard method. Three50 meter transects were surveyed,

Figure 8. Southern Slope in Spring 1999. 1) Western portion of slopeand 2) middle portion with erosion control experiment at top of slope.

Figure 9. East facing slope  in Spring 1999. 1) overview of slopeand 2) close up of slope with volunteer Phacelia sp. seeding.

with two on the south facing slope and one on the east facing slope. Overall percent cover was 55 percent (Table 5), with 22 percent being native plant species and 33 percent being nonnative plant species. Bare ground accounted for 45 percent cover. The Fish and Wildlife Service (FWS) has a criteria for recreating Palos Verdes Blue butterfly habitat based on percent cover of certain species over a certain amount of time. Results are shown in Table 6.

Table 5. Percent cover of plant species using the CNPS standard method.

Transect location
Native species
Nonnative species
Bare ground
South slope - east
South slope - middle
East slope
Average % cover


Table 6. FWS Palos Verdes Blue Butterfly habitat recreation criteria.

Actual % cover
FWS 2 year % cover criteria
BacPil, EriPal, IsoArb, SalMel
PVB Host Plants (LotSco, AstTri)
Other Woody spp
Other Herb spp



Soil samples revealed a general increase of nitrogen in the form of NO3 and NH4 between 1996 and 1997. Levels of total known nitrogen in the soil, however, did not increase. The proximity of the site to several oil refineries and other heavy industry puts it at risk for high nitrogen deposition. Although data collected and analyzed is only for one year, preliminary results show that the risk for high nitrogen deposition may in fact be a reality. Such a condition can create a profound shift from native perennial plants to exotic annuals in the absence of active management. One technique for restoration under these circumstances is to amend the soil with organic mulch in order to lower the carbon:nitrogen ratio. A low carbon:nitrogen ratio is essential for coastal sage scrub communities to remain intact. A shift towards more nitrogen in the soil encourages the growth of exotic annual species pre-adapted to flourish in disturbed areas. Evidence to support this scenario may be demonstrated in the hydroseed experiment. The native plant species seedlings had a significantly higher percent cover in the hydroseed/hydromulch and hydroseed/compost plots. These plots arguably received a greater amount of organic mulch, but more importantly, the mulch was distributed evenly. In the blown straw, punched straw and standard single application of hydroseed plots, distribution of organic mulch became patchy over time, providing less protected organic mulch areas for the native seeds to germinate effectively and leaving more areas open for nitrogen deposition to come in direct contact with the soil. Although not statistically significant, percent cover of exotic species was lowest in the hydroseed/hydromulch plot and highest in the standard single application of hydroseed plot. This suggests that increasing the organic mulch content during the seeding process increases the growth rate of native plant species while decreasing the growth rate of exotic plant species.

Density of native plant seedlings was not significantly affected bythe hydroseeding treatments, although numbers were highest in the hydroseed/hydromulch,followed by hydroseed/compost and punched straw/hydroseed plots. This suggeststhe greatest benefit of hydroseed/hydromulch and hydroseed/compost plotsis not necessarily that it increases the number of native plant seed germinations,but rather it provides a more favorable habitat for seedling growth. Artemisiacalifornica, Eriogonum fasciculatum, Heterotheca grandiflora andAmbrosia acanthicarpa appeared to benefit the most from this type ofhydroseed application.

Data was collected only from block three because, with the exceptionof the hydroseed/hydromulch and hydroseed/compost plots in block two, blocksone and two had virtually no seedling growth. This suggests that some otherfactors were at work. Soil analysis of the three blocks of the hydroseedingexperiment revealed that block one had significantly higher levels of N03and total known nitrogen. These soil conditions may have led to diminishednative plant species seed germination. Block two, on the other hand, hadthe lowest levels of NO3 and total known nitrogen, with slightlymore favorable seed germination results than block one. The hydroseed/hydromulchand hydroseed/compost plots in block two may have benefited from this,but not the other plots. Electrical conductivity, a measurement of soilsalinity, was significantly higher in block two and lowest in block one.

Whether these significant differences in soil properties pertain to seed germination, or suggest that the characteristics in block three represent ideal soil property values for seed germination, remains speculative at best. Thus, with an unclear connection to soil properties, the most likely cause for minimal seed germination in blocks one and two would more than likely have to be due to other factors. When the hydroseeding experiment was executed, the first block to receive the standard application of seeds was block three, then block two and finally block one. There is a good chance that, due to perhaps an inadequate mixing of seed and carrying agent, the majority of the native plant seeds were dispersed in block three. By the time the hydroseed truck had reached the middle of block two, it may have run out of seeds. Another factor involved was slope steepness. Block three had slopes averaging approximately 20 degrees or less. Blocks one and two, however, had steeper slopes, averaging over 30 degrees. Physical evidence of the hydroseeding after 14 months, in terms of the amount of carrying agent still intact, was more prevalent on the less steep slopes of block three. It is highly probable that no hydroseeding treatment was a viable method for blocks one and two because of the combination of steepness and sandy soil. Such a situation probably led to the removal of much of the seed through wind and water erosion and the impact of humans walking up and down the slope to view the softball games.

At this time, europium and plant tissue results are not available for analysis of applied ecological principles in relation to the dynamic interaction between soil and plant physiology's.

The surface erosion control experiment revealed that the type of erosioncontrol devices which completely covered the ground were more effectiveat reducing the rate of soil erosion after periodic rainfall events thanspot erosion control techniques. The materials most successful in reducingerosion were curlex, jute and straw mats. The cocoa mulch initially provideda complete coverage of the ground but after heavy rainfall tended to washdown the steep slope. The use of planted native species as erosion controlwas not as effective as the blanket types of erosion control. However,these plants were still young and had not yet reached mature size. Maturevegetation is very effective at holding the soil in place on slopes andresults may demonstrate the effectiveness of vegetation erosion controlonce the seedlings reach maturity. The least effective means of erosioncontrol proved to be the coir fences. These fences are quite porous, andproved to be ineffective in stopping the flow of the sandy soil duringprecipitation events.

Results from the companion erosion control studies at Point Loma somewhatmimic those at Palos Verdes. In terms of pounds of sediment lost per acreper one inch of rainfall, the following lists in order the least effectiveto most effective erosion control devices: pitting and cocoa mulch (4695lbs.), control (4287 lbs.), coir fences (3288 lbs.), cocoa mulch (2186lbs.), Jute cloth (1969 lbs.), vegetation (1457 lbs.), ENCS-2 (673 lbs.)and curlex (283 lbs.). Once again, the blanket type of erosion controlproved to be the most effective. Unlike at Palos Verdes, the planting ofvegetationwas nearly as effective as the blanket type of erosion control. The shrubsplanted at Point Loma were installed earlier in the growing season, allowingthem a chance to become established before sediment data was collected.The more established plants at Point Loma may have provided better erosioncontrol than the younger, less established plants at Palos Verdes.

To some extent, the hydroseeding experiment also provided erosion control. The hydroseed carrying agent remained intact on the more gentle slopes, but had a tendency to break apart on the steeper slopes, leaving the loose native soil exposed to rainfall and erosion. The treatments with the added layer of hydromulch or compost behaved in the same way as the standard single application of hydroseed. Erosion of varying degrees of intensity occurred in all plots, but more commonly on the steeper slopes. The punched straw remained in place, but rills occasionally formed in between the individual patches of straw. Rills also formed in the blown straw treatment. Here, the blown straw had a tendency to settle in the rills and other low spots, thus becoming concentrated, where in some cases, it slowed down small scale erosion.

Twenty-two months after installation, only two of the 167 Astragalustrichopodus were still alive. Over 75 of the deceased plants, however,had reached a height of two to three feet and had produced seed beforedying. A. trichopodus is classified as a perennial, however,in this case it appears to be a short lived perennial; with species existencerelying on propagation of seeds more so than a long lived sporophyte. Itis highly probable that the seed bank has a large supply of viable A.trichopodus seed from this original group of plants; however, sincethere has been no significant rainfall since seed set, germination hasnot occurred. A one year monitoring would have given better results asto the effects of added soil amendments and Rhizobia inoculationof A. trichopodus roots. Selenium could not be detected in the soiland tissue samples were not taken of the two live A. trichopodus plantsdue to the small sample size.

Approximately 96 of the 150 mature shrubs transplanted onto the east-facingslope were alive after 13 months. These shrubs received no further careafter being transplanted. They were, however, transplanted during an ElNiño winter where an adequate supply of moisture enabled many ofthe shrubs to survive the shock of transplanting. The area being salvagedwas close to the restoration site and allowed SERG personnel to accomplishthe work in one day. Thus, an increased number of plants were added tothe restoration site at a low cost, with average survivorship results.The locally salvaged plants had a higher survival rate than the containergrown plants, and were less expensive to install. Of the 544 native plantspecies planted in May 1998, approximately 300, or 55 percent, were stillalive after ten months. The only limitation of plant salvation is that,in this case, more container grown plants could be obtained.

Average percent cover of native species in November 1999, as a resultof both planting and seeding, is 22 percent. The contract required 50 percentcover of native plant species at the site by May 1999, two and a half yearsafter commencement. Planting and seeding of native species, however, wasinitially delayed one year due to a scant seed bank of local stock. Ineffect, the results from November 1999 are more in line with the one anda half year contract required percent cover, which was 25 percent. In addition,the percent cover results are also in line with the success criteria establishedby the United States Fish and Wildlife Service for the restoration of PalosVerdes Blue Butterfly habitat for two years after initial reestablishment.


Nitrogen deposition appears to be occurring at the site. When such a situation occurs, a common restoration management technique normally used is to add organic mulch to the soil. The use of added organic mulches to the seeded area at Palos Verdes appears to have increased the size, but not necessarily density, of native plant species seedlings by raising the carbon:nitrogen ratio. As an added benefit, the mulches reduced the number of exotic plant species from growing, making them easier to control with either hand weeding or herbicide application. Based on the results of this project, it is recommended that a second layer of hydromulch be added to the standard single application when using hydroseeding as a restoration procedure. The hydromulch mix is favored over compost because the compost clogged up both the hydroseed nozzles and pump of the hydroseeding truck.

Hydroseeding did not prove to be very effective on the steeper slopes.In the future, when treating steep (greater than 30 degrees), unstableslopes with hydroseed, the use of a blanket type organic material, suchas jute, is recommended for erosion control. This would allow for extendedcontact time between seed and native soil, which would then increase thechances of seed germination and establishment. Soil analysis of the hydroseededareas revealed contradictory results, with no clear connections being madebetween soil properties and plant propagation. Only a generalized assumptionthat raising the carbon:nitrogen ration in the soil increases native plantspecies growth can be made.

The use of a blanket type organic material appears to be the most successful means for erosion control, and is recommended for use on steep and unstable slopes. The blanket-type materials used in this experiment included jute, curlex and straw mats. The choice of material to use at a restoration site should be dictated by economics and availability of product. Loose mulches, such as cocoa, compost, blown straw and hydroseed should not be used by themselves as a means for erosion control. Punched straw has the good attribute of longevity, but does not provide enough ground coverage to slow down the erosion process significantly. Coir fences should not be used as a means of erosion control on steep or unstable slopes. Vegetation, though not providing for short term erosion control, should be viewed as a long-term biological erosion control method and used in conjunction with other methods previously mentioned. The use of straw bales dams in large gullies proved to be an excellent tool for stopping the erosion process, even during heavy rainfall events, and is highly recommended.

Since 45 percent of the slope remains free of vegetation, ongoing erosion control is still necessary. The use of blanket-type materials is not recommended at this point as the native shrubs are slowly filling in the slopes, thus making installment difficult. The main cause of erosion at this time would be by water collecting on the upper and lower shelves, then rushing down the slope and forming erosion gullies. Baccharis pilularis grows very well by cuttings. Placing cuttings of B. pilularis in the gullies in conjunction with straw bale dams would slow down the erosion process. Also, the use of catchments and straw wattles could be used to control the amount of water which collects on the shelves. This would help to prevent a sudden influx of water from rushing down the slope. On the bare areas of the slope, hand seeding followed by a light application of wood mulch could be an adequate erosion control technique. Species selected for seeding should be those which are scarcely present at the site and are important for the re-creation of Palos Verdes Blue butterfly habitat. These species include Artemisia californica, Astragalus trichopodus, Baccharis pilularis, Encelia californica, Ericameria palmeri and Salvia mellifera.

The weedy nonnative species (Brassica nigra, Centaurea melitensis, Chrysanthemum coronarium and Salsola tragus) continue to be problematic. Future attempts at eradicating these weeds should be continued until the native shrubs can adequately out-compete them. Many of the weeds are entering the site from the surrounding Navy lands open area which is currently being used as an off road vehicle area by trespassing civilians. This area is highly disturbed and is conducive to the growing of nonnative species. The area surrounding the work site is dominated by exotic annuals that provide an immense source of seeds that are blown onto the restoration site to then germinate. As a temporary solution, a screen could be placed on the fence to help minimize the passage of seeds into the restoration site. However, this is only a stop-gap measure and for a more permanent long term solution, the adjacent area of Navy lands should be closed to all off road vehicles and become a restoration site for the reestablishment of Palos Verdes Blue Butterfly habitat.

Planting container grown native plant species, is a good tool for restoration.Although more expensive than seeding, more control can be had over numberof species introduced to the site and planting location. Complementingthe container grown plants with the nearby salvaged plants increased theoverall number of plants, at a fraction of the cost for container plants.Where possible, the use of transplanting salvage plants is highly recommendedas a restoration method. No sightings of the Palos Verdes Blue Butterflywere seen on the A. trichopodus planted at the site; however,the new generation of seedlings that should appear from the seed bank madeby these first seedlings may help prompt a recovery of this endangeredspecies. Sightings were made, however, of the California gnatcatcher usingthe restoration site. Future research should look into why A. trichopodusdoes not do well as a transplanted species, yet healthy stands of the plantare found elsewhere.

The results of the restoration project conducted at Palos Verdes appears to verify the ability to reestablish coastal sage scrub habitat for the endangered Palos Verdes Blue Butterfly. The restoration site has met the success criteria established by the Fish and Wildlife Service two years after initial plant establishment/seeding efforts were accomplished. There is no reason to believe that, with time and active management of the site, percent cover of native coastal sage species will not continue to meet or surpass the success criteria established by the Fish and Wildlife Service. It has been our experience, through a number of previous coastal sage scrub restoration projects, that once a restoration effort has successfully survived the first two years, the site is able to survive and develop with minimal active management. In the case of the Palos Verdes site, due to the adjacent highly disturbed area, such active management will undoubtedly include weed and exotic species control until the disturbed area is itself restored.