LEGUME TREES IN SEMI-ARID AND ARID AREAS PETER FELKER'
ABSTRACT - A review of the research and development activities on nitrogen fixing trees for semi-and regions is presented. In the period from 1980 through 1983 mueh new information on pod productivity, on natural stand nitrogen £tion and on greenhouse mineral nutrition has become available. Early results from small plot plantation work indicate a biomass productivity of 10.13 dry metrie tons per hectare per year. Clones have been made of promising individual trees and research on rooting of cuttings is rapidly approaching commercialization. Tissue culture propagation of semi-and nitrogen fixing trees has only just begun. Procedures have been developed employing subsoiling and herbicides that yield more than 90% survival on sites receiving 500 mm annual rainfail.
A tabular review is presented of the major research organizations and donors involved in semi-and nitnogen fi,dng trees. Major technical constnaints are coping with semi-and soils that are generaily low in available phosphonus, iron, Se and manganese, range in soil texture from cracking clays to high bulk density sands or rock>' desert pavement, and that vary in pH from 6.5 to 9.5. Pest management issues (insects and diseases) have not yet been studied iii semi-and nitrogen fixing trees and should be addressed before large sede plantations are undertaken.
The excellent progress made from 1980 through 1983 suggests a ver>' bnight future for semi-and nitrogen fixing trees.
Index terms: N 2 fixation, nodules, saline soils.
LEGUMINOSAS ARBÓREAS PARA ÁREAS SEMI-ÁRIDAS E ÁRIDAS RESUMO - Foi feita uma revisão da pesquisa sobre leguminosas arbóreas para áreas semi--áridas. No período compreendido entre 1980 e 1983, muita informação sobre produção de vagens, fixação de N 2 em condições naturais e nutrição mineral tornou-se disponível. Os resultados iniciais obtidos em pequenas parcelas indicam uma produção de biomassa de Prosopis entre 10-13 toneladas de peso seco por hectare/ano. Propagação vegetativa
por cultura de tecidos é uma área de pesquisa recém-iniciada para as árvores fixadores de N 2 de regiões semi-áridas. Por outro lado, clonagens de árvores individuais mais promis-soras foram feitas, e a pesquisa sobre o enraizamento de estacas está rapidamente chegan-do ao nível comercial. Técnicas de manejo já foram desenvolvidas, empreganchegan-do subsola-gem e herbicidas, as quais permitem mais que 90% de sobrevivência de mudas em locali-dades com cerca de 500 mm anuais de chuvas.
As principais instituições de pesquisa e financiadoras envolvendo árvores fixadoras
de N 2 para o semi-árido são listadas. Os maiores problemas técnicos a enfrentar são os solos do semi-árido que, geralmente, são deficientes em fósforo disponível, ferro, zinco e manganês, além de apresentarem ampla variação de textura, desde os argilosos aos are- Coilege of Agnicultuie, Tens, A & 1 University- Caznpus Box 218, Kingsville, Texas 78363-USA.
nosos e rochosos, e variaç5o de pH de 6,5 a 9,5. O manejo das pragas e doenças n3o tem sido ainda objeto de estudo, e deve ser investigado antes que plantações em grande escala sejam estabelecidas. Entretanto, o excelente progresso obtido no per(odo de 1980183 suge-re um futuro brilhante para as árvosuge-res fixadoras de nitrogênio, adaptadas às suge-regiões semi--áridas.
Termos para indexaçâo: fixaçâo de N 2 , nódulos, solos salmos.
1 NTRO DUCT lO N
Excellent reviews on tree legume in semi-and regions have become available as for example ou
Prosopis cineraria in the Indian desert (Leakey & Last 1980, Mann & Saxena 1980), on utilization of
mesquite in the United States (Parker 1982), ou Prosopis tamarugo in the Chilean salt desents (Habit
et ai. 1981), the utiluzation of Frosopis iii Latin America (Mendes 1982), and a review of tree legumes lii
semi-and regions (Feilcer 1980). This review will primarily deal with Frosopis since this genus seems to
be the one receiving the most attention in semi-add regions.
In recent years substantial new information ou tree legumes has become available iii areas as
diverse as tissue culture propagation, animal feeding with Frosopis pods, nitrogen cycling studies, new
silvicultural teclmiques, and selection of superior genetic planting stock. The first part of this paper will review some of these developments, while the second part will examine unresolved issues with impact on further developments.
PROSOPIS POD PRODUCTION AND UTILIZATION
Some of the most innovative work ou Prosopis pod utilization has come from Meyer et ai. (1982)
and Becker (1982) at the USDA laboratory iii Berkely, California. These workers have devised
commer-dai scale milling equipment (500 kg/h) and procedures for separating Frosopis pods into four fractions
consisting of: (1) exo-mesocarp (31% sucrose, 35% dietary fiber, 11% protein, 0% galactomannan); (2) endocarp (5% sucrose, 61% dietary fiber, 6% protein, 0% galactomannan); (3) endosperm sphts (0% sucrose, 0% dietary fiber, 8% protein, 60% galactomannan); and (4) the colydenon (0% sucrose,
0% dietary fiber, 56% protein, 0% galactomannan). Thus, Frosopis pods can be fractionated frito a high
sugir fraction, a high fiben fraction, a high protein fraction, and a high galactomannan or gum fraction (similar to guar gum).
'tire mii used by Meyer et aI. (1982) cair be adjusted to produce nearly 100% unbroken but
mechanically scarifled seeds from Prosopis.
Cattle feedíng was examined in five leveis of substitution for wheat bran by ground Prosopis pods
(Silva 1982). A randomized complete biock design with four animais per replicate and four runs was
used. No significant differences iii weight gain were observed, but the Prosopis ration was the Ieast
costly. A study of the progressive substitution of Prosopis pods for sugar cane molasses in the diets of
sheep was carnied out by Barros (1981). As Prosopis pods were so higli in sugar, this study was designed
to test the usefulness of Prosopis pods as an energy ration. Tire ration which substituted 50% of tire
molasses by Prosopis pods had tire greatest daily gain, but tire lowest conversion coefficient efficiency.
Another Brazilian study (Nobre 1982) examined tire effect of substituting Prosopis pods for whole
wheat flour on the milk production ofHolstein-Zebu cows, and reported that rations with 20,40, and
60% substitution of wheat flour by Prosopis flour were not signiflcantly (1' = 0.05) different from each
other, but yielded significantly greater milk fat and total solid productions than the ration without
Prosopis pods.
A feeding trial iii the Sudan attempted to determine whether a diet consisting solely of Prosopis
pods could maintain body weight of desert goats (Abdelgabbar 1983). The Prosopis poda were finely
ground and fed to goats in confmement at 55%, 70%, 85% and 100 0/ of lhe diet respectively. Cotton seed cake and wheat bran were used to flui out the remainder of lhe diet. The goats lost weight when mesquite poda constituted 85% and 100% of lhe diet; however, gains of 0.17 kg/week and 0.27 kg/week were observed when mesquite pods accounted for 70% and 55% of lhe diets, respectively. Uttle differen-ces iii the digestibility of dry matter, crude protein, crude fiber or elher extract occurred between lhe
rations. The major shortcoming of lhe 100% Prosopis pod diet seemed to be lack of palatibility ofthe
finely ground pods. Additional studies should probably examine pods which have been coarsely chopped to break lhe seeds rather than grind them fmely.
A comparison of pod production of North and Soulh American, Hawaiian, and African Prosopis
accessions in young Califomia plantations is iri presa (Felker et ai. 1984). This study examined 13
species and over 30 Prosopis half-sib families over a period of four years. In one fleld study pod
produe-tion was stimulated at the expense of total biomass producprodue-tion when irrigaprodue-tion was withheld from the
trees in a winter rainfail dlimate. P. velutina accessions from southem Califomia and northern Mexico
consistently out produced olher species such as P. alba, P. chilensis, P. pai/ida, P. Tarnarugo, P. glarsdu-lora vai. torreyana etc., at 2 to 3 years of age on all three sites. However, many Arizona P. velutina
accessions had little or no pod production during this study indicating the necessity of fleld tests for progeny from soulhem Arizona to identify superior pod producers.
NITROGEN FIXATION/NITROGEN CVCLING STUDIES
Some excellent quantitative data on nitrogen cydling in a Prosopis dominated desert ecosystems
have been published by a multi-disciplinary team of ecologists and soil scientists (Shearer et al. 1983, Sharifi et ai. 1982, Virginia & Jarrell 1983, and Rundel et ai. 1982). These studies were conducted in a hot (470C maximum summer temperatures) dry (60 mm annual rainfail) region where Prosopis uses a
groundwater source 3.5 to 5 m below the soil surface. Prosopis accounted for 90% of lhe total plant
cover at lhe site, but stffl had a canopy cover of only 30% of the total area (Sharifl et a]. 1982). In spite
of the low canopy cover, Prosopis had a pod production of 3,650 kg/ha' /yr' which is very high for this
low rainfail (Sharifi et ai. 1982). Rundel et ai. (1982) measured lhe nitrogen compartmentalization of lhe biomass productivity of leaves, branches, trunk and reproductive tissues, and estuinated that 23.36 kg/ha' N/yr" were flxed on a stand basis of 33% canopy dover, and Rundel et ai. (1982) sugges-ted lhat approximately 150 kgfha Njyr could be flxed if 100% stand cover were achieved.
Due to the difficulties in sampling 3.5 m to 5 m deep root systems for nodule observations,
Shearer et ai. (1983) made nitrogen fixation estimates based on determinations of natural abundance
'5N/' 4 N ratios. This technique is based upon lhe principle that lhe nitrogen fixation process does not
discriminate between lhe 1 r4 and 14 N isotopes and thus a plant which fixes ali its nitrogen should have lhe same 15 N1 14 N ratio as lhe atmosphere. On lhe other hand, soil denitrifying organisms discriminate
between '5N and 14 N leading to "high" sou '5N1' 4N ratios. Thus, plants which take lheir nitrogen
from lhe soil should have higher 1
3N1 14 N ratios lhan plants which Lix their own nitrogen. Using this
technique and carefully pairing nitrogen fjxing and non-nitrogen flxing piants, Shearer et ai. (1983)
estimated that Frosopis fixed approximately 43 to 61% of its nitrogen on 6 of the 7 sites they examined.
These authors also demonstrated significant differences iii natural abundance Nisotopesbetween legumes and non-legumes on 7 different sues. Although 1 5 N1 4 N determinations are costly, these determinations yield an integrated indicator of nitrogen fixation for deep rooted perennial trees that is not possible to obtain by other methods.
A comparison of soil properties under the Frosopis canopy with those between lhe Frosopis rows
was carried out by Virginia & .Jarrell (1983). These authors observed significantly higher total N, nitrate N, ammonium N, organic C, and sodium bicarbonate extractable phosphorus under the trees. There was
also lower saiinity under the trees evidently because Frosopis excludes sodium during mineral uptake
at the leaf and root surfaces. 'me water infuitration was also greâter under the canopy of Frosopis than
between the trees. When the results of this multi-disciplinary study are taicen together, they indicate considerable potential for productivity for either food or fuel, for nitrogen fixation, and for soil impro-vement.
In addition to work ai the ecosystem levei on nitrogen fixation of semi-and tree legumes, some useful greenhouse experiments have also been published. Felker & Ciark (198 la) added another nitrogen
fixing tree to lhe list of and adapted trees: Olneya tesota (desert ironwood), and failed to observe
nodula-tion on Cercidium floridium and Atada greggt Cercidium occurs on vast areas in North and South America and its lack of noduiation on infertile soils seems out of place. Also in greenhouse experiments
it was found that Frosopis articu!ata and P. tamamgo growin sand culture (albeit very slowly) at salinities
equivalent to sea water (3.6% NaU) on a nitrogen free medium (Felker et ai. 1981). Soybeans and beans
(Fhaseolus spp.) tolerate salinities of 0.1% with difficulty and lhe most salt tolerant common legume, alfalfa, tolerates 0.6% with difflculty.
Nodules are seldom found on tree legumes in semi-and regions despite the fact that many of these tree legumes have been shown to nodulate in lhe iab or greenhouse. Felker & Clark (1982) simuiated a phreatophytically grown semi-and tree legume by growing ii iii a 3 m long column in the greenhouse.
With only one watening from the top and ali subsequent watenings (1% years) from below, lhe Frosopis
deveioped ali of its nodules at least 2.75 m from lhe soil surface. Funthermore, this legume fixed
nitro-gen aI leaf ah temperatures of 47 0C and leaf xyiem water potentiais of minus 33 MPa (Felker & Clark 1982). These environmental parameters are cleanly more extreme than can be tolerated by annual le-gumes.
MINERAL NUTRITIONAL REQUIREMENTS OF TREE LEGUMES
There are mililons of hectares of tree legumes growdng unmanaged in lhe semi-and regions of the
wozld (Felker 1980). The two major management practices for improving the productivity are: (1) to
correct mineral nutrient deficiencies affecting nitrogen fixation and growth; and (2) to improve lhe stand by eliminating inferior trees and replacing thêm with genetically superior stock. Before removing existing trees it wouid be useful to have a method for assessing nutrient deficiencies, for example, via folian analyses.
Reyes & Felker (manuscnipt iii preparation) recently examined lhe growth of Leu caena leu co
cc-phala K8 in a lime, phosphate, micronutrient factorial greenhouse, experiment designed to optimize biomass production. Leaf tissue macro and micronut.nient anaiyses were signiflcantly correlated with piant dry weights. ForPlevels this correiation was positive and for Na leveis negative. Biomass production Pesq. agropec. bras., Brasilia, 19 sfn: 47-59, jun. 1984.
was also negatively correiated with soil p11 (over the range of 4.9 to 7.2). Leal nitrogen content was not correlated with biomass production indicating that these piants were fixing suffieient nitrogen. Tabie 1 lists the optimurn leaf tissue parameters we observed for Lwcaena biomass produetion. These vaiues may prove usefui in diagnosing mineral deficiencies ira the field.
flue to the importance of phosphorus for legume growth and nitrogen fixation, one would expect mycorrhizae to benefit tree legumes on semi.arid soils since they are typically neutral to alkaline in pH with low phosphorus leveis.
Major nutrient isnba!ances occurred with leaf sodium leveis of 0.005 to 0.011% and greater, and when leaf phosphorus leveis were below 0.07%. Leal manganese leveis were sensitive to pH and were typicaily 40-50 jig/g at pH values greater than 7.
Preliminaiy results of a fieid Mal with Leucaena leucocephala K67 at 0, 60 and 120 kg/haP with
and without VÁ mycorrhizal (Glomus fasiculatus) inocuiation revealed that phosphorus fertilization had
little effect on growth whlle mycorrhizal inoculation nearly doubled the growth of three month old transplanted seedlings (Mbugua & Rhodes, unpublished observations). No differences ira leal tissue phosphorus concentrations were observed, but aimost double the quantity of phosphorus occurred ira the mycorrhizal seedlings (Mbugua & Rhodes, unpublished obs). Á poor soil (white clay with very low water infiltration) was chosen for this trial and ptrhaps mycorkhizae will have no benefit on better soil types. This preliminary experiment needs to be repeated on several sites over a period ofyears.
FIELD PRODUCTIVITV STUDIES
A. Natural stands
Ás mentioned earlier, Sharifi et ai. (1982) have reported fite pod productivity of a native Prosopis
glandulosa var. torreyana stand in the California desert to be 3,650 kg/ha for 30% coverage. 11ese authors suggested that higher productivities would be possible since the stand density is primarily limited by establishment rather than water availabiity from the 4 meter deep water tab!e.
TABLE 1. Leaf tissue nutrient conccntrations associated with optimum biomass production
ofLeucae-na leucocephala 1(8 in fite greenhouse (data from Reycs & Fetker manuscript lia prepara-tion).
Nutrient Leal concentration
phosphorus 0.16% nitrogen 4.3% potassium 1.5% magnesium 0.18% sodiunt 0.001% calcium 0.16% iron amo 22 ii g/g manganese 95 JgIg copper 2,29919
In contrast to this high productivity, Braun et ai. (1978) estimated a net primary pod productivity
of 1,60 kg/ha 1 /yf 1 for a Proso pia flexuosa dominated community lii a 260 mm annual rainfali region
of Mendoza, Argentina. This was a matute climax vegetation site and perhaps greater yields would have been obtained in younger growth stages.
B. Plantations
A comparison of the biomass production of 32 Frosopis accessions representing North and South
American, Hawaiian and African germplasm, under three different irrigation regimes, has just been
published (Felker et ai. 1983). 'fie Prosopis accessions were watered when the soil water potential at
the 30 cm depth reached either 60, 200 or 500 kPa. After three seasons growth there was little differen-ce in biomass production ainong irrigation treatments for the fast growing acdifferen-cessions ofP. chilensis and
P. alba. However, there were large differences in dry matter productivity between the accessions of P. chilensis from Argentina, showing the highest yield (40 t/h&1 c3/yr'i), and the leafless P. kuntzei also
from Argentina, the lowest.
The total irrigation plus precipitation received during the three seasons was 1,390 mm and therefore a water use efficieney of 345 kg per kg of dry matter was calculated (Feilcer et ai. 1983), which compares very favorabiy with C4 grasses like sorghuxn. A similar yield (14 t/hi'/yr' ) was estimated forltosopis leaf and woody biomass production irrigated at 30% of pan evaporation in the Cailfornia Imperial Vailey (Felker et ai. 1983). A screening trial of 55 tree legume accessions carried out in the Imperai Valley showed a 100 fold range in biomass production among the accessions.
The cold tolerance of Prosopis selections representing high biomass producing iines, such as P.
chilen ais, P. alba and P. articulata, were compared to mesquite native to California, Arizona, New Mexico
and Texas (Felker et ai. 1982). Prosopis juliflora from West Africa and P. pallida from Hawaii were
observed to be truly tropical species tolerating maximaily - 1.5 °C. Most P. alba, F. chilensis and P.
articu-lata accessions had greater biomass productivity, but less cold tolerance than species in native North
America. However, sufflcient variability existed to substantiafly improve Prosopis biomass productiori in the United States.
Practical fieid management techniques are being deveioped which avoid irrigation using the promis-ing ciones identified in California field triais (Felker et ai. 1984). A heavy duty mechanical transpianter was modifled from a subsoiler to plant 38 cm long seediing containers entirely beiow ground levei. Five
herbicides at a high and a low rate were examined for use in transpianting Leucaena and Prosopis.
Oryzalin (2.8 kg active ingredient per hectare) gave a 3-4 foid increase in biomass production
ofLeucae-na and Prosopis, and increased seedling survival from 75 to 95%. This experiment was carried out with only 150 mm rainfail within 110 days after transpiant and ifiustrates the importance of weed control iii semi-and systems (Felker et ai. 1984).
A field tnial was recently conducted to compare growth and survival of Prosopis and Leu caena seedllngs using two seedling containers (Felker & Smith, unpubllshed observations). A 20 cm long by 3.8 cm wide plastic dibbie tube that is remcved prior to transplanting was compared to a 38 cm long cardboard container that was not removed prior to transplanting. An adverse site with a ciay soil that Ind a 10w water infiltration rate was chosen.
The sunvival was 100% with cardboard containers and 73% and 46% with plastic dibble tubes for
Pra sopis and Leucaena, respectively. However, the Leucaena seediings, which survived when transplanted
without a surrounding container, had much greater biomass than seedlings which still had lhe container around the root system. The contrast, Prosopis seedlings had greater biomass production when the card-board containers were dili Ieft on. These differences may be dueto the fact lhatFrosopis has a tap root while Leucaena has a fibrous root system that is dose to lhe soil surface.
• Our observation has been that lhe most expensive planting system that excludes partial ar hill irrigation is approximately ten times cheaper than lhe most inexpensive irrigation system. Thus, we have chosen to use long (38 cm) seedling containers, disking, subsoiling, herbicides and mechanical transpianters capable of rapidiy pianting long root systems. Iising these techniques our Frosopis pianting survival was greater than 90% on each of lhe five sites we planted izi the last 2 years.
VEGETATIVE PROPAGATION
The breeding mechanisms of semi-and tree legumes dictate whether or not lhey will propagate Ime to lype from seed. Leucaena leucocephala is a seif-fertile, highiy inbred piant that produces copious quantities of seed aI one to Iwo years of age (Brewtaker & Hutton 1979). Thus, isolated seed orchards
ofLeucaena are the method of choice for providing bulk quantities of Leucaena propagules.
In contrast, Frosopis and many Acacia species are self-incompalible (Siinpson 1977) and are therefore obligate outcrossed species that cannot propagate true lo lype from seed. In addition, some of lhe best biomass producing F. chilensis species do nol begin lo produce seed until they are five years of age (Feilcer etal. 1984).
Thus, rapid (and hopefu]ly inexpensive) vegetative propagation techniques that produce Irue to type seedlings are required. Six aí lhe tested species sometimes rooled weli if young greenhouse grown stockplanls were used (Feiker & Ciark 1981b). However, these workers noted lhat mature fieid trees yieided 1% or lesa rooted cuttings and that even greenhouse grown stock pianls yielded 10% or lesa rooted cuttings in the winter. Thus, lhe effecl aí environmental parameters on lhe rooting aí cuttings was examined with lhe use of growth chambers. Kiass (manuscript in preparation) found that cuttings wouid nol rool aI 20°C, they wouid root poorly aI 27 0C and the optimum temperature was approxi-mately 35 °C. The rooting percentage was 6 to 9% at 9,000 iumens/m 2 (900 ft candles) and 70% at 25,000 iumens/m2 (2,500 ft candies) when fluorescent Iights were used. Photoperiods (8, 12, 18,24 h) did not exerl a inarked influence on lhe rooling percentages, aithough lhe 12 and 18 li photoperiods were sligjilly better lhan lhe 8 and 24 li photoperiods.
When lhe optimal !ighl, lemperature and photoperiod were achieved indolebutyric acid and napthaieneacetic acid powders, in the range aí 0,3% to 3,6%, did not signiflcantly increase the percenta-ge rooting over a contrai aithough they did increase lhe number aí roots per cutting (IClass manuscript iii preparation). This indicates lhat environmental factors have more influence on rooling aí cuttings than do exogenousiy supplied auxins.
Ànother altemative lo vegetative propagation is lo produce plantiets via tissue cuiture propagation. Iii arder lo ensure lhe exact reproduction aí lhe desired parental lype, lhe pianliets shouid be derived from lateral ar apical meristems rather lhan undifferentiated calius. While il is often reiatively easy to produce tissue cuitured piantlets from seedling hypocotyis it is obviousiy nol possible to clone desirable field lrees from hypocotyi explanis.
The laboratory with the most experience to date on tissue culture of semi-and tree legumes is that of Dr. H.S. Aiya of the Botany Department at the University of Jodhpur, India. Dr. Arya's
labora-tory has worked on the tissue culture ofAcacia, Frosopis, Tecomelia and Zyzphus.
Research on the tissue culture propagation of Prosopis alba, F. tamarugo and Leucaena
leucoce-phala at Texas A & 1 University developed for initial axillary bud explant sterillzation, and hormone
formulations have identified that routinely yield leaf proliferation. In about 15% of the cases single shoots of 2 cm maximum length have been produced. Multiple shoots have not yet been produced nor have any of the shoot explants been successfully subcultured. As tissue cultured plantlets can be pro-duced over 1,000 times more rapidly than cuttings oncegoodtechniques have been developed, it ismanda-tory to continue this une of research. However, one must be very patient, for results in this area do not come quickly.
Wood technology and harvesting
Some of the and/semi-and tree legumes species have superb wood qualities. Both desert ironwood
(Olneya tesota) and Texas ebony (Pithecellobium flexicauli) are more dense than water and possess an
almost black colored heantwood. Wood supplies of these trees are very limited and are in considerable demand by wood carvers, custom fumiture makers and hobbyists.
Fine furniture craftsmen in both Argentina and the United States have come to appreciate the excellent technical qualities of Frosopis wood. In a recent Frosopis workshop, Rogers (1983) compared mesquite wood properties with five other hardwoods and found in to have 25% of the volumetric shnink-age of the other hardwoods and a hardness 50% greater than any ofthe other hardwoods.
The semi-and lands are somewhat unique in their potential to produce large scale firewood plant-ations because of the low population density and extensive areas that are suitable to biomass farming. With advanced technical packages yields of 10-15 metric tons of wood per hectare per year have been obtained. Felker (1984) estimated that wood produced from semi-and energy farms might cost as little asUS$ 28per drymetricton or US$ 1.40perGflUS$ 1.48permillionBtu).
Technical constraints to further development
Since genotypes with good biomass productivity and water use efficiency have been identified, it is important to develop field management techniques that will enhance nitrogen fixation of the farm
or iii the plantation.
Soil physical properties may exert profound influences over the growth of nitrogen fixing trees. Low productivity is to be expected on shallow (30 cm) soils over clay or calcium carbonate hard pans, on gravei soils, on clay solis with low water infiltration rates, and on sandy soils that have not been subjected to deep piowing to reduce bulk density and increase root penetration. Sandy soils often appear to the inexperienced observer as though roots should have little difficulty penetrating them, but sands have the greatest bulk density of all soil types. In an excellent monograph, Charreau & Nicou (1971) have demonstrated for millet and groundnuts linear regressions between crop yieid, root density and soil bulk density iii the sandy soils of West Africa. There is no reason to believe that root develop-ment (and above ground biomass) of nitrogen fbdng trees would not also be restricted by high soil bulk density in semi-and sandy sons.
Soil chemical properties profoundly influence nitrogen fixation in annual legumes and would also be expected lo influence nitrogen fixation in trees. In contrast to the wet tropics, lhe soils of semi-and regions of the tropies have pH vaiues generaily greater than seven with high base saturation (Dregne 1976). Under these conditions, phosphorus, iron, zine and manganese could ali be expected to limit cnop production. While considerabie data are available on soil fertffity leveis of rainfed cuitivated agricuiture sites, which have generaily reeeived some fertilizer axnendments, there is virtuaily no fertility data base for
lhe non-cultivated soils supporting mixed Acacia/Prosopis stands iii the semi-and regions ofAfnica, India,
Nonth America or Latin Amenica. Recent reviews of Afnican soils (Dregne 1976, Ahn 1977) gave no vaiues or references for sodiuzn bicarbonate extractable phosphorus ieveis which are generaily recognized as lhe most usefui soil phosphorus test for alkaline soils (Olsen eI ai. 1954). Sodium bicarbonate extractabie phosphorus leveis for three non-fertllized rangeiand sites in lhe United States were iess than 2 p g/g (Ei-Ghonemy et ai. 1978, Biack & Wigjit 1972, Virginia & Jarreil 1983) and thus considerabi below lhe 10 pg/g value to which a response to phosphate fertilization is certain for annuai crops (Olsen
et ai. 1954). The responses of native stands and plantations ofAcacia and Prosopis to phosphorus and
trace element fertilization need to be measured and related to ieaf tissue leveis. Baseline data (maps) on soul phosphorus, pH and trace eiement concentration lhroughout lhe non-cultivated semi-and iands aiso need te be established.
Given lhe low concentrations and low avaliabilities of key nutnients for nitrogen fixation it is critical to use them as efflcientiy as possibie. Mycorrhizae enhance uptake of immobile piant nutrients,
as discussed earlier, stimuiates lhe gnowth of young Lwcaena seedlings. As various species and strains of
mycorrhizae occur naturaily it wouid be usefui lo screen various sources for their ability te stimuiate gnowth at iow soil nutnient concentrations.
CONCLIJSIONS
Much progress has been made since 1980 te deveiop lhe resources of nitrogen fixing trees in semi. -and regions. Seed cieaners have been identified that produce 100% unbroken seed from pods fed aI the rate of 500 kgfh; firm ecosystem nitrogen fi.xation data have become available demonstrating 30-40 kg N/hi 1 /yf 1 fixed in the California desert; haiophilic nitrogen fixing trees have been identifled that grow on 50% sea water, and clones have been identified lhat produced 50 kg of dry pods/tree in
two seasons. Vegetative propagation of Frosopis by rooting of cuttings is approaching commercialization,
and tissue cuiture propagation is seriousiy underway. - A very refreshing increased awareness of lhe probiems and potentiais of nitrogen fixing trees in semi-and systems is now evident from university researchers, nationai forestry directors and intemationai donors.
Semi-and nitrogen fixing-tree species possessing adequate nitrogen fixation and biomass production have been identified on selected sites and reasonabiy good stand establishment procedures that avoid irrigation at transpiant are now availabie. Research is now requined lo determine lhe iong tenm producti-vity of lhe seiected strains and te determine lhe growth responses on a vaniety of soU types with and without fertilization and wilh and without mycorrhizai inoculation.
Ver)' exciting genetie material for woodfuel and pod production has been identified. Great technic-ai strides have been made in the last few years and more people than ever before are engaged lii research and development of these trees, and funding agencies are looking more and more favorably upon requests
for research and development itt this area. The future for semi-and nitrogen fixing trees looks very
bright indeed.
ACKNOWLEDGEMENTS
This review paper has borrowed heayily on fite unpublished research results of the Texas A & 1 team including (alphabetically): Dr. Richard Beck - tissue culture, Dr. Yashpal Goyal - tissue culture,
Sharon Klass - rooting of cuttings, David Mbugua - Mycorrhizae studies, Peter Oduol - Frosopis pod
studies, Isidro Reyes - Leu caena, Dwigth Rhodes. mineral nutrition, and Don Smith - field manager. My
gratitude to them is gratefully acknowledged. The fmancial support provided by the U.S. Department of Energy, subcontract 9066,by the U.S.Agency for International Developnient,Grant DAN 5542.GSS-2099--00, and by the Caesar Kleberg Wildlife Research Institute, is also gnatefuily acknowledged.
REFERENCES
ABDELGABBAR, A.I. Feed utilization and body composition of Sudan desert goats fed on mesquite (Prosopis
chilen-rir). si, University of Kartoum, 1983. 130p. Tese Mestrado.
A1-IN, P.M. SoU factors affecting rainfed agriculture lii semi-and regions with particular reference to the Sahei zone of Africa. Ia: PROCEEDINGS OF AN INTERNATIONÂL SYMPOSILIM ON RMNFED AGRICULTURE IN SEMI--ARII) REGIONS. Riverside, California, University of California, 1977, p.128-65.
ARMITAGE, EM.; JOUSTRÀ, P.A. & SALEM, B.B. Genetic resouices of tree species In and and semi-and areas. Rome, FAO, 1980. ll8p.
BARROS, N.A.M.T. Efeitos da substituição progressiva do melaço por vagens de algaroba (Frosopis/uli)7ora (s.w.D.C.)
na alimentaçfo de ruminantes. Joib Pessoa, Univ. Eed. da Paraíba, 1981. 97p. Tese Mestrado.
BECKER, R. The nutnitive value of Prosopis pods. In: MESQUITE. Utilízation. si, s.ed., 1982. Proceedings of a
Symposium at Texas Tech. University, Coilege of Agriculture, Lubbock TX, M1-M9, 1982.
BLACK, AI. & WIGHT, 3 .R. Nitrogen and phosphorus availability ia a fertllized rangeland ecosystem of the northern
great plains. J. Range Manage., 25:456-60, 1972.
BRAUN, R.Il.; CANDIA, Ri.; LEIVA, R.; PAEZ, M.N.; STASI, C.R. & WUILLOUD, C.F. Productividad primaria aerea neta dei algarrobal de Nacuaan (Mendoza). Deserta, 5:7-45, 1978. Published by Instituto Argentino de Investigaciones de Zonas Andas (IADIZA).
BREWBAKER, J.L. & HUnON, Ei{,Leucaena-t'ersatila iropical treo legume. In: RITCHIE, G.A. cd Ncw agnicultural
crops. Westview Presa Boulder Cobrado, 1979. p.207-59. (Am. Assoc. Ady. Sci. Symp., 38).
CHARREAU, C. & NICOU, R. L'amelioration du profil cuitural dans les sois sableux et sablo-arglleux de la zone tropical seche Ouest Africaine et ses incidences agronomiques, L'Agron. Trop., 26(5):565-631,1971 -
DREGNE, H.E. Soils of and regions. New Yorlc, Elsevier Pubi. Co. 1976. 237p. Pesq. agropec. bras., Brasilia, 19 s/n: 47-59, jun. 1984.
EL-GRONEMY, A.A.; WALLACE, A. & ROMNEY, E.M. Nutrlent concentrations ia the naturalvegetation of the Mojave desert. Sol! Sci., 126(4):219-29, 1978.
FELKER, P. Economic environinental, and social advantages of intensively xnanaged short rotation mesquita (Prosopis spp.) biomass energy farms. Biomass, 1984. (In press).
FELICER, P. Uses of tree legumes ia semi-and regions. Econ. Bot., 35(2):174-86, 1980.
FELKER, P.; CANNELL, Gil.; CLARK, P.R.; OSBORN, J.F. & NASI-I, P. Biomass production ofProsopis species
(vlesquite), Leucaena and other leguminous trees grown under heat/drought stiess. Forest ScL, 29(3):592-606,
1983.
FELKER, P.; CANNELL, Gil.; OSBORN, J.F.; CLARX, P.R. & NASH, P. Effects of irnigation on biomass production of 32rrosopis (mesquite) accessions. ExpI. Àgric., 19:187-98, 1983.
FELKER, P. & CLARIC, P.R. Nodulation and nitrogen fixation (acetylenereduction) indesertinonwood (Olneya tesota).
Oecologia,48:292-3, 1981a.
FELKER, P. & CLARX, P.R. Position of mesquite (Prosopts spp.) nodulation and nitrogen fixation ia 3 m long
phrae-tophytically siznulated soll columns. flant and Sou, 64:297-305, 1982.
FELKER, P. & CLARK, P.R. Rooting of mesquita (Prosopis) cuttings. J. Range Manage, 34:466-8, 1981b.
FELKER, P.; CLARK, P.R.; LLAG, A.E. & PRATT, P.F. Salinity .to!erance of the tree legumes mesquite (Prosopis
grandulosa vai torreyana, E. velutina, and?. articulata), alganrobo (P. chilensis), Kiawe (1'. pallida), and tamaxugo (P, tamarugo) grown ia sand culture on nitrogen fnee media. Plaat and Soil, 61:311-7, 1981.
FELKER, P.; CLARK, P.R.; NASEI, P.; OSBORN, J.F. & CANNELL, G.H. Screening Prosopi: (mesquite) for cold
tolerance. Forest Sei., 28(3):556 -62, 1982.
FELKER, 1'.; CLARK, P.R.; OSBORN, J.F. & CANNELL, G.H. Prosopis pod production a comparison of North
Amenican, South American, Hawailan and Mrican germplasm ia young plantations. Econ. Bot., 1984. (la press). FELKER, P. & ROBERSON, S. Mesquite woodworldng proceedings of a wokshop on the cutting, drying, and
fabrica-tion of mesquita wood into flooning, handcrafted items, and furniture. lcingsville, TX, Caesar Kiebeng Wildlife Research Institute Fublishers, 1983. 35p.
FELKER, P.; SMITH, D.; SMITH, M.; BINGHAM, R.L. & REYES, 1. Evaluation of herbicides for use la transplanting
Leucaena leucocephala and Prosopis alba on semi-aiid lands without irrigation. Forest Sei., 1984. (In press).
HABIT, MÁ.; CONTRERAS, D. & GONZALEZ, R.H. Frosopis tamatugo: fodder tree for and zones. Roine, FAO,
1981. liOp. (FAO. Plant Production and Protection Pape; 25).
LEAKEY, R.R.B. & LAST, F.T. Biology and potential of Prosopis species ia and environmental wlth particular
referen-cato?. cinerarla. J.Àxid Eaviron., 3:9-24, 1980.
MANN, H.S. & SÀXENA, S.K. Kliejri (Frosopis cinerafla) ia tke lndiaa desent - 1h role la agroforestxy. Central Ad
Zone Reseaich Inatitute, 1980. hp. (Jodhpux Monograph, 11).
MENDES, B.V. Simpósio Brasileiro sobre Algaroba 1.Natal, RN, EMPARN, 1982. 407p. (Document, 7).
MEYER, D.; BECKER, R. & NEWKOM, H. Milhing and separation of Frosopis pod components and their application
ia food products. In: PARKER, H.W. Mesquite utilization 1982. Lubbock, Texas, Texas Teeh. University. Coilege of Agricultura, 1982. L1-L12. Proceedings of a Symposium.
MIJSTAFA, A.R.F,M. Growth responses of Pra soplschilensls (molina) atunta seedlings. Wad Medani, Sudan. University
ofGezira, 1983. 129p. Tese Mestrado.
NOBRE, E.V. Àlgaroba na alimentação de vacas em lactação. João Pessoa, Univ. Fed. da Paraíba, 1982. 74p. Tese Mes-trado.
OLSEN, S.R.; COLE, C.V.; WATANABE, F.S. & DEAN, L.A. Estimation of available phosphorus lii soils by extractio'i with sodium bicarbonate. USDA circ., 939:1-19, 1954.
PARKER, H.W. Mesquita utilization- 1982. Lubbock, Texas, Texas Teci. University, 1982, 245p. Proceedings of a Symposiwn.
PEDERSEN, 11.0. & GRÀINGER, A. Bibliography of Frosopis. Int, Tree Qops J., 1(4):273-87, 1981.
RUNDEL, P.W.; MLSEN, E.T.; SUARIFI, M.R.; VIRGINIA, RA.; JARRELL, W.M.; KOUL, Dii. & SHEARER, G.B.
Seasonal dynamics of nitrogen cyclirig for a Frosopis woodland ia the Sonoran descrt. Plant and Soil, 67:343-53,
1982.
SHARIFI, M.lt; NILSEN, E.T. & RUNDEL, P.W. Biomass and net primary production of Frosopis glandulosa (Fabaceae)
ia thé Sonoran desert of California. Amer. J. Bot., 69:760-7,1982.
SIIEARER, G.; KOHL, Dii.; VIRGINIA, R,A.; BRYAN, BÁ.; SKEETERS, J.L.; NILSEN, E.T.; SIIARIFL, M.R. & RUNDEL, P.W. Estirnates of N a fixation (som variation in tio natural abundance of 15 N ia Sonoran desert ecosystems. Oecologia, Berlin, 56:365-73, 1983.
SIDDIG, M.M.B. Phenology and regeneration of mesquita (Prosopis chilends) ia the Sudan. Wad Medani, Sudan,
University of Gezira, 1983. 94p. Tese Mestíado.
SILVA, D.S.D. Substituição do farelo de trigo (Triticum vulgare komarnistsky) pelo fruto triturado da algarobeiza
(Prosopis juliflora S.W.) na alimentação de bovinos de corte em confinamento. João Pessoa, Univ. Fed. da Paraíba, 1982. SOp. Tese Mestrado.
SIMPSON, B.B. Breeding systems of dominant perennial plants of two disjunct warm desert ecosystems. 0eco1oa, Berlin, 27:203-26, 1977.
ULICU, W.L. Economics and technology of harvesting whole-tree mesquita. In: PARKER, H.W. Mesquite utllization 1982. Lubbock, Texas, Tens Tech. Univ. Coilege of Agricultura, 1982. B1-1313. Proceedings of a Symposium. VIRCINIA, R.A. & JARRELL. W.M. Soil properties ia a mesquita dominated sonoran desert ecosystem. Soil Sci. Soc.
Am. J., 47:138-44,1983.
APPEN DIX
ORGMOIZATIONS PRIMARILV IMVOLVED IN SEMI'ARID TROE LEGUME WORE
0000000.01 *4468 08011111080
0o.J. AI.., *68...' 0141 á Do,. *004. 0.aOo,Io. UnI,..! 5.04v. 91.1* knOnid,olIol Ao.,pio Ao..S
E.aa.t. 84010 900 *020. 0.00l
Do. doy. 0.94.0010(10, UnI,. .1000npOq.J61.pl., 01,418 Ti...0Itl,l. 6.,4..PhopoácIOaho
Do. 0. 0.140. Enoy 1.4468.1,7.0.10004. LM4n 01 CM8 L,,.00.y ....,h Pl..wà, .40.0!.
Cosi. 1140.OooiI4.00,CO34S
68.0. 0o.0.o U!D&W*4C 000 8004110. 8*, 0.,SW. N00101#o 00 000 10011110 811104 0.01*01.. 90.80.. 011117.
CAO47IO
68.04.0.0.,,.. U.I.Agnvf,8 101. DonIno,..nIS6TIPNO D000o.o.00y o*dooa
D.pOol 0100. W.hI.ø4a. 0.0.20823
818. 18111 0.,*o Lo0.ni. WId Ro1101 2000 P*ok *0..t 91.1.6868100 E..' & 400.0 La04068. AoS, '1080!.
N.noY.,104.Y.10810 *1010601,401
D...LL 8n.ot.k. M, 9101 ,. 10*. *804.90. Eoo 802 0..o.* 8000111004..
WSma.I.. III 08786
Do UM8.. 00.110 91.8001900004.0000 MInodC.io0 A,. lo. Polo. Doed.. PISA 86.0(00 1080 o.
Edil. C,'s,da 04(08.1083.14.001.10
Do. wo1 D0.No*10.. C40000.0915TOM 0.9.1300 Dás. 8a•o•I *4t4& 0811006 C100!(llnA .40.018
Do,P. FolIa Tlo A & 1 Uo*.Sny. Ca,o.a 01 310 94014608. ...01, 0108p8 Loto...
kI1,Q0141I90 TIO 70262
PoliciA. 0.800 80400*9*. Ed. 81100! C.nw 14.01,010 30030 P1.OMIWI000000....0d0 Pr000k. dias. L.o..o.
801 -0. 8. 804. 0. 00? 80 70232 awill.. DO, E,..11
P.E.J. Dono Wo,ld 8..k 1010H. 00.005.01. 0080000000v 0,1008
Wk8,m, 0.0. 20423
Do, 04.001.0 5.01..S 01.4.01,0100181682101 0.i,ø18s4., R.00á.fl
Ao. 01.01810.015.0.0.20418
08,0. Oto.4 Cal,. 5.1100.1 do 50*04*8 Fo,.oIk.. 91.0.00100 ..0á4410000i1 Ao.o4.
0.9. 2312 D01.o'Has. 000
Do. A. 008100,4 90,1.0 0000.,ot, 0.niío 9.0.800011 PISA 01010 .44.06.01.,!.
000.010811*. SvSo
Do. OH. HwoI18r D.1,.,10l,s.00 do Cloolda 01.1000.0 T.00,OOIY cyloIoçv 910.0(40
UoI10nId.d do 040101 AI0 1428
0,08,10*11... *8600108
Do. 968 M,.I.y ICOAF 9.0.8.. 20077 5*0184, K.inv. *0°á, o...011404060010r 104080
Do.W. .iwd0 D.pi.8o# & 60v. 0,4.. UO10. 081. 5Io.sd. NI,oç.o .1,410.008080010 .4ao', C.oáo..RoopIo
01.00140.0* 84710
W. .100,844*1180008,,- 0a.v.840 5.1, 66. CAIE 04091,00*6. 910,0.040. 68101.01,-O I,14m1118.Ood.s Ao.c!., P11000I.
OSoo 00,4, 50 10810
8*. LO. Lo..d 0000010010.,S 0.40. 9100. 0.ot* 0.. 8508 Doo., 000010v 001008
0,1068.010280-60.4.
Do. 0.9. 1.01h00; 0.00018*4.001100-do 10040401 *9901.411014
w. lodia
Do. 04. Mo.,Io 908,101.0,040861 DS0I. Doai 00.1000. PISA 04.1010 odd 68108 000896.44005
1.0000,001,0, 801,90
Do. 8.14. SIi,tSt 51,1810(01140. Ou. I..,loa. 8,-d109001.1d 841400110014010000.01 000.0(4
90.40.0415822 016.800,0, 00081001,00. SI.
0. '.0,6.,. 9.,..0 Dsp.. 9*0 OS doo. 10.0.1. R....044.o,d50400 0.00101.40 1011.00100 004008
0081181.08100 0oí, losy
818,91.000.00 To... 9,8.009.4* Lá. 8400410. TIO Wood 104000100v P,o.opk
100±1011168 Noolond PIo,a0.1.8,..8640 041,4.01404. AçI.dnn.. 6 D.o.dooIo P0.o0.OI000.008 ?lo..ph
Coshi.2008'0i.lto. Eoado,
08.00000414 Opnøoo DoloSo Hwo.i,P.0. 0" 072 9106804..flSlS.,a0 0..!., 08a*i., 01080,,
Po,110J'PrI68, 04.401
0 ,4*.,. 011010.. Not TIods 051088. Ao *0410 00,800.102' XS.p. L.So.t.,y 60118468.1*
Veia,.. 0400100
100. Tonis,TolIo R. 6.8. V. 91001* AoOdo. *00 0o4d.00 M.o.g.n.0000010010o 068* P000pM; .40.06.108...
08.04.8. To,,- 01400008115482.0.011.0., 0h11. RM0001. 000.106 0001000. *otpà aOS
Do. J.W.Too,h'II COlHO øhdSooPoo..t 000.P0.8004088 O- .,lI.odoo á Innooy A0'S. 80...di*
0.6.,. *0.1. 2000'A*tvSI.
0. Solo,. VoMido 0Jo1n.Sdod do Uni.' 140,.. P.on, .4068010110400. 900095
0,q. 9.40484688114.11. D.00ttooalo d.CSooS. A,-1. 000dooIIoolI000 1100000
U044.004dod 1840018 do 0008,806 048004410310001004704.0010.0t8.. *1*000110
0, G.E.91100.,. OoyoI 00100100*0000.00.000400.1.4.001 1410,40,,. '.014,, .0 0,10 lalA 01.010 .004.1.
8041.,dlWO2AO
Do. 90w 1 Wo.d 0,nii,.o.o.SOh Pó.., 0,4.0,004 0.0*4 boA 'Saibo. 0.01.101,80400 Aflos. Ao,,p(.
000001 001 291081101*4
1,0. Lo'. 15.000. 0.411. 2908 0001 90 .0001180, 0h11. r.o,aol0001an.14.. 14089!. 60080000
