Soil Ecology Restoration Group

last update December 30, 1999

Mesquite Dune Experiment Area in the Yuha Desert, California.

Mesquite Dune Reconstruction in the Colorado Desert

David Bainbridge

In March 1995 we began a project to create a mesquite accretion dunes in an abandoned borrow put at the southwest corner of the intersection of highways 86 and 78, adjacent to lower San Felipe Creek near the Salton Sea. This project was partial mitigation for damage caused by widening and other improvements of highway 86. Creation of the dunes was initiated because they are biologically important for both ecosystem structure (primarily habitat) and process (soil accumulation and nutrient cycling). It was also hoped that the studies would offer new insight into how best to control wind erosion. We anticipated that creation of these mesquite mounds would accelerate soil recovery (soil accumulation &127 and microbial inoculation by capture of wind transported particles, nitrogen accumulation by fixation), promote re-establishment of the original surrounding vegetation, and provide high quality habitat for local fauna.

The project was also designed to test the effectiveness of alternative methods of stabilizing erosion prone materials in the desert. Five covering materials and a control were tested with four replicates each: 1) straw punched into the sand with shovels; 2) coir (coconut husk fiber) netting covering with the windward two thirds of the mound, 3) xanthan gum sprinkled onto the sand and subsequently moistened to create a crust; 4) bark chip mulch covering the mound; and 5) plastic barrier fence placed immediately upwind of the mound. Eight seedlings were planted on the mounds, with the exception of a set of bare mound controls.

An irrigation system was set up using plastic tanks on wooden towers, PVC main line, 1/2 inch laterals and spaghetti tubing, and drip emitters. This proved marginal. Pressure was to low for effective water delivery and emitters easily clogged and needed to be cleaned repeatedly. This limited survival of plants. We continued watering, maintenance. and monitoring of the mesquite dunes through the fall and winter. A solution to watering problems was finally found by John Tiszler when he hooked up a small gas water pump in line to boost pressure, but mortality had already occurred for some mesquite seedlings. After some settling and reshaping under strong spring winds in 1995, the dunes appeared to settle in. They were watered periodically in 1996 and twice in 1997, not at all in 1998, and once in 1999 at the time of monitoring 6/21.


After more than 4 years only the punched straw and plastic erosion fences have held up, although two of the fences had partly blown down as wood posts were eaten by termites. Sunlight, winds, and foot traffic during watering and monitoring over the years contributed to the failure of the other systems. The coir was finally broken down almost completely, the xanthan gum crust was seen only in small patches, and all the bark had blown off the mounds.

Mound Stability

There were differences among treatment types and these had changed over time, but these were not statistically different. The coir covered mounds showing the largest drop in height (as the coir starts to break down) and bark mounds showing the smallest drop, Table 2-1. Bare mounds with no plants lost only 3.5 inches, suggesting that some of the loss in the other mounds is related to foot traffic related to watering and measurement. The bare mounds may also erode less because there is less turbulence than on the planted mounds. The mound that was dug up, perhaps by a coyote trying to excavate a rodent in 1998, had largely healed with sand blown in off the sides and top.

Table 2-1. Mound height stability

Mean Cumulative loss (inches)

  Year 1 Year 2 Year 3 Year 4
Bare 4.3 3.5 3.6 4.3
Bark 2.7 2 2.2 3.0
Coir mat 0.7 0.0 10 2.0
Barrier fence 2.6 1.9 3.6 2.6
No plant, bare 1.0 1 5.8 1.0
Straw 1.6 0.8 6.7 1.9
Xanthan gum 2.6 1.9 5.6 2.4
Mean 2.2 1.6 5.3 2.4

The loss in mound height under all treatments is relatively small, but several mounds have formed east pointing "tails" of finer materials, suggesting downwind movement is occurring. High winds had occurred at least once as fences were blown down and new sculpting was noted on some mounds. Full cross-sectional measurements will this winter determine sand loss volume.

Seedling survival

The mesquite seedlings have suffered intense herbivory since they were planted despite cages and treeshelters, primarily by jack rabbits (Lepus californicus), figure 1. At least 86 were alive in 1998 and 83 in 1999, but with mesquite it is sometimes hard to tell in midsummer and 86 may still be alive. If even one survives on each mound the mound may develop properly. In June 1999, only one mound had just one survivor, almost half of the mounds had four or more survivors. The biggest problem has been herbivory by jack rabbits, figure 2-1..

Figure 2-1. Herbivory by Jack Rabbits

These were recaged with taller wire cages in spring 1999 to increase growth. A deep pipe irrigation system was also installed on each mound, and each pipe was given a tablespoon of bone meal (P) and a teaspoon of fertilizer (18:6:12).

Overall plant survival averages 43% over all treatments. Survivors by treatment are shown in Table 2-2. There has been no positional trend associated with mortality and our observations suggest death may result from several phenomena including: sever herbivory, root clipping or exposure after burrowing activity below the seedlings, exposure of the roots from erosion from the wind, and desiccation. Mortality was very low this year despite the very low rainfall

Table 2-2. Percent Survival 1996 1997 1998 1999 Rank
Xanthan gum 86 66 53 50 1
Barrier fence 86 62.5 53 46 2
Bark chunks 58 69 47 44 3
None 75 47 43 41 4
Crimped straw 73 59 40 40 5
Coir netting 78 56 31 38 6

Growth of at least one, and sometimes more, tree has been good on most mounds, figure 2.2

Figure 2-2. Mesquite seedlings have continued to grow larger


The mounds provided an immediate improvement in habitat diversity and many organisms have colonized the mounds. There are a total of 225 burrows this year, up from 86 in year 1 , Table 2-3. Most mounds have multiple burrows although the mounds with mesquite are more heavily populated. Determining which animal, insect, lizard or snakes uses a hole can be difficult, and no detailed assessment has been done. Species that have been noted in the burrows include: sidewinder, lizards, a horned toad, desert iguanas, small mammals (perhaps including Kangaroo rats and mice), and insects, including some large spiders. The burrowing mammals were probably responsible for some if not most of the mortality of plants on some mounds. Birds have also been observed sitting on the mesquite. It is clear that the mound topography and mesquite has adding richness to the once flat and largely barren borrow pit.

Table 2-3. Burrows (means) 1996 1999
Barrier fence 3 7.5
Crimped straw 4 7.25
Xanthan gum 3 6.75
Bark chunks 2.5 4.75
None 1.5 5.5
Coir netting 1 3
Total 86 166

There was a decline in burrows from 1998, probably as a reflection of the serious drought. The coir netting was a poor choice for habitat improvement. The crimped saw, xanthan gum, and barrier fence appear to provide more favorable conditions for burrowing insects, lizards and mammals.


The site did not get much rain this year and remains very dry. The rainfall at the nearest station was 1.77 inches for the 12 months of the 1998-99 season, while the measured evaporation was 76.39 inches. As a result even the native creosote bush near the mounds is very stressed. The differences between treatments have not been dramatic, although there were differences in erosion, with straw now the best treatment. The punched straw is still holding up well, although termites have attacked the straw on the mound. The straw treatment is inexpensive and provides good performance on a range of criteria. It would still be our recommendation for most projects of this type.

However, the choice for a particular location may depend primarily on the desired goal. If short term stabilization of erosion is critical (especially for 1-2 years) the coir may be needed. If a more complex goal of erosion control, plus revegetation and burrows is the goal then straw, bark, barrier fences, or xanthan may be equally appropriate.

In 1998-99 we began cleaning-up the site, removing irrigation lines and plant shelters. Additional mesquite will be planted in drainage features on the site. For future projects of this type a 3-4 foot high fence around the entire mound with 1 inch mesh would be recommended. Some of the plants should also be treated with a window screen mesh shelter for protection against insect damage. A number of trees have now escaped the herbivores and with new cages others will also start to add height and size in the next year.

The strategy chosen for these mounds appears to be marginally successful. If we were asked to do it again we would grow the mesquite trees first and then bury them with sand. We believe this would more quickly create a more natural and more stable mound.