The project, which began in the spring of 1996, was initiated by the National Resources and Environmental Affairs Directorate (NREA) at the MCAGCC as an effort to develop and test methods of native plant revegetation in disturbed areas while simultaneously controlling erosion and dust. A tank trail parallel to Del Valle Road, just to the west of that road and north of the Berkeley Avenue turnoff, had become unnecessarily wide and was increasingly encroaching on an adjacent area of undisturbed creosote bush scrub. The tank trail was divided in half lengthwise, and the western half, adjacent to the undisturbed vegetation, became the project site. The eastern half of the tank trail remains active and is separated from the site by a railroad tie barrier. The site is approximately 1600 m long by 50 m wide, running roughly north to south.
For the purpose of this report the tank trail site has been divided into two units. Unit one, roughly the southern half of the site, was plantedbetween March 1996 and May 1997. Unit two, the northern section of the site,was planted in October 1997. This report contains the plant survival data for allplantings, the results of the various statistical tests which were performed, andother pertinent data.
MATERIALS AND METHODS
UNIT ONE
In January 1998, the southern or island area of the tank trail site was replantedwith 100 plants, which were placed in shallow basins in the three previouslyplanted plots without supplemental irrigation equipment. Thedifferent irrigation methods which had been applied to these plants at theoriginal replanting in March and April 1997 were disregarded statistically asthey had become irrelevant over time due to the loss of straw mulch to windand because the five-gallon reservoir systems had proven ineffective in thelong term due to a lack of durability and a tendency to become clogged.
Germination data was also collected in January 1998 in the nine unplanted soiltreatment plots, in order to measure the success of the seeded plots and todetermine the relative effectiveness of the soil preparation techniques andamendments in promoting both seeded and natural revegatation. Within eachplot, ten randomly located one meter square quadrats were established, andall plants present were recorded by species, number and estimated percentcoverage of the quadrat. Twenty quadrats were recorded within each of thethree plots containing areas of added bark mulch; ten within barked areas andten without bark. The plants recorded were categorized as seeded natives,non-seeded natives, and non-natives. Analysis of Variance(ANOVA)was performed to compare ripped and non-ripped plots, seeded and non-seeded plots, and to do a three way comparison between imprinted, pitted, andcontrol plots. The mulched quadrats within plots 2, 6 and 10 were comparedwith non-mulched quadrats from the same three plots.
The mesquite mounds were replanted in November 1997 with approximately25 plants. Each plant was provided with a four inch wide deep pipe as wasdone previously on the mesquite mounds. Driwater was not used as wesuspect that it does not allow for optimal water dispersal, thus confining thegrowing roots to an inadequately small area of temporarily available moisture.
In January 1998 a small area within the mesquite mounds was utilized tocompare the survival of three native species which were germinated in theSERG greenhouse, each with and without native mycorrhizal inoculum.Species used were Hymenoclea salsola, Atriplex canescens and Enceliafarinosa for a total of 59 plants. Chi-square analysis was conducted tocompare survival of each species between inoculated and non-inoculatedplants.
The gravity feed area in unit one, consisting of four 55 gallon containers, hasnot been replanted or altered in any way since it was established in May 1997.ANOVA was used to compare survival between emitter-fed plants and capsule-fed plants.
Unit Two
This area was planted in October 1997, using a variety of irrigation methodsand a total of 615 plants. Starting from the southern boundary ofUnit one, the initial planting was a series of nine 55 gallon containers, each ofwhich supplied 24 plants with water, 8 plants on each of three sections of .irrigation tubing extending from near the bottom of each container (Figure 2).Each section of tubing was branched into eight outlets, each of whichterminated in a ceramic capsule which was buried near the root zone of eachplant as it was planted. The containers were arranged to provide water to eachcapsule through gravity-induced pressure. The intent was to provide water asefficiently and in as timely a fashion as possible, with the ceramic capsulesreleasing water through diffusion at a rate dependent on the aridity of thesurrounding soil. While the irrigation tubing was standard commercially-available material, the cartridges were manufactured by SERG using astandard bottle mold. The 55 gallon containers were placed close enough tothe active tank trail to allow easy access by a commercial water truck.
In the next section a total of 316 plants were planted using three inch diameter flexible perforated plastic pipe as the water delivery medium. Fifteen 100 foot long sections of pipe were buried in ten inch deep trenches with approximately 20 plants placed along each side of each pipe (Figure 3). Both ends of the pipe were elevated so that the pipe would hold water and release it through the perforations. The intent was to experiment with providing water below ground where it would be both needed and less susceptible to evaporation. These pipes were placed with one end adjacent to the active tank trail for ease of filling by a water truck.
In a natural basin area close to the southern end of Unit 2, ten Chilopsislinearis (Desert willow) were planted without supplemental irrigationequipment. This is a wash inhabiting species that tends to do relatively well inlow-lying areas which collect water. In an extension of this idea, a group ofcatchments, roughly four meters square, were created near the center of Unit 2.Each catchment consisted of two berms forming a water collection point at thelow corner of each catchment, each with three or four plants for a total of 32plants. These were placed so that they were not only connected but wereprogressively lower as they moved further from the tank trail, creating a terracedeffect. This allows a water truck operator to fill all of them from one spot nearthe tank trail.
The last planting in Unit 2 consisted of 91 plants which were planted with avariety of irrigation devices, including deep pipes of both 2 and 4 inch widths,deep pipes with ceramic containers attached at the bottom, ceramic containersalone and controls which received no treatment. These plants were plantedabout a meter apart from each other in a roughly square plot.
All the tank trail plants were watered on a regular basis by a contracted water truck or, where necessary, by hand by SERG personnel. With few exceptions, the watering regime has been every three weeks, never exceeding one month between waterings except during the wet season(November through March). Survival was recorded in November 1997, April 1998 and July 1998. After the last survival recording in July 1998, a Chi-square test was performed to compare survival in the largest irrigation treatments, the perforated pipe plantings and the nine gravity feed systems, both of which were installed in October 1997.
Vertical Short Take-off and Landing(VSTOL) Site
In addition to the tank trail site, a nitrogen mulch interaction study was undertaken by SERG in March 1998. This study site was located at the Vertical Short Takeoff and Landing area (VSTOL) of MCAGCC, near the Sand Hill area of the base. This site is on an abandoned landing strip with runways and aircraft parking areas treated with Soil Sement, which forms a hard concrete-like surface. As part of an effort to revegetate this area, SERG designed an experiment on a portion of the main runway which had been ripped to a depth of 3 feet by base personnel and equipment. This experiment consisted of three identical 10 meter square plots, each containing four soil treatments, including alfalfa straw mulch, wheat straw mulch, sawdust mulch and a control. In addition, half of each plot was fertilized with Ammonium nitrate. This effectively created eight subplots, with each plot having four soil treatments, each of which had fertilized and non-fertilized sections. Within each of these eight subplots, four Ambrosia dumosa were planted. In addition, each subplot was seeded over a one meter square area with both native and exotic seeds.
The different mulch materials represent a gradient in Carbon:Nitrogen ratios.Hypotheses to be tested include the possibility that a particular C:N ratio willprove to be significantly more effective in promoting the development of micro-organisms, subsequently lowering the amount of available soil nitrogen. Thefertilizer is intended to simulate the atmospheric nitrogen deposition whichoccurs in polluted areas. The main hypothesis being explored is that theadditional nitrogen provided by atmospheric deposition in polluted areascreates an advantage for exotic species, which are often able to utilize an over-abundance of nitrogen more effectively then native plants, which have a morelimited capacity for nitrogen usage having evolved under low-nutrientconditions.
The data being collected includes soil nitrogen levels, both at the surface and below ground Nitrate and Ammonium levels will be measured in each sub- plot, using ion-exchange resin bags. Survival and height of the out-plantings will be measured, and germination in the seeded plots will be quantified. These data were recorded beginning in May 1998, and will be collected on a bimonthly basis for about one year. At the present time, the collected data are not yet available for analysis.
RESULTS
Unit One
Overall survival in the island plantings, Section 1 of the tank trail plantings, was 78%. ANOVA performed on the soil treatment plots revealed a significantly higher percent coverage of non-seeded native species in the ripped areas than in the non-ripped plots, with P = 0.03. There was a significantly higher number and percent coverage of seeded native plants in the imprinted plots compared to both the pitted and control plots, with P = 0.0001 for number and P = 0.0008 for percent coverage. There was also a significantly lower number of nonnative species in the imprinted plots than in the control plots, with P = 0.003.
Several significant differences were found between the plots with added bark mulch and those without mulch. These included a higher percent cover of seeded native plants in the non-mulch plots (P = 0.004); a higher number of non-seeded natives in the non-mulch plots (P = 0.006); and a higher number (P = 0.0001) and percent coverage (P = 0.0001) of nonnative plants in the non- mulched plots.
Survival in the mesquite mounds was 67%, with no significant difference in survival between the ripped south section (64%) and the non-ripped North section (69%).
47 of the original 59 plants in the mycorrhizal study survived, for an overall rateof 80% survival. A significantly higher number of inoculated Atriplex canescensplants died than did non-inoculated A. canescens, with P = 0.03.Survival in the four initial gravity fed planting areas, planted in May 1997, was64%. Survival was not significantly different between the emitters (66%) andthe capsules (62%).Unit Two
The gravity feed system with nine 55 gallon containers had a survival rate of81%, with 166 out of 206 plants surviving. The perforated pipe section had 274plants survive out of an original 316, leading to a survival rate of 87%. Acomparison between these two irrigation methods via Chi-square analysisdemonstrated a significantly lower survival rate for the 55 gallon containerirrigation system (P = 0.06). Although the survival rates were similar, therewere many more plants in the perforated pipe group.
Eight of the ten Chilopsis plants planted together in the basin area survived for a survival rate of 80%. 100% of the 32 catchment plants survived, and 79% of the group of plants with various individual watering mechanisms survived.
DISCUSSION
Unit One
The higher percent coverage of non-seeded natives in the ripped plots compared to the non-ripped plots was caused mainly by the seedlings being larger in size. The ripped plots probably developed larger plants due to their enhanced ability to retain water and windblown detritus and with the reduced soil strength found in the mechanically ripped plots, seedling roots were better able to push through the soil to acquire both water and nutrients. The number of plants might be restricted to a total relatively close to that of the non-ripped plots by the fact that the area available for optimal germination is restricted to the lowest points in the plot, which provide the most shelter and moisture to the seed. Thus, both types of plot contained about the same number of plants, but the ripped plots were able to support significantly larger ones. It is also possible that since many of the non-seeded natives were fast growing annuals, an individual plant in a physically favorable location may have grown quickly enough to out-compete other nearby plants, reducing the total number of plants in the ripped plots.
A similar situation may have existed in the imprinted plots, which producedhigher numbers and percent cover of seeded natives then both the pitted andcontrol plots. Imprinting creates a regular pattern of numerous smallindentations which may provide suitable germination conditions for a greaternumber of plants then pitting, which tends to create an irregular pattern offewer, larger indentations. The fact that non-seeded natives were not in greaterabundance in the imprinted plots is possibly an indication that the shallowindentations in those plots are not particularly effective in trapping windblownseed, although they were able to retain seeds deposited there by hand.
The bark mulch plots had no advantages over the non-bark plots, and actually had significantly fewer non-seeded natives, a lower percent cover of seeded natives, and a lower number and percent coverage of non-natives. This may be due to the recalcitrant nature of the bark, which probably takes three to five years to decompose, and has not as yet released into the soil significant quantities of the desired nutrient and fungal amendments. Advantages to native shrubs provided by recalcitrant mulch being added to disturbed soil normally do not appear until a large amount of the mulch has decomposed and entered into the mineralization cycle. In the meantime, the bark may actually reduce seed germination by physically preventing seeds from germinating due to size and bulk of the bark.
The five gallon gravity feed (diffusion irrigation) devices and the straw mulchtreatment used in the initial island plantings proved ineffective over the long-term. The basins which were constructed during replanting were relativelysuccessful, but require a large amount of water and may be too small to trapsignificant amounts of windblown seed and organic detritus and initiateadditional plant colonization.
The mesquite mound plants, once established with deep-pipe irrigation,seemed to do relatively well. Some of the older plants have achievedimpressive size and can probably be considered independent of additionalsupplemental watering. As stated above, Driwater seemed relatively ineffectivein supporting root development in young plants, although it seemed to promoteshort-term survival and prevent transplant-shock to some degree, as noted inthe initial annual report for this project. The four inch wide deep-pipes can holdabout two gallons of water and deliver it into the soil directly to the root level,well below the surface, reducing loss to evaporation. Plants with these devicesare probably encouraged to produce deeper roots and may ultimately have ahigher long-term survival rate then those with Driwater or surface watering.
Atriplex canescens, which had a significantly higher survival rate in non- inoculated plants then in inoculated plants, is a member of the non-mycorrhizal plant family Chenopodiaceae. The added fungal matter may have actually had a pathogenic effect on this species, casting doubt on the common assumption that mycorrhizal inoculum is universally beneficial.
The last treatment in Unit one, the four gravity feed systems, had a survival rateof 64% in June 1998. The two different types of water delivery devices, theceramic capsules and the commercial emitters, performed about equally aswell. These plants were planted in May 1997, when the weather had alreadybecome quite hot and dry. Initial survival was recorded in November 1997 andhas remained consistent since. It seems that the plants which survived theinitial transplant shock and the ensuing summer weather have survived quitewell. This water delivery system required few adjustments post-planting andmay be considered effective for this application, particularly when plantingoccurs earlier in the year.
Unit Two
The plants in this unit were planted in October 1997, which proved to be extremely favorable due to the El Nino phenomenon which occurred that winter. The area received abundant rainfall and after initial watering it was not necessary to irrigate these plants until April 1998. Eight of the ten Chilopsis plants survived, probably due to the moisture holding capacity of the basins they were planted in. The perforated pipe group had a significantly higher survival rate then the next most numerous group, the nine gravity fed plantings. This discrepancy might possibly be due to numerous leaks which developed in the gravity feed system when the containers were filled and pressure was applied to the system. Specifically, the connection between the ends of the irrigation tubing and the ceramic capsule were not strong enough, and leaks developed each time the system was used. Unfortunately, one leak is capable of draining an entire reservoir, leaving all 24 plants without water. Efforts to do field repairs ultimately proved ineffectual, and these plants have been individually watered since June 1998. This system would probably be quite effective after replacement of the faulty connection. The perforated pipe system was comparatively trouble-free and may be the most efficient system of those tested for delivering water quickly to a large number of plants. It does use a large quantity of water, however, which could be reduced by using non- perforated pipe and placing holes in the pipe adjacent to the plants. This would allow the pipe to fill faster due to the reduced number of outlets, and would direct the water to only where it was needed.
The catchment plantings were successful, although too few in number tocompare statistically with the other plantings. All 32 plants survived, probablydue to the large quantity of water which was used to irrigate them. This is themajor drawback to large-scale application of this technique. Unless significantquantities of water are readily available or it is used in areas where naturalrainfall is trapped and the plants are planted in the fall or early winter and cantake advantage of the rain, this system would probably not be economicalexcept for small scale projects.
The 91 plants which were planted together with various individual wateringdevices did relatively well (79% survival), again aided greatly by the wet winter.The control plants suffered about the same mortality rate as any of the othergroups, implying that the various treatments have yet to effect survival. Due tothe relatively subtle differences between these devices, it may take severalseasons for a difference in irrigation capability to become apparent. Results ofanalysis will be available in the next report.
Overall, the contract requirement of 60% survival one year post-planting will bemet. The abundance of native annuals which germinated in the plots duringthe spring of 1998 was encouraging, as they will add both to the soil seed bankfor future years and necessary nutrients to the soil to help establish a self-sustaining mineralization system to the project area that had been destroyed bythe impact of both wheeled and track vehicle use for many years.
RECOMMENDATIONS
Results from the Tank Trail experiments demonstrate the benefits of ripping acompacted area to enhance the site's ability to revegetate. By reducing the soilstrength through ripping, both water infiltration and soil mineralizationincreases and a plant's ability to move its roots through the soil improves. Theend result of such actions is what was found on the ripped section of the TankTrail; larger seedlings and increased coverage when compared to the non-ripped section of the Trail. The speed of the restoration effort is increasedthrough the physical effort of ripping the soil, even if no other actions arecompleted since even volunteer seedlings that become established from windblown seeds will flourish in the enhanced environment. It is highlyrecommended that any site suffering from heavy compaction, such as the TankTrail, be thoroughly ripped prior to any other restoration activity beingconducted. In certain situations, such as heavily used but narrow roadways,ripping may be all that is required with volunteer seedlings becomingestablished on the site from nearby seed sources.
Tank Trail results also seem to indicate that imprinting a large area inpreparation for direct seeding can greatly improve both initial germination andoverall success rates. The Tank Trail seeded/imprinted areas demonstratedboth higher numbers and greater coverage of native species than either controlor pitted plots. The shallow depressions formed by the imprinter apparentlyenhance a direct seeding effort, but do not seem to aid in the establishment ofnative volunteer seedlings. It is recommended that any restoration effort that isto rely on direct seeding as the primary source of revegetation include the useof an imprinter during site preparations. If the imprinting and seeding effort canbe timed to occur during a wet winter season, such as occurred in 1997-98,results should be excellent with a minimum of money and effort beingexpended.
There has not been enough time, as of yet, to accurately analyze the effects ofadding mulch to a restoration site. Though previous research has shownmulch to be of benefit to the reestablishment of native shrubs in slightly moremesic habitats, such as coastal sage scrub, the lack of rainfall and moisture atTwenty-nine Palms appears to have possibly slowed down the mineralizationrate of the recalcitrant mulch to where any benefit will not be apparent for quitesome time. As such, it appears that the one noticeable effect the mulch hashad on the Tank Trail restoration site has been to prohibit the germination ofseedlings, from both direct seeding and windblown seeds. Though it isrecommended that data continue to be collected on the mulch plots todetermine if the benefits may appear over a longer period of time, it is alsorecommended that recalcitrant mulch not be included with any direct seedingrestoration efforts since it appears to be counterproductive in this situation.
The mesquite mounds have been slow to develop. Mesquite seedlingssuffered heavily from initial die-off and were slow to become established. Theprimary reason for this problem is believed to be the timing of the moundsconstruction and planting. The first group of mesquite seedlings were plantedduring the La Nina season of 1996-97, with little subsequent rainfall. This notonly led to large die-off of seedlings, but little growth of those that survived. Ithas only been during the wet 1997-98 winter season that any growth has beennoted. Those seedlings that survived and have demonstrated a belated growthspurt have been the ones with deep pipe irrigation systems. Dri-water andsurface irrigation has not been successful on the mesquite mounds. It isrecommended that the mounds continue to be monitored to determine if thoseseedlings that survived and have shown growth over this last year will continueto develop into the type of well covered mounds that once thrived over much ofthe Twenty-nine Palms area.
The use of perforated pipe as a methods to deliver water to native seedlings has shown to be the most efficient and least expensive irrigation procedure tested so far. Survival rates were significantly greatly with perforated pipe than other methods, and the system has been virtually maintenance free. Though ceramic capsules appear to have excellent possibilities, problems with loose fittings and system leaks have yet to be overcome. Continued research should be done on the use of ceramic capsule irrigation systems since, with further experimentation, ceramic capsules may yet prove to be a cost efficient irrigation method for desert restoration; however, the use of perforated pipe appears to be the most successful method to date. It is recommended that for restoration efforts that allow a water source such as a water truck to be used, perforated pipe be selected as the means of irrigating the seedlings. The only change recommended is to use solid pipe and punch the drip holes into the pipe at selected locations near the newly planted seedlings. By doing this instead of purchasing pre-drilled pipe, water will be delivered only to the specific points where a seedling is located. This minor adaptation will improve the efficiency of the system even more.
The use of catchments for groups of plants appears to be a successful method of irrigation, but is not the most efficient use of water. Since catchments rely on surface watering, much is lost to evaporation during the hot summer months when water is most necessary. If water availability is no problem, and large amounts of water are added to the catchments on a regular basis, results on the Tank Trail demonstrate that the use of catchments leads to high seedling survival rates. However, due to the need for such large amounts of water, the use of catchments is recommended only for small restoration efforts located near an available source of water. If such conditions are met, the use of catchments can lead to a successful restoration effort.