Plant Protection

for plant establishment 
on arid and semi arid sites
Prepared for the
California Department of Transportation 
as part of the Desert Plants Project
David A.  Bainbridge
Biology Department
San Diego State University
San Diego, California 92181


Establishing plants on arid and semi-arid sites is very challenging. In addition to providing water and soil capable of supplying the nutrients required for growth (including necessary soil symbionts such as Rhizobia and Mycorrhizae if appropriate), the planter must provide protection from grazing, high winds, abrasian by blowing sand, mechanical wind damage, and temperature extremes (Bainbridge and Virginia, 1990; Bainbridge, 1991;Bainbridge et al., 1993).  This paper reviews the nature of these problems, results from field tests and observations, and suggests solutions.


Even when soil moisture and nutrients are favorable for plant establishment hervbivory may limit survival and growth on desert sites.  Herbivory is increasingly recognized as a critical factor in tree survival in arid environments (McAuliffe, 198; Bainbridge and Sorensen, 1990; Bainbridge and Virginia, 1990). Newly established seedlings are often the most succulent plants available and rodents, rabbits, reptiles, domestic livestock and insects can prove fatal to young plants unless adequate protection is provided.

Rabbits (Sylvilagus auduboni), blacktail jack rabbits (Lepus californicus), and rodents (presumably Citellus and Neotoma spp. ) have been the most significant herbivores on our trial plantings. 1993-1994 was reported as the worst year for herbivory in 30 years and tested systems to the extreme.  Grazing of many species has been severe and rabbits have browsed even old resinous creosote bush heavily, chewing stems up to 1 cm thick on mature plants at Red Rock Canyon State Park.

A series of field studies suggests that the order of preference is: mesquite, palo verde, bur sage, cheese bush, creosote bush, bladderpod and smoke tree.  While rabbits seemed to be the principle herbivore, tooth mark analysis on several sites suggests smaller rodents, such as rats, Neotoma spp., and squirrels, Citellus spp., may be important as well.  In some areas, feral animals can be very destructive.  Burros have eliminated Ambrosia dumosa from many thousands of acres of range by repeated, heavy browsing.  The seedlings of most species tested can survive repeated gnawing to within a half centimeter of the soil surface in they are on an irrigation regime, otherwise, browsing is often fatal.

High Winds

Protection from the wind can be critical in extreme environments (Virginia and Bainbridge, 1987).  High winds and blowing sand damage and kill plants (Mosjidis 1983).  Observations of eroded  buildings and utility poles in the Coachella Valley make it clear how abrasive wind carried sands are in the desert.  The potential for damage at a site can be evaluated placing a vertical piece of railroad chalk mounted on a metal pin at the ground surface.  This large diameter chalk clearly shows sand impact and maximum damage direction.

In addition to sand blast effects plants may be damage and killed by the mechanical action of high winds.  Multiple branching has been observed as a common response to damage from wind, either at the ground surface or at the top of wind shelter protection .  Young tree seedlings (with only cotyledons) have been blown out of the ground in some locations.  Sand carried by the wind also fills in plant collars and makes irrigation more difficult.

Moisture Stress

Drying winds also increase the moisture stress on young seedlings.  Potential evaporation measured near the Salton Sea was 3, 627 mm (Hughes, 1963), more than 40 times the annual precipitation.  Seedlings which emerged following a rain event in August 1989 became moisture stressed after about six months and then declined rapidly.  Protection can reduce evapo-transpiration and reduce moisture stress on seedlings.  This appears to be most critical in the first 6-8 weeks after outplanting.

Temperature and radiation extremes

Protection can minimize adverse effects from temperature extremes.  While freezing is not often considered in the low desert, many species are very sensitive to low temperatures, and freezing (Bowers, 1980) may limit plant distribution.  Freezing is not uncommon on winter nights with clear skies, but hard freezes occur only once a decade.  A particularly severe freeze in 1978 resulted in wide-spread damage and mortality for many desert species. Lows during this freeze included 10 F at Barstow, 25 F at Blythe, and 26F at Palm Springs (Lenz an Dourley, 1981).  Microclimate variation and cold air drainage can produce much lower temperatures than the climate station data would suggest.  Frost damage was observed in both 1989 and 1990 near Travertine Point where leaves of young seedlings were blackened and shriveled. 

Soil temperatures may reach more than 60 C in the summer in the low desert.  Soil temperature in the shade may be about air temperature unless the wind is blowing, when soil temperature may be elevated 4C above ambient (Wallace and Romney, 1972). High radiant loads increase the impact of high air temperatures, which may exceed 35C in any month in the low desert.

The options

Many strategies can be used to provide plant protection, including treeshelters, rock mulch, plastic or metal screens, plant collars, dead plants, repellent, straw stubble, or shade screens.  All of these may prove worthwhile for specific site problems and species (Figure 1). Their advantages and disadvantages should be considered and evaluated in the field before specifying their large scale use.  These options are listed in approximate ranking for use with tree seedlings on an typical exposed site in the low desert with windblast.


a. Tree Shelters

Many companies have introduced tree shelters in the last two years. These are commonly plastic tubes of various configurations and materials. The benefits and costs of using have generally been very favorable but many questions remain (Young, 1992; McReary and Tecklin, 1993;Windell, 1993;Sorensen, 1993;Bainbridge, 1994).

Commercially available models include:

Treessentials Supertube

Treessentials Supertubes are 3" diameter twin-wall translucent plastic tubes available in a variety of heights. The tree shelters are placed over seedlings and secured by pushing into the ground. Treessentials Supertubes worked very well with mesquite (Prophis glandulosa) with trees reaching 12 feet in three years. Although wind scour exposed the bottoms of some of the tubes and contributed to the death of some of the plants, the overall survival and growth of the trees has been very impressive (figures 1,2). Treessentials Supertubes used in conjunction with sub-irrigation resulted in rapid improvement in health. Within a week plants that were stunned and wilted with open plastic mesh protectors filled out resumed growth.

The Treessentials Supertubes have worked very well with both minimal and intensive irrigation. Comparable results have been observed in tests in Arizona and California with mulberry, mesquite, creosote bush, saltbush, bur sage, and cheese bush. The benefits for creosote bush were limited to the first few months. Eventually, the shelters effect became adverse, possibly a result of increased temperatures combined with decreased moisture levels.

The major drawback to these shelters is cost. They nest and are easy to ship, handle, install, and reuse. Available in tan and white, from 1-6 feet tall.

&127Treesentials Company &127
75 Bidwell Street, Suite 105
St. Paul. MN 55107
(800) 328-4827 ext. 1906.


The Treepee shelters are made of recycled plastic. The mounting pins are integral to the protector and installation is fast. The top cone piece is added in the field. The cone is larger diameter than other protectors. This is perfect for most sites and species as it provides more room for plant growth. A plastic mesh extender is sold to fit the Treepee, providing additional height protection.

PO Box 550
Laytonville, CA 95454

Tree Pro

The Tree Pro shelters are made of a single face polyethylene and are assemble on site. The top is flared to reduce damage.

Tree Pro Tree Protectors
445 Lourdes Lane
Lafayette, IN 47905

Tree Sentry

The Tree Sentry is an open rolled tube made of recycled polyethylen. This allows opening the shelter to look at the seedling, but this feature may make the shelter susceptible to filling with sand.

Tree Sentry 
PO Box 607 
Perrysburg, OH 43552
(419) 872-6950


The Blue-X shelters are made of rolled recycled X ray film. They can be cut to the size desired. The rolled tubes are relatively stiff. They are attached to a wooden stake.

All Season Wholesale Nursery
10656 Sheldon Woods Way
Elk Grove, CA 95624
(916) 689-0902

Tree shelters have worked well in the desert but they are not appropriate for all situations and species. The value of the tree shelter is related to the species, irrigation system, water schedule. Plants with upright growth forms and rapid growth rates seem particularly well suited for tree shelters. The morphology of spreading plants changes in tree shelters. Ambrosia dumosa appears mushroom shaped when initially removed from treeshelters. After shelter removal the shrub eventually recovers its normal shape.

The treeshelters create a low-light, high temperature microclimate. As little as 30% of outside light is transmitted through the tan TUBEX tree shelters tested in Southern California (Sorensen, 1993). High temperatures may be some negative effects on enclosed plants. Plant response to these conditions will depend on the ability of the plant to acclimate or at least tolerate low light and high temperatures. Although transpiration often keeps shelter temperature below ambient air temperature the leaf temperature may rise when plants are dry. Shelter may limit rainfall, reducing soil moisture at the seedling roots. This may be a concern with the conical Treepee.

The long term affects of tree shelters were called into question in a 1994 review of survival of a 1991 planting. Initial survival and growth were much higher in shelters but the long term survival is much lower for the treesheltered plants (Bainbridge, unpub. data). This suggests that a hardening off process may be required or that timing of removal may be critical to allow p&127lant metabolism to recover. As herbivory is the primary cause of death it may also reflect changes in plant biochemistry.

The extreme grazing pressure in 1993-4 revealed limitations on the effectiveness of the smaller shelters. The shorter 6" TUBEX shelters were kicked over by rabbits (even when buried 3-4 inches at the base) and the plants were eaten. Two experimental plantings were lost. The 8 inch shelter, which seems better for some plants, also provided inadequate protection as animals were able to pull the plants out by grasping the shoot. Many roots were found loose in the shelter with the tops sheared off. A two part system with a short tree shelter and a wire cage has been very effective. Health and growth were better than with the cage or tubex alone, figure 2. The Treepees and inverted top cones from Treepees were more successful. Browsing was restricted to me tops of plants with these protectors.

Figure 2. Herbivory rated on a scale of 1 to 4 (1:severe herbivory; 4:no herbivory). Results are six months outplanting. N=10 for each treatment.

Tree shelters may trap birds, mammals and lizards. Netting or cross-threaded fish line near top of treeshelters will keep out most small animals. A vertical stick placed inside the shelters helps lizards escape. Birds nested on top of seedlings in one trial in Arizona, which reduced growth, but did not kill trees.

b. Wire screen fences, tubes, and cages

Wire cages made with a variety of meshes often provide adequate protection from mammal herbivores. The cage diameter can be easily tailored to the required size and shape. Two foot tall cages, staked with rebar and rock braced along the bottom may exclude most rabbits and rodents, but 3 foot wire is better. Cattle and burros can be controlled with six foot cages made with reinforcing wire mesh and t-posts. Unfortunately, wire cages are costly to build and install. Rodents have burrowed under cages or climbed in open tops. The tops can be pinched in and sealed for increased protection, however, closed tops and smaller cages can lead to sever plant entanglement if they are not removed in time. Removing cages from entangled plants can be frustrating and time consuming. Racin (1988) and Virginia (per. comm) have reported similar problems with wire cages. In any case the cages must eventually be removed. If they are well built they can be repaired or reused.

We have found that 12" diameter wire mesh tubes (1/2"-1" mesh), tied to 1/4" reinforcing bar stakes ("pencil rod") worked well. The tubes are rolled and fastened with aviary clips (available locally at San Diego Hardware). Racin (1988) used 18" tall tubes but these would not be tall enough for leader type seedlings. Large diameter cages 2-3 feet tall with polyethylene plastic film, bubblepack, or floating row cover material wrapping are most suitable for spreading plants in severe wind blast areas. We are now using 12" diameter, 3' tall cages made using stucco mesh. This mesh is heavier gage and is cut to have points facing up. The loose edges are threaded together with a piece of pencil rod. The cages can be installed quickly and taken apart for packing, storage and re-use.

The major drawbacks clipped wire cages are the cost of assembly, installation and removal and difficulties in transport and handling. Racin (1988) found that two workers could make 1,000 wire mesh protectors in 5 days. At minimum wages the cost would be about $300 for labor, $1,000 for the wire, $240 for the stakes, plus cage clips, ties, etc. Total cost may be from $2.00-$5.00 ($1.50 for the threaded stucco wire cages) each plus installation. Cage making is generally a good activity for volunteers. If a perimeter fence can be installed it can minimize cost, but it may be difficult to eliminate herbivores inside larger areas. Nabhan (pers comm.) has reported great difficulty in clearing a relatively small fenced area in the botanic garden in Phoenix.

c. Rock mulch

Rock mulch provides good protection from climatic extremes and some protection from herbivory. Three or more medium to large rocks are arranged around each seedling. The thermal mass of the rocks protects from heat and cold, the rock reduces moisture evaporation and blocks the wind, and if properly placed, limits grazing. As the plants mature, the rock mulch can be left in place. If rocks are available the cost is low and the aesthetics are pleasing. The rock mulch improved survival of basin irrigation plants, but reduced survival of buried clay pot irrigated plants, perhaps by limiting root extension.

d. Plastic mesh

International Reforestation Suppliers (IRS) manufactures plastic mesh tubes for plant protection. These tubes have been effective on some sites but marginal on others. Plastic mesh screen provided limited protection for mesquite, the most preferred browse. Many screens have been chewed through and several plants have been pulled out of the ground and completely consumed.

The open mesh also provides little protection from drying winds and blowing sand. The plastic mesh is photodegradable (different lifetimes are available) and if staked with bamboo can be left in place. The mesh is also available with a fertilizer in the mix.

IRS sells a 1 mil plastic bag which fits over the tube and provides some of the benefits of the rigid treeshelters. These are awkward to install and short lived.

The 3" mesh tubes are better for leader plants that bushy or spreading plants which can be deformed by the plastic or damaged during installation (Racin, 1988). The larger mesh size may also allow mice in.

The advantages of these tubes is the low cost and easy handling of large quantities.

International Reforestation Suppliers
PO Box 5547
Eugene, Oregon 97405
(800) 321-1037

e. Plant collars

Even fairly short collars of plastic, peat, or paper can improve survival. These reduce bending at the ground, improve watering efficiency, reduce sandblast and evapo-transpiration of young seedlings. They also reduce grazing. PVC pipe collars (3" diameter collars, 3" tall) have improved survival on some sites. Larger diameter collars 4" -6" tall have worked better on sever sites where shorter collars fill in with sand. The 3" diameter collars are easy to integrate with the IRS mesh protectors. The IRS peat protective collars survived well in the desert but were too small to provide significant protection for seedlings. Cottage cheese cartons with the bottom cut out and bottomless cups have also been used.

f. Repellent

Repellents may provide some protection in the desert. These include both commercial and homemade solutions which make plants less palatable. Dry desert climates reduce washoff repellants, a common problem in wetter forest use.

An earlier repellent test was abandoned after rabbits severed almost all of the drip tubing without eating any plants. Repellent may prove more useful in protecting spaghetti tube in drip systems. The repellent may be mixed with latex paint or some other binder.

Smoke tree (Psorothamnus spinosus) or bladderpod (Cleome isomeris) may provide extracts that will mask the more palatable mesquite and palo verde. Systemic repellents are also available and may prove useful. These tablets or solutions may be placed in containers in the nursery or are placed in the planting hole. They release chemicals which are taken up by the plant. Workers should wear gloves and a respirator when working with bulk quantities or repellents such as Anapel. An extensive test of repellents in 1993-4 showed no statistically valid benefit from 3 commercial repellents, figure 3. The grazing was slowed slightly but long term damage was comparable. Repellents may be more effective with rapidly growing plants on irrigation systems or when applied in the nursery mix. This will be evaluated in more detail in 1994.

Figure 3. Herbivory rated on a scale of 1to 4 (1:severe herbivory; 4 no herbivory). Results are three months after outplanting. N=10 for each treatment.

ASG Consultants
7868 11th Avenue
Burnaby, BC V3N-2N3
(604) 521-0864

g. Dead plants, straw, etc.

Dead plants, straw bundles, and other shading and screening device can provide protection for plants. Plants covered with plant debris in washes experienced mush less herbivory than plants in the open. Dead plants and branches provide less than perfect protection from herbivory and are most appropriate where commercial materials are unavailable or not visually acceptable. Thorny and inedible plants like smoke tree branches provide the best protection. Buried straw bundles or loose straw has provided similar benefits in some projects.

h. Shade screens

Shade screen provided some improvement in survival of Palo verde but not creosote bush. Shade cloth made screen cages more vulnerable to high winds. Some benefits may be obtained for riparian plants or transplants.


Plant establishment on arid and semi-arid sites requires careful attention to many factors. Neglect of any one of these may lead to complete failure or very low survival. Plant protection have rarely received sufficient attention. If careful preparation is combined with good protection excellent survival and growth may be achieved with minimal water use, low maintenance, and low cost.

The higher cost of better protectors appears to be offset by improved survival and growth. Treepees, Subertubes, or plastic wrapped wire cages may make limited irrigation feasible even on extreme sites. Systematic repellents applied in the nursery appear promising but have not been tested in the desert. The ultimate goal is an integrated container/shelter system that minimizes root disturbance and planting cost.


Special thanks to Pam Beare, Rober MacAller, John Crossman, Laurie Lippitt, Ronnie Clark, Mark Faull, Matt Fidelibus, Jeff Thomas, NaDene Sorenson, John Rieger, and Bill Steen. Support provided by the California Department of Transportation, California Department of Parks and Recreation, and Ecocultura.


Bainbridge, D.A. 1991. Successful tree establishment on difficult dry sites. pp 78-81. In Proceedings of the Third International Windbreak and Agroforestry Symposium. Ridgetown, Ontario.

Bainbridge, D.A. & N. Sorenson. 1990. Seedling establishment in Colorado Desert washes. Third Biennial Desert Conference Program, California State University, Dominquez Hills, page 4 (abstract).

Bainbridge, D.A. & R.A. Virginia. 1990. Restoration in the Sonoran Desert. Restoration and Management Notes 8(1):3-14.

Bainbridge, D.A., R.A. Virginia and N. Sorenson. 1993. Revegetating desert plant communities. pp. 21-26. In Proceedings , Western Forest Nursery Association Meeting. USDA Forest Service GTR RM-221.

Bainbridge, D.A. 1994. Treeshelters improve establishment on dry sites. Treeplanters Notes (in press)

Browers, J.E. 1980. Catastrophic freezes in the Sonaran Desert. Desert Plants 2(4):232-236.

Hughes, C.L. 1963. A study of the evaporation from the Salton Sea, California. U.S. Geological Survey Open File Report, Yuma, Arizona.

Lenz, L.W. and J. Dourley. 1981. California Native Tree Shrubs. Rancho Santa Ana Botanic Garden, Claremont, CA.

McAuliffe, J.R. 1986. Herbivore-limited establishment of a Sonoran Desert Tree, Cercidium microphyllum. Ecology 67 (1):276-280.

McReary, D. and J. Tecklin. 1993. Tree shelters accelerate valley oak restoration on grazed rangelands. Restoration and Management Notes 11(2):152-153.

Mosjidis, J.A. 1983. Detection and control of sand blast injury to jojoba seedlings. Desert Plants. 5:35-36.

Racin. J.A. 1988. Revegetating desert cut slopes with containerized native shrubs. Cal Trans Transportation Lab, FHWA/CA/TL-88.09. Sacramento, CA 88p.

Sorenson, N. 1993. Physiological ecology of the desert shrub Larrea divaricata: implications for arid land revegetation. M.S. Thesis. San Diego State University, San Diego CA

Virginia, R.A. and D.A Bainbridge. 1987. Revegetation in the Colorado Desert: Lessons from the study of natural systems. pp 52-62. In proceedings of the 2nd Native Plant Revegetation Symposium, Society for Ecological Restoration and Management, Madison, WI.

Wallace, A. and E.M. Romney. 1972. Radioecology and ecophysiology of desert plants at the Nevada test site. U.S. Atomic Energy Commission TID 25954, NTIS 439 pp.

Windell, K. 1993. Tree shelters for seedling survival growth. USDA Forest Service TEch Tips 2400. 4 p.

Young, J.H. 1992. Development of techniques to establish thorn scrub seedlings for restoration of ocelot habitat. MS Thesis, Texas A&I University. MAI Abstracts International 31(1):205