In desert species, the quality of seed is highly variable from year toyear and must be evaluated before collecting large quantities of seed. AtFort Irwin, this was very true for collections made of Atriplex canescens.Atriplex collected during September 1996 was immature and had lowerviability, confirmed by X-ray analysis completed by the Seed Lab operatedby the California Department of Forestry and Fire Protection (CDF&FP Lab).Test results confirmed later collections made in October were of betterquality.
Often both inexperienced and experienced seed harvesters are deceived whenseeds are aborted early in development. Again a quick X-ray evaluationcan confirm viability before large collections are made. Atriplexpolycarpa proved to be instructive in this situation. Collections ofAtriplex made during September 1997 proved to be only tissue with noviable seed. Because seed production is unpredictable and some years maybe especially bountiful, X-ray evaluation can also be used to determine ifseed is of exceptional quality and if extra effort should be spent onadditional collections. June 1997 proved to be a very good year for seedproduction by Eriogonium fasciculatum in the area just north of GranitePass. Normally seed viability for Eriogonium is averages between 5 and10%. Seed collected during June was found to have a viability of 79%.Had a subsample been tested and received immediately after collection, itwould have been worth making additional collections because the seedquality was so high.
The timing of seed collection can be crucial for desert plants. Ripe seedmay be available for only a few days before predators and high windsdissipate it. At Fort Irwin this is true for Larrea tridentata wherecontinuous monitoring was necessary to determine seed maturity before thelight seeds were dispersed by wind. Seeds which ripen and fall quicklycan be collected by placing the seed head in a section of nylon stocking,cheese cloth, or netting. This has worked very well for collectingocotillo seeds. Young and Young suggest that clipping twigs withfruits are attached, just prior to maturity, is one means of harvesting.The remaining stem and leaf tissue is able to provide enough energy tofinish seed development. At Fort Irwin this may be useful to obtain seedfrom rare species or from individual shrubs in distant locals not likelyto be encountered again. A variation of this method was used to collectseed from Bebbia juncea . Immature seed heads very close to maturity, butunopened, were clipped off. Bebbia, like dandelion seed, is difficult tocollect, because the seeds are light, filimentous and disperse with thelightest breeze. The seed heads were allowed to air dry in uncoveredtrays. During drying, the heads opened and the seed was liberated. Asimilar phenomena was observed for both Encelia farinosa and Enceliafrutescens. Harvesting both species in this manner can greatly speedcollection.
Several different collection methods were also tried at Fort Irwin. Amodified weed wacker with a bag to catch and hold seed was tried onAtriplex canescens. This commercial implement, designed for use onprairie grass species, was less effective than collecting by hand. Forboth Larrea tridenta and Ambrosia dumosa a bucket into which seed couldbe brushed or shaken by hand proved to be a useful though slightly painfulmeans of harvesting. On one occasion, a small portable shop vacuumplugged into a generator was used to harvest Ephedra viridis seed. Thisproved reasonably effective because large quantities of seed could beremoved from the ground. One draw back to this method was that theharvested seed required screening to separate it from soil and debris.Another was the weight of Ephedra seeds, because they were so large andheavy they often clogged the vacuum hose. This method probably has thebest potential when used with small and light seed like Larrea tridentata.Another method that proved useful for Ephedra seed while it was stillattached to the shrub, was spreading a sheet around the shrub and knockingthe seed off on to the sheet by hand. This method also proved useful forcollecting Atriplex spp. and Lepidium fremontii as well.
Seed gathering from the soil surface is tempting, but is usually of lowquality, requires excessive cleaning, and is more commonly contaminatedwith pests and disease. Despite this sound advice,we succumbed to the temptation and collected Larrea tridenta seed that hadfallen and accumulated in roadside depressions. The result waspredictable. Seed collected from the ground had a lower viability, 44 %,compared to 70 % from seed harvested by hand.
Some of the desert species appear to produce stress crops with more,higher quality seed. When it is too dry however, they may produce littleseed and supplemental water may be helpful. Native stands of Jojoba(Simmondsia chinensis) produced many more seeds when provided withsupplemental water from micro-catchment basins.Once additional stands for seed collection have been identified at FortIrwin, supplemental water sprayed on via a water truck or by constructingmicrocatchment may be worth considering to increase seed production.
Processing
After harvesting, seeds often require cleaning and processing. All seedcollected from Fort Irwin was air dried in flats at room temperature for aperiod of two and four weeks depending on seed moisture. This wascompleted prior to storage and analysis to reduce the possibility offungal infestation during storage and permit any green seeds to ripen.Where possible seed was kept in paper bags prior to cleaning to promoteair circulation. In only one case, a bag of Atriplex canescens that wasair dried for just a little over one week, was there evidence of mold.Care must also be taken to protect personnel against biting insects duringprocessing. Desert gnats present in Ambrosia dumosa collected at FortIrwin were brought back to San Diego State University. After the seed haddried and was being packaged for cleaning and analysis, the gnats emergedand inflicted countless painful bites on everyone present. This seed waslater fumigated at the CDF&FP seed lab after a repeat incident.
Seed used for direct seeding does not necessarily need to be ascarefully cleaned as smaller quantities stored for growing containerplants. Young and Young recommend screening twice. Once to removelarger debris and once to remove debris smaller than the seed of interest.Screening, along with analysis for purity and viability, may be all thatis necessary for seed to be used in direct seeding. Typical lowtechnology methods of processing bulk seed other than hand screening mayinclude rubbing on concrete and running over seed with a car to break openseed coats and pods. Originally SERG processed all of its own seed byhand (i.e. cutting off pods) but now relies on vendors as much aspossible. The expertise and equipment at the CDF&FP Lab in Davis hasprovided many lessons about the value of mechanical processing. Handseparating Prosopis seeds from pods can produce several hundred cleanseeds per hour, compared to many thousands of seeds per hour using amodified meat grinder. Because of their high sugar content, the fleshypods must be thoroughly dried or thoroughly wetted before processing.
Some types of desert seed can also be placed in a mesh bag (or a plasticbag with a few small holes in it) and compressed air jets used to blow thewings off seeds. In the past we have done much of ourseed quality improvement by hand sorting, but if possible commercialfacilities air separators and sorting equipment are used.
We have learned, the hard way, to test every seed batch for purity andpercentage of sound seed after processing. If the seed is to be storedfor long periods of time it may be wise to test periodically duringstorage, and again just before sowing. We have learned through backtesting of failed seeding experiments (some costly in terms of time) thatpoor seed quality was probably the cause.
Percentage of sound seed can be determined by different methodsdepending upon the species. Many can be evaluated by eye, but if possible seedshould be sent out for X-ray evaluation, a non-destructive method ofassessing seed fill and potential viability. By combining X-rays withcutting tests it is possible to quickly determine seed quality.Determination of desert seed quality can be challenging because even dryseeds are sometimes viable.
While tetrazolium tests can be very fast and effective on some species,lack of uniformity of staining, failure to detect seeds that willgerminate abnormally, and difficulty in interpreting different degrees ofstaining are important drawbacks. Tetrazolium stains have proven to beuncertain on seeds with very effective dormancy -- which includes mostdesert seeds. Staining may work better on these seeds followingstratification.
The most reliable method of determining potential germination is togerminate a representative sample of the seed lot, but this is onlyeffective when the stratification and scarification requirements arerelatively well understood. This is not the case with many desertspecies. The proper germination temperature is species specific and needsto be determined by experimentation before running germination trials formeaningful results.
X-ray analysis and cleaning was carried out on all large batches of seedcollected at Fort Irwin. Smaller batches of seed collected on severaloccasions were consolidated for testing and cleaning. Seed collected fromthe following species was tested and cleaned: Ambrosia dumosa, Ephedraviridis, Atriplex hymenelytra, Senna armata, Larrea tridentata, Atriplexcanescens, Petalonyx thurberi, Isomeris arborea, and Eriogoniumfasciculatum.
Storage
Brucids, a boring insect, can persist within seed for many years slowingdestroying seed viability as it feeds. Brucids can usually be controlledby drying infected pods immediately after collecting and then freezingthem to kill the larvae. When possiblerefrigerate seeds, but space is often at a premium and seeds must bestored at room temperature. This does not always work and largequantities of seeds can be lost to insects and rodents. Storage of dryseed in a sealed plastic containers offers some protection, though storagein an ice chests, or a sealed metal cabinet offers better protection.
We suggest using insecticides as necessary. One seed vendor dips seedis a 20% solution of Malathion in water followed by a drying period and asubsequent dusting with 5% Sevin. Kay et al.used Phostox (aluminum phosphide) to protect the seeds of Mojavedesert shrubs.
Desert species appear to be very susceptible to fungi and diseasebrought on by the presence of moisture. They have adapted to growing insoil which has been sterilized by solar radiation and heat. Seeds can often be sterilized by soaking in a 40% solution ofhousehold bleach in tap water (2 parts bleach in 3 parts tap water) forten minutes, then rinsing thoroughly in running water for at least 48hours. Unfortunately, bleach can be phytotoxic. Use of a long runningwater rinse (48 hours) may reduce levels of pathogenic fungi to a similarextent without affecting seed viability. For the runningwater rinse, a shower head can be placed in the bottom of a bucket, and anaquarium bubbler used to increase the concentration of dissolved oxygenand promote water circulation.
Although desert seeds are often long-lived and may exhibit multipledormancy, many seeds have their best germination potential at the momentthey reach maturity. Proper storage conditions are critical to maintainseed viability over an extended period of time. Post drying moisturecontent of seeds from a variety of Mojave desert shrub species ranged from1.5%-9.4%.
Breaking seed dormancy
Seed dormancy is ecologically important in desert species. It can be adifficult obstacle to revegetation efforts where prompt, uniform, andcomplete germination is desirable. Impermeability of the seed coat orprotective fuzz and jackets are probably the most common form of seeddormancy. Many desert species are adapted to scarification in floodevents or wind abrasion. Small batches of seed can be scarified by hand,using a file or razor knife to make a nick or slice in the seed coat.Seed can also scarified by sanding, but have shown increasedsusceptibility to fungus and mold, presumably from the presence of thesmall particles of seed coat left after sanding.
Boiling water dips or soaks can often provide similar results. Lippittprefers briefly dipping seeds in boiling water. The boiling waterdip is safer than acid scarification, and seed coat thickness is less of aproblem. The same length of treatment can be used from year to year whichis risky with acid.
Many desert seeds require exposure to either high or low temperaturesbefore being placed in conditions favorable for germination. Many desertseeds have water soluble growth inhibitors that prevent the seed fromgerminating unless there is sufficient moisture to establish a seedling.If all of the inhibitor is not washed away at once, the seed produces moreinhibitor. Graves et al. found that Ambrosia dumosa(Mojave desert seed sources) seed germination was improved bystratification in activated charcoal or moist sand at 36¡F (2¡C) for 30days.
Other species may require high temperature 120ûF (50¡C) stratificationrather than the low temperature/moist stratification commonly used (Capon& Van Asdall, 1966). If a seed lot is to be planted during the sameseason that it is collected, it may be beneficial to try high temperaturestratification. If the seed needs to be stored for longer periods oftime, lower temperatures may be desirable.
Conclusions
High quality site-adapted native seeds are essential for successful revegetation or restoration projects. Project planning must begin with sufficient lead time, often one year to a year and a half, to procure suitable seed or a local seed bank must be started and maintained. Beginning and maintaining a local seed collection is an investment in time and effort, but the payoff is in high quality site adapted seed that will increase the probability of revegetation establishment and subsequent reseeding. The Arc View database created by SERG containing species and collection location data will facilitate future seed collection efforts at Fort Irwin. When integrated with additional seed collection sites, a suite of collection locations for each species can be monitored annually and used to collect larger quantities of seed with a diverse genetic makeup. Tracking seed viability over time and collecting and storing additional quantities during good years will help revegetation efforts when seed cannot be purchased. The seed harvested this past year will supply Fort Irwin with a diverse collection of local species, many unavailable for purchase, in quantities suitable for nursery stock production. Already much of the seed is being grown for use at Fort Irwin this spring. Cleaning, and upgrading seed before storage can improve the storage life, increase germination, and make production and planting easier. If seed collected at Fort Irwin is to be used for direct seeding, then it is well worth the extra time to send a sub sample off for X-ray analysis to confirm viability before time and effort are invested in collecting large quantities. Multiple dormancy is common in the seeds of many desert species and experimentation is often necessary to determine the best way to break seed dormancy. This can be complicated by year to year and plant to plant variation. Seed lots should be tested for viability before storage, periodically during storage, and prior to sowing.
