You are on page 1of 12

Microcatchment water harvesting for desert-revegetation

Prepared for
The California Department of Transportation
District 11m 2829 Juan Street
San Diego, CA 92138
Jul 199!
"atthe# $% &ideli'us and Da(id A% )ain'ridge
The desert regions of southern California experience low and sporadic annual rainfall. Low
annual rainfall and unfavorable distribution of rainfall throughout the year limit soil moisture in
this and other arid and semi arid zones (Boers, !"#$ and plant production is often limited by
water availability (%hanan,!&!$. To establish and maintain plants, techni'ues for improving
capture and utilization of local water supplies must be developed (%hanan, !&!$. (lthough
irrigation can overcome water shortages if surface or ground water is available, a drainage
system may be needed to maintain a favorable salt balance (Boers, !"#$. ( complete irrigation
and drainage system is not always practical or desirable if the immediate region lac)s water, or if
plants are being established in a low or no maintenance area. (n irrigation system may also be
undesirable from an environmental point of view when preserving or establishing native plant
stands (*hrler, !&"$. +ater harvesting, a method of exploiting available precipitation, is
practiced in many arid and semi, arid regions for both economic and environmental reasons.
-icrocatchments could be much more widely used in California to reestablish native vegetation.
(ncient desert civilizations rarely had the s)ills or resources to transport water over long
distances, but they did develop very effective systems for utilizing water (%hanan,!&!$. -iddle
*astern civilizations, for example, used flashfloods to irrigate the flood plains of streams, and in
some cases, chec) dams and flood bypass structures were constructed to control water flow.
These were the first runoff farms. %imilar techni'ues were used in (merica, .orth (frica,
-exico, and %outh (rabia. /t is unclear why the extensive systems in the mid *ast were
abandoned, but it appears li)ely that it was for political, rather than environmental reasons.
-ichael *venari, Leslie %hanan, and .aphtali Tadmore have studied the uni'uely adapted
farming methods employed several thousand years ago on the .egev 0esert of /srael for many
years. 0uring the reign of 1ing %olomon and his successors between the tenth and sixth
centuries B.C., the .abataens appeared as traders in the .egev (2all, !&!$ The .abateans
eventually developed the technology to collect and direct runoff form the hills to crops. 0uring
the Byzantine empire from 345 B.C. to (.0. #65m .abataean agriculture reached its pea) with
improved water harvesting methods that were practiced on a large scale (as much as 655,555ha$.
Crop production from these extensive systems is impressive in a desert with annual rainfall of 6,
7 inches ("5,55mm$ per year8 similar to the normal annual rainfall recorded in the Colorado
0esert. /sraeli researchers were intrigued by the ancient farms, and one of the farms, +adi
-ashash was restored in the early !#59s in an attempt to :relearn: the ancient techni'ues of
water harvesting. The positive results of the +adi -ashash restoration inspired /srael to invest in
water harvesting research as a means of supporting many types of crops (%hanan, !&!$.
-icrocatchments have also been used to establish vegetation for par)s and roadside rest areas.
These studies formed a solid foundation for modern microcatchment development.
-odern microcatchments systems harvest runoff water in a laminar flow , water depth less than
;" inch (,3 mm$, and flow velocity less than 3.&4 inches;sec (& cm;sec$ (%hanan, !&!$. The
catchment area can be left unaltered, altered (defoliated and compacted$ or treated with living
soil crusts or water resistant compounds to enhance runoff. The collected water use then
effectively :stored: below the soil surface where it is available for plants but protected from
evaporation higher relative water yield per unit surface area than the larger runoff farm
catchments. The size of catchment can be tailored to provide an optimal runoff volume for
individual plants. -icrocatchments enhance leading and often reduce soil salinity (%hanan,
!&!$.The use of microcatchments techni'ues in (rizona has continued the productive use of
land retired from ground water irrigated agriculture (1arpisca), !""$
-icrocatchment systems provide many advantages over alternative irrigation schemes. They are
simple and inexpensive to construct and can be built rapidly using local materials and manpower.
The runoff water has a low salt content and because it does not have to be transported or pumped
it is relatively inexpensive. The hydrological data needed for an efficient design can be collected
through observations over two to five years even in areas with limited rainfall. =inally, the
systems are easy to operate and maintain and relatively safe from failure.
-any types of domesticated crops have been successfully established using microcatchments,
but water harvesting has also been used to supplement rainfall for water stressed native
vegetation. *hrler (!&"$ compared the growth and seed production of >o>oba (%immondrisia
chinensis$ plants that were given one of three treatments? T5$ no water harvesting catchments8
T$ cleared, smoothed and rolled 35 meter s'uare catchments8 and T3$ li)e Tplus a water
repellent coating. The plants receiving supplemental water form micro catchments, especially the
treated catchments, were larger in volume and produced flowers and seeds than the untreated
plants. =igure . *ven native plants that are well adapted to hot and dry conditions will usually
benefit from supplemental water provided by microcatchments (as illustrated by the increased
height of creosote bush, Larrea divarivata, alongside desert highways$.
Rainfall and Climate of the California Deserts
(lthough rainfall patterns vary in the California deserts, rainfall can generally be described as
bimodal with a pea) precipitation occurring in @anuary, and declining gradually to allow in @une
when there is essentially no rainfall. ( second pea) occurs in many parts of the Colorado 0esert
in (ugust as a result of tropical storms. /n all areas, the potential evaporation far exceeds the
average precipitation. /n !&4, at the Anited %tates 0ate Bardens at /ndio, for example, total
precipitation amounted to only three inches (&#mm$, while free evaporation (from an open pan$
was 3.!# inches (3"#! mm$, more than forty times the years of precipitation (Lenz, !"$.
-ost of the Colorado 0esert (Coachella Calley and related area$ normally receives less than five
inches (3& mm$ of rain per year. This aridity is the result of several environmental factors? the
stable Dacific high pressure cell which diverts storms to the north8 the cold California current
which precipitates rain before it reaches land, and series of mountains along the western
periphery of the Colorado 0esert which create a rain shadow effect (Bainbridge E Cirginia,
!"#.$ %torms generated from tropical air masses moving north from the Bulf of California can
drop rain e'ual to the yearly average in a matter of minutes (Bainbridge E Cirginia, !"#$.
%torms such as these recharge soil moisture and are important in the spring landfall for
establishing seedlings. /n !!, late spring rains resulted in considerable creosote bush
germination and establishment at a site in /mperial County and at Fed Foc) Canyon %tate Dar)
in the -o>ave 0esert.
The high temperature and limited annual rainfall of CaliforniaGs deserts result in plant water
stress. (s temperatures rise, the rate of evaporation increases and with it the water demands of
the vegetation (.ir, !&7$. Dea) temperatures in the 559s H= (6" HC$ occur from @une through
%eptember when less than .4: (6&.4mm$ of precipitation normally occurs.
Temperature and precipitation values can be combined to graphically portray the aridity of a
region. ( graph of precipitation values in millimeters plotted against temperature values in HC
with a scale of HC ?3 mm is a common climatic yardstic). +hen the curve of precipitation
values is below the temperature curve, then the amount of precipitation is smaller than twice the
temperature length and intensity of the dry season. Braphs of temperature and precipitation
values for several weather stations in the Colorado 0esert illustrate the year round water deficit,
figures 3, 6.
+ide temperature variation, daily and annual, is characteristic of both the -o>ave and Colorado
deserts of southern California. /n the summer months, the difference between mean high and low
temperatures can be greater than 65H= (&HC$, 55H= to &5H= (6",3 HC$ (.I(($.
(mbient air temperature and soil temperature have a pronounced effect on plants. The high air
temperatures of the summer increase the plants water demand for evapo,transpiration. /n
@anuary, mean low temperatures are between 64,75 H= (.4,7.4 HC$, with highs between #5,&5
H= (#,3 HC $ a span of 34 H=.
=reezing temperatures are not uncommon on clear winter nights, but hard freezes only occur
about every decade. %everal hours of below 3# H= temperatures can damage sensitive plants
(BainbridgeE Cirginia, !"#$. /n (ugust !"! a rainfall initiated growth of many palo verde
seedlings in the Colorado 0esert. The ma>or causes of death in the initial months were herbivore,
frost, tramplings, and flood damage. Thus freezing can affect these plants as significantly as high
temperatures (Bainbridge ECirginia, !!5$.
Bare soil temperature in the sun may reach 4&,#7 H= (7,"HC$ or greater above ambient air
temperature under windless conditions. +hen the wind is blowing, exposed soil temperatures
may be only 7 HC above ambient air temperature (+allace E Fomey, !&3$. Iptimum root
growth for mes'uite and many other summer active desert plants occurs at soil temperatures in
the high "59s H= (3& HC $, usually not achieved in until early summer.
Microcatchment design
There are four types of microcatchment systems? micro,watersheds, runoff strips, contour bench
terraces, and catchment basins (%hanan, !&!$. If these, runoff strips and contour bench terraces
are best suited for agriculture as they re'uire extensive mechanical re,shaping of the surrounding
terrain and create regular patterns which are inconsistent with a natural appearance. /n runoff
strips, a series of Jsaw,teeth: ridges are built by moving soil laterally and longitudinally with a
grader to increase the natural slope and form a contour strip to collect runoff on a parallel
cultivated strip.
/n contour bench terrace systems, the runoff strips of the contributing area are left unaltered, but
the cultivated strips are leveled longitudinally to distribute the runoff evenly (%hanan, !&!$.
-icro,watersheds and catchment basins can be built using hand labor and are this less expensive
and more adaptable to revegetation pro>ects.
-icro,watershed systems include mound and strip collectors. %trips can be built with mechanical
e'uipment of by hand. The steps are bordered on each side by ridges ",35 inches (35,45 cm$
high and #.4,#.4 feet (3,4 meters$ apart. The result is a series of linear strips well suited for
crops such as corn. The waffle gardens of the Kuni in .ew -exico are a form of crossed strip
/n mound systems the soil surface is shaped by hand into 7,35inch (5,45 cm$ tall mounds
spaced 3,4 meters apart. +hen organized into a regular pattern, this system is suitable for many
types of farm crops, including melons and s'uash (%hanan, !&!$. The mounds are also effective
when arranged in a more random manner for revegetation and restoration efforts.
%ince catchment basins involve a minimum amount of labor and less manipulation of the
surrounding terrain, they are probably best suited for revegetation and restoration. ( gently
sloping plain (ideally with slopes less than 4M$, is divided into plots by small earth ridges 5,35
cm high and 35,65 cm wide (%hanan, !&!$. The ridges are constructed with the soil excavated
from a planting basin about 75 cm deep or from the uphill side of the ridge. These can be
constructed by hand, or with a small plow. Catchment basins are susceptible to siltation and
erosion if undesired runoff is allowed to enter the system, so protective diversion ditches are
often constructed above areas sub>ect to extensive ground flow.
The gradients of the micro catchments should be between ,&M.%'uare or rectangular plots are
easy to sta)e out so they are commonly used, however basin shapes for revegetation sites should
be tailored to suit the geography of the site to reduce disturbance and retain a more natural
Dotential water yields must be estimated before other decisions regarding microcatchment design
can be made (%hanan, !&!$. To determine expected yields from microcatchments three rainfall
characteristics must be evaluated? $ the average annual rainfall, 3$ pea) rainfall intensity, and 6$
the minimum expected annual precipitation. The average annual rainfall must be )nown to
predict the long term water availability.
Consideration of a site for microcatchment construction must also include four main
physiographic factors? the runoff producing potential8 the soil surface condition (cover,
vegetation, crust, stoniness$8 the gradient and evenness of slope8 and the water retaining capacity
of the soil in the root zone profile (%hanan, !&!$. These all contribute to the runoff threshold
coefficient which is a )ey factor in determining the optimum size for catchment (2owell, !"!$.
Ither factors affecting the infiltration capacity of a particular area include? the moisture content
of the soil8 macropores in the soil as result of decaying roots or burrowing animals8 and the
compaction of the soil (%hanan, !&!$.
The optimal size of the microcatchment for each species depends upon many factors including
normal precipitation, the soil 'uality, and the slope (*vanari, !&$. The size and depth of the
planting basin in relation to the size of the catchment area is also important. These factors
determine the size of the surface area wetted by runoff and the volume and depth of the water
column in the soil. /f the infiltration rate of the soil and the water demands of the plant are
)nown, the size of a catchment basin can be estimated. /f a particular species of shrub re'uiring
65 inches of rain per year is being grown in a region of 4 inch average annual precipitation, and
then an additional 4 inches of rain is needed. /f the catchment soil has a runoff of 5M (a typical
runoff volume for untreated desert soils$, then a 55 s'uare foot catchment should yield enough
water to meet the water re'uirement for the shrub (2owell, !"!$.
Estimating runoff
*ffective precipitation is defined as that which produces runoff (Toy, !"&$. %urface infiltration
is often proportional to vegetation cover, so as cover decreases, infiltration decreases. This
results in greater volume and velocity of runoff for a given rainfall event. Catchments areas are
often defoliated (a condition faced in most revegetation, restoration efforts$, and given additional
treatments to increase runoff.
To illustrate the potential advantage of treated microcatchment areas, rainfall can be reported as
a volume rather than as a depth (-yers, !#&$. Ine millimeter of rain e'uals one liter of water
per s'uare meter assuming 55M runoff. %everal runoff enhancing treatments have been
evaluated on microcatchment basins. /n !&3, =in) et al. compared runoff yields untreated basins
with basins treated with paraffin wax basins covered with butyl sheets. Daraffin (candle wax$ is
applied to the soil by hand, in the form of granules, at a rate of one to two pounds per s'uare
yard. +ax with a low melting point, 35,45 H= (7!,#4HC$, will melt within a few days in hot
desert environments to form a solid coating on the soil surface (=in), !&6$. The wax treated
soils yielded !5M runoff compared to 65M runoff on untreated soils, and 55M runoff from a
butyl covered plot. +ax treatments are best for sandy soils, and some plots have remained
effective after five years, sufficient time for plant establishment.
/n soils with clay content between 4,65M, sodium salts (including sodium chloride$ effectively
reduce permeability by dispersing clay particles (Cluff and =orbel, !&"$. +hen used in a
suitable soil, the sodium is absorbed by clay, and the 'uality of the runoff water is tolerable. The
'uality generally improves after the initial one of two runoff events. Theses treatments only wor)
on the clay fraction of a soil, however, and will not be effective on sandy or course grained soils.
%alt treatments will often be inappropriate for revegetation sites.
-any types of synthetic membrane materials have also been used to increase runoff. Dlastic
membranes, such as polyethylene and vinyl, have been used on extensive revegetation pro>ects in
China8 they are very effective but generally last less than one year. Butyl rubber and chlorinated
polyethylene sheeting lasts much longer, but these materials are expensive and must be well
secured to protect them from wind damage.
(sphalt, concrete and other hard surfaces in urban and suburban areas can also be used to
channel water to catchment basin plantings. Landscaping on streets in Tucson is increasingly
watered this way and the potential for highway plantings is high, despite water 'uality concerns
from road surface runoff (2owell E 2owell, !!8 Cirginia E Bainbridge, !"#$.
*nhancing the development of natural surface crusts has applications for increasing runoff. /n
many soils, crusts formatter rains8 either by mechanical action due to impacting raindrops or by
chemical action such as hydration and dispersion (=reebair et al,. !!$. /n areas with segregated
soil types a thin layer of well oriented clay forms after wetting, enhancing runoff (Tac)ett
EDearson, !#4$. -icrophytes such as bacteria and lichen living on the soil surface also
contribute to crusting. Fesearcher in /srael have observed that colonies of blue,green algae (cyan
bacteria$ secrete threads that entrap sand particles producing a thin, grayish soil crust (Billis,
!!3$.This crust my decrease infiltration and increase runoff. Brotherson and Fushforth (!"6$
found that microphytic crusts in northern (rizona allow deeper penetration of applied water, and
reasoned that the crusts seal the soil surface, reducing evaporation (+est, !!5$. 2arper and
-arble, (!""$ however found that sites heavily crusted with dar) colored cyanobacteria lost
significantly more water from the upper &,4 cm of soil than intermixed, scalped plots. They
concluded that the dar) crusts absorbed more solar radiation than the light colored, uncrusted,
soil leading to higher evaporation. -ore research on the effects of microphytic crusts on soiled,d
water relationships is needed, but their potential use as biological :coatings: for microcatchments
appears promising.
Water demand
Lac) of water is often a critical problem for revegetaion efforts of arid lands. %uccessful
transplanting and direct seeding may re'uire temporary supplemental water. (fter an initial
establishment phase, the demand for water is much reduced for survival but remain important for
growth. %palding (!!5$ found that Larrea transpiration rates increased ",! times when irrigated.
+hile most of these species respond positively to supplemental water, they generally need well
drained soils due to low resistance to fungal infections and oxygen deficiency (=isheret al.,
!""$. Anli)e the shallow diffuse roots of typical domesticated plant species, many Colorado
0esert plants develop deep tap roots that eventually reach the water table. /n (rizona, live
mes'uite, Drosopis glandulosa, roots have been found more than 45 feet below the soil surface.
2owever, phreatophytes such as mes'uite and more drought tolerant species such as fourwinged
salt bush ((triplex canescens$, white bur sage ((mbrosiadumosa$ and creosote bush (Larrea
divaricata$ all need relatively high surface soil moisture to germinate and become established.
Conditions suitable for plant establishment may only occur once or twice a decade when
favorable moisture patterns develop (Bainbridge E Cirginia, !"#$.
(mbrosia dumosa li)e Larrea, is very drought tolerant and prefers well drained soils, but it is
more tolerant than creosote bush to low soil oxygen levels !Lunt, !&6$ that might be
encountered in catchment basins. +hite bur sage is drought,deciduous and loses its leaves during
drought stress (Bamberg, !&4$. Ander favorable conditions it exhibits high photosynthetic and
transpiration rates. (mbrosia dumosa, a good candidate for revegetation in the Colorado 0esert,
should respond favorably to micro catchments.
( triplex canscens also appears to be a good candidate for micro catchments. ( study showed
that (triplex seed emergence was always greater in shade and correlated with rainfall (2ennessy,
!"7$. %choll (!"4$ observed that stands of saltbush in .ew -exico established themselves in
basins and furrows on a mine site where soil moisture content was higher. (de'uate soil
moisture is also very important for transplanting this species.
-es'uite li)e fourwing saltbush, white bur sage, and creosote bush re'uires supplemental water
during the first few months of establishment. Iver,irrigation may decrease growth and flooding
may )ill mes'uite trees, so the trees should be planted on a mound in the basin or on top of the
di)e where they will be safe from prolonged flooding.
Construction and Maintenance
The first step in constructing microcatchment basins is clearing the catchment area of weeds. The
catchment and basin area should then have it contour lines sta)ed so that the catchment area can
be shaped into evenly sloped basins to maximize runoff and minimize erosion (%hanan,
!&!$.%haping can then be carried out with hand tools, plows or graders. %haping should be
underta)en when soil moisture is near field capacity if possible (0utt, !"$. The catchment area
is smoothed with hand ra)es following rough shaping. %oil treatments, if used are applied after
smoothing. The soil is compacted following the first rain storm. Compaction can be done by foot
or tools on small catchments or with rollers.
Basin size depends upon re'uirements. %mall basins (less than!mN3$, can be constructed by hand
labor, but larger catchments should be built with e'uipment (%hanan, !&!$. Basin should be
shaped to form inverted truncated pyramids. The soil removed from the basin area is deposited
and spread on the border ridges. =our wor)ers can excavate a basin 6.4mN3x35 cm deep and
raise border chec)s 35 cm9s high around the 345 mN3 catchment area in about one hour (*venari,
-icrocatchments are traditionally planted with shrub or tree seedlings. To ta)e advantage of
pea) precipitation and favorable temperature, plantings of native shrubs in the Colorado 0esert
should be planned before precipitation pea)s. The soil around the planting spot should be
loosened before planting as compacted soils retard tap root growth which can be essential for
successful establishment (Bainbridge E Cirginia, !!5$.
Traditionally, the tree shrub is planted in the basin near the lowest point of the catchment, where
the water would be deepest and the ridge highest. 2owever, for many desert species this is not
desirable (Irev, !""$. (fter an intense rainstorm, the catchment will fill with water, and may
remain flooded for several days. 0uring this time, the roots and lower stem suffer form lac) of
aeration and may succumb to fungal diseases. The infiltration rate of the soil will help determine
the best planting spot. =ast draining soils would be better candidates for planting at the bottom of
the basin. %low draining soils should be planted on the di)e or basin border. The water collected
in the basin will moisten the soil above the water line by capillary action.
/n many cases, the best place to plant desert shrubs in catchments is immediately above the mean
high water line, or near the top ridge where the topsoil from the basin is placed, figure7.
Ibservations of conventional plantings reveal both weeds and shrubs appear to thrive on this site
and are absent from the places where water stays longer (Irev, !""$.
-icrocatchments wor) well and should be more widely used. -icrocatchment basins have been
utilized for crops for thousands of years. .ew technology has improved soil runoff potential and
increased the positive effects of microcatchment basins. -icrocatchments are relatively
inexpensive and re'uire little maintenance. They are now being extensively used in /srael for
planting crops and have excellent potential for revegetating sites in the California 0eserts and
other arid and semiarid regions.
Bainbridge, 0.(. E F.(. Cirginia. !"#. The Colorado 0esert. Fevegetation in the Low 0esert
Fegion of California with *mphasis on the %mall Leaved +oodlands. %an 0iego, C(, A%(?
%an0iego %tate Aniversity. p 3
Bainbridge, 0.(. E F.(. Cirginia. !!3. 0esert Dlants. %an0iego, C(, A%(. %an 0iego %tate
Bainbridge, 0.(. !"#. %and tan)s for dryland water storage. 0rylander ($?&.
Bainbridge, 0.(. E F.(. Cirginia. !!5. Festoration in the %onoran 0esert of California.
Festoration and -anagement .otes"($?
Bambridge, %.(., B.*. 1lein)opf, (. +allace, E (. Collmer.!&4. Comparative Dhotosynthetic
Droduction of -o>ave 0esert %hrubs. *cology 4#?&63,&6#.
Boers, Th -., 1. Kondervan, E @. Ben,(sher. !"#. -icrocatchment water harvesting for arid
zone development. (gricultural +ater -anagement. 3?3,6!.
Brotherson, @.0, E %.F. Fushford. !"6. /nfluence of crytogamic crusts on moisture
relationships of a soil in .ava>o .ational -onument, (rizona. Breat Basin .aturalist. 76?&6,&!.
Cluff, C.B. E F.1. =robel. !&". Catchment construction methods. +ater harvesting Catchment
and Feservoir Construction -ethods. Aniversity of (rizona, Tucson.
0utt, B.F. !". *stablishment of .aCl treated catchments. Fainfall Collection of (griculture in
(rid and %emiarid Fegions.=arnham 2ouse, p !.
*hrler, +.L., 0.2. =in) E %.T. -itchell. !&". Browth and yield of >o>oba plants in native
stands using runoff, collecting microcatchments. (gronomy @ournal &5? 554,55!.
*venari, -. !&4. =ields and Dastures in 0eserts? ( Low Cost -ethod for (griculture in %emi
(rid Lands. Bundesrepubli) 0eutschland? *duard Foether.
*venari, -.L. %hanan, E ..2. Tadmor. !&. The .egev.Boston? 2arvard Aniversity Dress p
=in), 0.2., 1.F. Cooley, E B.+. =rasier. !&6. +ax treated soils for harvesting water. @ournal
of Fange -anagement3#(#$?6!#,6!".
=isher, =.-., @.C.Ka), B.L. Cunningham, E+.B. +hitford.!"". +ater and nitrogen effects on
growth and allocation patterns of creosote bush in the .orthern Chihuahuan 0esert. @ournal of
Fange -anagement 7(4$?6"&,6!.
=reebairn, 0.-., %.C. Bupta, E +.@. Fawls. !!. /nfluence of aggregate size and micro relief
on development of surface soil crusts. %oil %cience %ociety of (merica @ournal 44($?"",!4.
Billis, -.(. !!3. /sraeli researcher planning for global climate change on the local level.
Bio%cience 73("$?4"&,4"!.
2all, (.*. !&!. (griculture in %emi,(rid *nvironments. pp7,#. %pringer,Cerlag, .ew Oor).
2arper, 1.T., E@.F.-arble. !"". ( role for nobascular plants in management of arid and
semiarid rangelands.pp 64,#!./n? (pplication of Dlant %ciences to Fangeland -anagement
and /nventory. -artinus .i>hoff;+.@un), (mserdam.
2ennessy, @.T., F.D. Bibbens, E -. Cardenas. !"7. The effect of shade and planting depth on
the emergence of fourwing saltbush. @ournal of Fange -anagement 6&($?33,37.
2owell, 0. !"!. 2ow to harvest water with microcatchments. Dermaculture 0rylands @ournal 4.
2owell, @.E0. 2owell. !!. +ater conservation in the home. Dermaculture 0rylands @ournal ".
1arpisca), -.-. !"". (rid Lands? Today and Tomorrow. pp3&6,3"5. +estview Dress, /nc.
Boulder, Colorado.
Lawton, 2.+. E D.@. +il)e. !&!. (griculture in %emi,(rid *nvironments. pp 7,& %pringer,
Cerlag, .ew Oor).
Lenz. L.+. E @. 0ourley. !". California .ative Trees and %hrubs. Fancho %anta (na Botanic
Barden. 363 pp.
Lunt, I.F., @. Letey, E %.B. Clar). !&6. Ixygen re'uirments for root growth in three species of
desert shrubs. *cology? 64#,6#3.
-eyers, L.*. !#&. .ew +ater %upplies from Drecipitation 2arvesting. /nternational Conference
on +ater for Deace8 !#& -ay368 +ashington, 0.C., A.%.(. pp 6!56!&,
.ir, 0. !&7, the %emi (rid +orld? -an on the =ringe of the 0esert. pp #&,&&. Longman,
Irev, O. !"". %ome considerations in the planning of runoff farming. 0esertification Control
Bulletin (#$?6,#.
%choll, 0.B. !"4. Cegetation and soil water response to contour furrows, dozer basins, and
surface additions of topsoil or power plants ash on arid coal spoils. Bridging the Bap between
%cience, Fegulation, and the %urface -ining Iperation. (merican %ociety for %urface -ining
and Feclamation. Drinceton, +estCirginia. pp 3&,335.
%hanan, L., E ..2. Tadmor. !&!. -icrocatchment %ystem for (rid Kone 0evelopment.
2ebrew Aniversity, @erusalem.
%hanan, L., E ..2. Tadmor, -. *venari, E D. Feiniger.!&5. Funoff farming in the desert. ///.
-icrocatchments for improvement of desert range. (gronomy @ournal #3? 774,77".
%palding, C.-. !5!. 0istribution and -ovement of 0esert Dlants. Carnegie /nstitute,
+ashington 0.C. 77 pp.
Tac)ett, @.L., F.+. Dearson. !#4. %ome characteristics of soil crusts formed by simulated
rainfall. %oil %cience!!(#$?75&,73.
Toy, T.F. 2adley. !"&. Beomorpology and Feclamation of0isturbed Lands. p 74,&. 2arcourt
Brace @ovanovich, %an 0iego.
+allace, (., *.-. Fomney. !&3. 0esert Dlant *cology. Fadioecology and *cophysiology of
0esert Dlants at the .evadaTest %ite. Ia) Fidge, Tennessee. 76!.pp
+est, .. !!5. %tructure and function of microphytic soil crusts in wildland ecosystems of arid
to semi,arid regions. (dvances in *cological Fesearch 35?"5,35#.