last update April 22, 1998
Arid and semi-arid ecosystems have evolved to survive under difficult and adverse conditions. The biological adaptations required to survive extreme drought and temperature differ from those that permit disturbance adapted species to readily regenerate following environmental change. In the desert, natural regeneration is confined to years with unusually high precipitation (In the southwest United States this is associated with the phenomena known as El Nino). In the Mojave desert, the window for regeneration is further narrowed by below freezing temperatures during winter. Sufficient moisture is available for plant growth and seed germination when total monthly precipitation exceeds average monthly temperature. For the Mojave Desert, sufficient moisture is available only during the months of December and January. However, during December and January night time temperatures routinely dip below freezing, killing any tender new growth.
Thus, it is not surprising that establishment in this severe environment is naturally slow and disturbance can make conditions for plant establishment many times more difficult. Disturbance often begins when human activity disrupts and strips the soil surface of both plant cover and the thin cryptogramic crust, composed of fungi and algae, that forms on the soil surface. With these natural soil stabilizers removed wind and water erosion can remove the thin and fragile top soil. The dust released from these exposed soils often releases particulate matter smaller than 2.5 microns, posing a significant threat to human health. Alternatively, soil compaction by wheeled and especially tracked vehicles can reduce water infiltration rates, slowing plant recolonization and speeding soil erosion and the formation of gullies. It is little wonder that desert areas may take centuries to recover from human disturbance and not surprising that individual tank tracks dating from World War II training maneuvers are still visible today. Without active intervention and restoration efforts many areas may never recover within the average human life span
As a result of ground maneuvers, live fire exercises, and force on forcemilitary trainingactivities much of the vegetation cover on National Training Center (NTC)and Fort Irwin,located 30 miles north of Barstow California, has been diminished or inmany areaseliminated. Within the next five years, many areas will requirerehabilitation in order tomaintain training realism and improve health and safety factors. TheIntegrated TrainingArea Management (ITAM) program is charged with maintaining and improvingthe qualityof training lands on post. Based on the results of demonstrations andstudies in progress,the ITAM program will choose revegetation techniques to be employed in alarge scalerestoration program. This rehabilitation will include seeding of disturbedareas with nativeplants, visual blocking of excess trails to prevent additionaldisturbance, physical soiltreatments to improve water infiltration and reduce soil erosion, and theestablishment ofwindbreaks and tactical deception islands.
The objective of this project was to develop, implement and monitormanagement/revegetation plans for five different sites to serve as modelsfor future landrehabilitation projects. These projects include the creation of tacticaldeception islands athills 696 and 780 (hills site), installation of windbreaks along roadsidesat the intersectionsof three and four corners (three corners site and four corners site),trail closures on the westside of Bicycle Dry Lake (bike lake west site) and berm removal and areastabilization onthe east side of Bicycle Dry Lake (bike lake east site). In addition, twoinvestigative experiments were initiated, one investigating threeirrigation treatments at the hills site and second investigating theefficacy of the commercial irrigation supplement, Driwater(tm), at thebike lake east site.
Results
Overview
To describe any generalities that exist between revegetation success andsoil condition at thevarious revegetation sites, survival data was collected and analyses ofsoil chemistry andphysical properties was performed. To help describe the belowgroundmicrobialcommunity for bacterial and fungal soil populations, europium staining wascarried out.Europium staining was performed twice, once in February 1997 and again inOctober1997. The vegetation survivorship and soil analysis results reported hereare from a finalsampling carried out in October 1997. To determine if significantdifferences exist betweensites and/or experimental treatments, one way analysis of variancestatistical computationswere performed using a 90~o confidence level ( oc< 0.1). Where there was asignificantdifference, least significant denominator (LSD) multiple comparison testswere computed.
Vegetation
In October, overall survival of the 1518 container plants installed over the entire project was 67%. At the bike lake east site, a total of 136 plants were installed and 77 were alive in October. Percent survival at the bike lake east site was 57% nearly 9 months after planting. At the bike lake west site, 266 plants were installed and 206 plants were alive in October. Percent survival at the bike lake west site was 77% approximately 5 1/2 months after planting. At the hills site, 425 plants were installed, and 346 were alive in October. Percent survival at the hills site was the highest of all project sites at 81%, 6 1/2 months after planting. At the three corners site, 286 plants were installed and 130 were alive in October. Percent survival at the three corners site was 45% 5 months after planting. At the four corners site the southern planting berm was planted first. Here 141 plants were installed still 58 were alive in October. Percent survival on the southern berm was the lowest of all the sites at 41% almost 5 months after planting. At the four corners site northern berm, 104 plants were installed and 89 were alive in October. Percent survival at the four corners north berm was 86% almost 3 three months after planting.
To determine which of the native species used had the highest outplanting success and survival. Percent survival was calculated for each species across all sites and treatments. Ambrosia dumosa was planted at all five sites with percent survival of 72%. Larrea tridentata was also planted at all five sites with an average percent survival of 72% also. Hymenoclea salsola was planted at all five sites, with a percent survival of 65%. Ephedra nevadensis was planted on three sites and had an average percent survival of 41%. Isomeris arborea was planted on three sites with a survival of 54%. Eriogonum fasciculatum was also planted at three locations and had a percent survival of 49%. Prosopis flandulosa was planted at two sites and average 62% survival. Atriplex canescens and Encelia farinosa were both planted at only one site each, and had a percent survival of 77% and 62% respectively.
Hills lrrigation Experiment
One way ANOVA were calculated for each of the six species used in theirrigationexperiment. Out of the six species only Ephedra nevadensis had asignificant treatmentresponse (p<0.10). LSD multiple comparisons test revealed survival on theperforated pipetreatment was significantly higher than the deep pipe treatment (p<.05)and almostsignificantly higher than the hand water treatment (p<0.11). Ambrosiadumosa showed atrend (p<0.17) with the hand water treatment slightly higher than theperforated pipetreatment.
Driwater Experiment
A one way ANOVArevealed that there was a significant treatmenteffect. ( p<0.02) between the number ofDriwater cartons and survival of Prosopisglandulosa (mesquite) seedlings planted in thebike lake east Driwater experiment. LSDmultiple comparisons tests revealed that theseedlings planted with two cartons hadsignificantly higher survival than the seedlingsplanted with zero (p<0.003) and one carton(p<0.04) and a tendency toward higher survivalthan the seedling planted with three cartons(p<0.12).
Soil Analysis
Major soil nutrients (nitrate, phosphorus and potassium), soilmicronutrients and cations(sodium, magnesium and calcium), soil pH, conductivity, organic matter,texture, bacterial population counts and fungal hyphal lengths wereanalyzed for each site. To determine ifsoil properties were significantly different at each of the project sites,one-way ANOVAwere calculated for each soil analysis. Three replicate soil samples werecollected from eachsite. The data reported are the mean and standard error of the threesample replicates.Based on field observations it was suspected soil on the southern threequarters of the bikelake site were different from the remaining upland portion, so two sets(three replicates perset) of soil samples were taken from the bike lake east site. One set wastaken from theupland portion bike lake east, called bike lake east north, and one fromthe remaining threequarters of the site on the edge of the dry lake, called simply bike lakeeast.
Soil nitrate was significantly higher in the soil samples taken from thebike lake eastsampling location than any other site (p<0.01). Soil nitrate here was 510ppm. At the hillsite, the next highest, it was 6.3 ppm and it was lowest on the bike lakewest site,averaging 3.6 ppm. Soil phosphorus did not significantly differ betweensites, but washighest on the hills site averaging 12.3 ppm bicarbonate extractablephosphorus.Potassium was significantly different at all sites (p<0.01) with theexception of the bikelake east north sampling location and the hills site. Soil potassium washighest at the bikelake east site, averaging 701 ppm, and lowest at the bike lake northsampling location andthe hills site, averaging 131 and 163 ppm respectively.
For the less important soil micronutrients and cations, sodium wassignificantly higher onthe bike lake east site at 2742 ppm than all the other project sites,ranging from 22-127ppm (p<0.01). Magnesium was significantly higher (p<0.01) on the bike lakeeast andfour corners sites, 176 ppm and 195 ppm respectively, than the rest of thesites, whichranged from 75 to 130 ppm. Like magnesium, calcium was significantlyhigher (p<0.01)on the bike lake east and four corners sites, 5195 and 3721 ppm, than therest of the projectsites, 1141 to 2450 ppm.
Percent soil organic matter was significantly higher on the four corners, bike lake west, and bike lake east sites than on the hills, bike lake east north, and three corners sites (p<= 0.01). The highest percent organic matter was at the four corners site averaging 0.86% and lowest on the hill and bike lake east north sites averaging 0.33%. Soil pH was significantly lower at the bike lake east site than all the other sites (p<0.01). Soil pH was the highest and lowest at the bike lake east site ranging from 8.5 at the upland sampling location, bike lake east north, and the lowest at 7.3 at the dry lake edge. Electrical conductivity, a measure of soil salt concentration, was significantly higher at the bike lake east site, at 15.533 mmhos/cm, than all the other sites ( ranging from 0.76 mmhos/cm at the three corners site to 0.30 mmhos/cm at the hills site) (p<0.01).
The percentage of silt and clay were significantly higher at the bike lakeeast and fourcorners sites (p<0.01). The bike lake east site averaged 17 % silt and 24% clay. The four corners site averaged 7% silt and 18% clay.Alternatively, the three corners site and bikelake north sampling locations had the highest percent sand (and lowestsilt and claycomponents) averaging 98% at both sites.
Europium analysis
Europium is a rare earth element that fluoresces under ultraviolet lightand has an affinityfor living and recently dead DNA. When used in combination withfluorescencemicroscopy it can be used to estimate living and recently dead soilbacteria and fungalpopulations. To estimate bacterial and fungal activity at the projectsites, europium stainingwas carried out twice - once in February and again in October 1997. DuringFebruaryonly the three corners, four corners and bike lake east sites wereavailable for sampling.During October all the project sites and both the bike lake east north andbike lake eastlocations were sampled.
Both fungal and bacterial activity and populations are directly related to soil moisture, so it is important to consider percent soil moisture when the samples are taken. Percent soil moisture was not significantly different between the project sites during the February sampling. During October, soil moisture was significantly lower at the bike lake north sampling location (p<=0.05). Soil moisture was also significantly higher (p<=0.05) at the four corners site. In February fungal- hyphal length was significantly higher (p<=0.05) than in October.
During the February sampling hyphal length was significantly higher at the three corners site(p<=0.01). In October hyphallength at the bike lake east northsampling location and bike lakewest was significantly higherthan the remaining sites (p<=0.5).However, hyphal length was notsignificantly different betweenbike lake east and the bike lakewest sampling locations. Bacterial counts inFebruary were significantly lower at the bike lake east site (p<=0.1).During Octoberbacterial counts were notsignificantly different.
Discussion and Conclusions
The conclusions that can be drawn from the vegetation survivorship, soiland microbialanalyses are anecdotal and only descriptive because our primary goal wasto produce arevegetation product based on the resources available at each site and atthe time ofplanting, not a careful and controlled experiment between sites. Care mustbe taken toavoid drawing any more than the broadest conclusions because vegetationsurvivorshipprobably integrated a wider range of variables than those discussedhere.
Survivorship was the lowest, at 41%, on the south berm of the four cornerssite. Thesingle most important factor was probably the time of year that the sitewas planted, lateMay. At that time of year extremely high temperature and very low humiditycan quicklyreduce xylem water potentials below what the underdeveloped root system ofmosttransplanted seedlings can supply - even when there is sufficient soilmoisture present. Thesoil at the four corners site probably contributed to the lower survival.The soil texture atfour corners site was very similar to the soil from the bike lake eastsite in that they bothhave very high silt and clay components, this probably made rooting moredifficult as thesoil began to dry out. Once dry, the soil at the bike lake east site hadthe consistency ofcured concrete.
Survivorship was much higher on the north berm at the four corners site at86%. The northberm was planted in early August. The only other difference between thenorth and southberms, other than the date they were planted, was the addition of 20 cmlong deep irrigationpipes. These deep pipes increased irrigation efficiency by deliveringwater directly into theroot zone. The deep pipes are probably the most important factorresponsible for thedifference between the two plantings. However, rainfall in Septemberprobably was animportant contribution to survival. The month of September received anunusual andsubstantial amount of rainfall. This storm system,even if it did not drop rain at the four corners site, undoubtedly loweredtemperatures andraised the relative humidity on site contributing to higher survival.
Survivorship was second lowest at 45% on the three corners site. Both thethree cornerssite and four corners south berm were planted in late May . Like the fourcorners southberm, the time of year the site was planted probably had the mostsignificant impact onsurvival. Anecdotal observations at three corners site suggests that howplanting holes aredug is probably important to survival. The two berms that were augeredwith the tractorappeared to have a higher survival rate than the two berms that could notbe augeredbecause of the buried fiber optic cable. This is probably due to thegreater area ofdecompacted soil present in the holes that were augered.
At bike lake east site the mesquite were planted in November 1996 and theremaining plantsin February 1997. It is disappointing that survivorship was not higher -because the sitewas planted early in the growing season to permit seedlings establishmentbefore the arrivalof hot weather. The soil at bike lake east site was probably responsiblefor the poorsurvivorship. Nutrient analysis, conductivity, pH, texture, organic matterall suggest thatthe majority of the site was located on the dry lake bed itself andtherefore not suitable forrevegetation because of the extremely high salt concentration and heavyclay texture. Soilanalysis indicate that the upland portion of the bike lake east site wassuitable forrevegetation work and accordingly success was higher there. Unfortunatelythe Driwaterexperiment was setup on the southern three quarters of the site where thesoil wasunsuitable. Still, it is interesting that plants with two Driwater cartonssurvived better thanthose with zero, one and three. This is consistent with observations madeat some of ourother restoration sites. In general, we have observed that Driwater doesnot encourage rootexploration into the surrounding soil, but instead encourages localgrowth. Most desertshrubs have a broad extended root system that is capable of scavenging lowsoil moistureover a large area. The two Driwater cartons were placed opposite eachother, on either sideof the root mass during installation. Perhaps this placement of the twoDriwater cartonswas important in encouraging lateral root exploration. While Driwater maykeep containerplantings alive in the short term, it may do so at the expense of longterm survival.Alternatively the seedlings installed with two Driwater containers mayhave run out ofDriwater at a critical time window that facilitated root exploration. Thiswindow ofopportunity may have been missed by seedlings installed with one and threeDriwatercartons.
Survivorship at the bike lake west site was 77% and would have been higherif ninevegetation islands had not been washed away when the site flooded inSeptember 1997.These islands were placed on the southern edge of the site in the path ofa trail/wash. Careshould be taken in future trail closure projects not to place anyplantings in washes ordrainages where flooding may occur. A better option is to use largeboulders to blockaccess, not plants. Construction of large basins around the vegetationislands permitted theapplication of larger than usual quantities of water, 20 to 25 gallons perbasin which wasprobably the main reason for the higher survival. The close proximity tothe cantonmentarea and an overhead water source made application of such largequantities feasible. Asimilar scheme would not be feasible at remote sites where the cost ofwater transport mightnot be offset by the cost of planting more container plants. It isinteresting to note that thevegetation islands in the northern portion of the project were fertilizedduring installationyet when these islands were sampled in October, they had the lowest levelof availablenitrate of all the project sites. One explanation for this could be themulch placed in theisland basins. Decomposition of this mulch may be responsible forimmobilizing theadditional nitrogen. Increased decomposition would also help explainhigher bacterial andfungal activity at the bike lake west site.
Survivorship was highest, at 81%, on the hills site. This very high survival was probably due to a combination of when the plantings were installed and the loose and uncompacted soil on the site. The hills site was planted in early April 1997 when soil moisture was very high and temperatures relatively mild. The soil on the hills site was loose uncompacted sand with the highest nitrate, 6 ppm, and phosphorus, 12 ppm, of any of the sites other than the extremely saline dry lake portion of bike lake east site. This undoubtedly facilitated deep rooting prior to the arrival of hot weather. Unfortunately, getting a water truck out to the hills site was problematic and the majority of the watering was completed by hand directly into the tree shelter This is the most probable reason for a lack of response to the deep pipe and perforated pipe irrigation treatments. Additionally, high winds at the site were responsible for filling many of the deep pipes and protective tree shelters with sand. This burial obviously contributed to mortality but may also may have interfered with the experiment.
Europium sampling in February revealed more fungal activity at the threecorners site thanfour corners and bike lake east during the same time interval. Thissuggests that the threecorners site may have been less disturbed or more likely lower hyphalactivity was aresponse to higher cation concentrations, like magnesium and calcium, atthe four comerssite.
In general all the sites had low microbial activity. During February threecorners had 750mm of hyphae/ gram of soil. This amount of hyphae is still indicative of adisturbed site.Dave Bainbridge found at Red Rocks Canyon State Park that restorationsites with less than1000 cm hyphae per gram of soil were in need of remedial measures. Themulchincorporated into the mounds at three corners may be a suitableremediation technique,however based on the October sampling this is unclear. Long term samplingis needed todetermine the effect of mulch on soil microbial activity. Low activity inthe Octobersampling compared to February is probably simply a seasonal response tothe harsh desertconditions and consistent with observations elsewhere. In quarterlysampling at RedRocks Canyon State Park microbial activity remained lower in October thanin February.During the elevated temperatures of summer soil fungi are probablydormant. Samplingagain at the three corners and bike lake west site in February of March of1998 or 1999would help determine if incorporating mulch into the soil is a suitableremediationtechnique.
On the project sites, it appears that soil moisture is less related tomicrobial activity than tothe amount of clay in the soils. It is likely that sufficient soilmoisture was present forbacterial activity. Though not statistically significant,the higherbacterial activity at the bikelake west than the three corners site suggests that leaving mulch on thesoil surface ratherthan incorporating mulch into the soil may help to create a more favorablemicroenvironment.
Recommendations
Based on the success of the basin island planting scheme at bike lake westsite, werecommend this scheme over individual plantings anywhere water is readilyaccessible. Theisland basins were easily filled using a fire hose and water truck. Theonly draw back isthe quantity of water required to fill the basins. By filling one basingit was possible toirrigate three to four plants at one time. Mulch placed in the basinsappears to improveirrigation efficiency by reducing surface evaporation. If mulches areused, a preplantfertilizer should also be added to prevent nitrogen immobilization andsubsequent lack ofnitrogen for plant growth. Any fertilization should be employed withcaution as the highernitrogen content in leaf tissue could cause preferential browsing and theeffects on longterm survival are unknown.
Based on the results at three and four corners, we recommend timingplantings to coincidewith natural rainfall. No amount of irrigation water can substitute fornatural precipitation.Large plantings should be confined to the months of December, January andFebruary.Also, based on the results of four corners, we recommend the use of deeppipes to facilitateand improve irrigation efficiency on roadside berms. These deep pipes willincreasewatering efficiency by reducing surface evaporation, put water where it isneeded in theroot zone, and also permit larger quantities of water to be applied at onetime.
We also recommend longer project cycles. Most large revegetation projectsare funded onfive year cycles. This long-term approach permits enough time to plan andgrow anappropriate selection of container species, and more importantly, takeadvantage of erratichigh rainfall years. In an arid environment where natural regenerationprobably occurs inonly one out of every seven years, long-term planning to take advantage ofhigh rainfallyears is critical.
We recommend focused restoration on the fringes of training sites thathave not beencompletely degraded. The hills site is an excellent example of howsuccessful rehabilitationcan be initiated on a fringe site that is not completely degraded.Rehabilitation of sites thathave been in use for extended periods of time and are severely degraded,like three and fourcomers, will take considerably longer, cost significantly more and run ahigher risk ofdamage from training activities. Rehabilitation of fringe areas will alsohave a higherhabitat value by serving as a buffer for less disturbed habitat.
The vegetation at Fort Irwin is not disturbance adapted. Instead, it isadapted to survivalunder harsh conditions. Rehabilitation and restoration at best can onlyaccelerate naturalregeneration. Research into how to best accelerate natural processes needsto take place.Avenues worth investigation may include: avoiding container plantingaltogether bygerminating and growing seedlings on site; how to best inoculate aseverely disturbed sitewith soil microorganisms; research of alternative irrigation techniquesand determining theappropriate amounts of water and fertilizer to apply to restorationprojects.
Research needs be completed to determine what restoration and revegetation techniques are appropriate for what site conditions. For example, the catchment berms at the hills site may not be particularly useful in catching water because of the high percentage of sand in the soil. Are catchments more effective and appropriate on sites with a higher clay content? Why were deep pipes at the hill site not particularly useful in augmenting survival but at the four corners site they were extremely useful? Questions like these need further research. Another avenue of research at Fort Irwin could be into "artificial vegetation assemblages or cover species" in areas with high training activity where most native species cannot survive. If realistic training is to continue, rehabilitation efforts must keep pace with degradation. Research to determine the most effective restoration techniques may be the best way to keep ahead.