Previous Page Table of Contents Next Page

Chapter 8
Management of Silvopastoral Resources

P. juliflora Management at the Brazilian Northeast

Paulo César Fernandes Lima

embrapa-Agriculture and Livestock Research Center for the Semi-Arid Tropic (cpatsa)
Petrolina, Pernambuco


The versatility of Prosopis juliflora (Sw) dc, whose leaves and pods are used as animal fodder; its fruits as a source of food in the form of flour, beverages and syrup; its wood as fuel, charcoal, railroad sleepers, stakes and poles; its effectiveness in erosion and desertification control, dune stabilization, reclamation of salinized soils and as support for apiculture, calls for a thourough study on its management in our semi-arid region.

This paper analyzes some practices performed by farmers, companies and research institutions, as regards establishment, growth and care of P. juliflora in the Brazilian Northeast.

Seedling Production Methods

P. juliflora reproduces through seeds or cuttings. The most widespread method in the Northeast is raising seedlings from seeds.

For the establishment of populations exhibiting high output of pods with good protein and sugar content, free of thorns, it is recommended to raise seedlings from cuttings obtained from parent trees having the characteristics desired, as the cross pollination of Prosopis makes it difficult to keep the traits of the parent tree.

Seed extraction

In this region, P. juliflora selection has been made by obtaining seeds from trees with the desired phenotypic characteristics such as absence of thorns, large pods, fruit shape, size and healthiness. Collection is performed directly by picking up fruit from the ground.

The common process for seed extraction among farmers is to soak the pods in water for some 12 hours, thereafter slicing the pod lengthwise with a knife along the narrow side (Gomes, 1961).

Seed extraction with chemicals is expensive and difficult to handle by peasants. Souza et al. showed the viability of obtaining P. juliflora seeds with a fodder machine, from sundried pods. This process reduces costs of manual extraction by almost 45 %, even considering the amount of seeds lost.

Minute cracks form in the tegument of the seeds so extracted, a feature which helps scarification. Thus, seeds do not require further pregermination treatment, and germination rates exceed 70%. Trials are being performed on seeds obtained with this process and, 30 months after storage in plastic bags both in cold room and at room temperature they show perfect phytosanitary conditions and germination capability.

The number of P. juliflora seeds per kilogram is about 28,400 (Carvalho, 1976), and they must be fumigated prior to storing to prevent insect attack. Morais et al. (1981) confirmed the presence of Bruchidae.


Sowing depth used for this species is 1.0 cm, enough to cover the seeds, and is performed directly in plastic bags, cans or other containers, in a 2:1 soil-manure mix. One or two seeds are placed in each container. Should both seeds germinate, the least vigorous seedling is culled, placing the remaining seedling at the center of the container.

This region's most widely used method for breaking seed dormancy is to submerge seeds in hot water for 3 to 5 minutes. This low-cost treatment provides over 90% germination and does not pose any handling hazards as compared to methods using chemicals such as sulphuric acid.

P. juliflora seedlings are raised without shading, during the period preceding the rainy season. The plants remain in the nursery for 45–60 days, until they reach 20–25 cm height.

In order to prevent the appearance of fungi or other pathogenic agents, the soil mix to be used must be previously fumigated.

At time of sowing, it is advisable to inoculate with Rhizobium. Research carried out by Franco (1982) permitted to select strains Br 4001, Br 4002, Br 4003, and Br 4007 as highly efficient for nodulation and nitrogen fixation in P. juliflora. In this region, seedlings of this species have been observed with nodules although no prior specific Rhizobium inoculation had been performed at time of sowing.

During the seedling raising stage, an average of three waterings per day is carried out, so as to keep moist the soil used as substratum and to facilitate seedling development. As seedlings grow, both amount and frequency of watering are decreased, in order to acclimatize the plants to the region's drought conditions.

Seedling quality

The root system is one of the factors with bears upon seedling quality, and which can depend on type of container. As soon as they germinate, P. juliflora seeds issue a fast-growing tap root which, depending on type of container, reaches the bottom of the container in a matter of days. Upon the impossibility of growing farther, it starts to curl upon itself.

Seedlings with wrapping roots entail subsequent plant strangling or root malformation in the field, with likely upturning of the plant.

Fallen adult P. juliflora have been observed in the Northeast. cpatsa is investigating the precise reasons behind these falls. Seedling quality is one of the hypotheses being considered, for these falls have occurred at man-made plantations. Two other factors could be soil depth, grounds here being generally flat, and water availability.

P. juliflora possesses a root system with a deep tap root which seeks the phreatic layer and anchors the plant to the ground, and fasciculate roots extending below the canopy area at a depth of some 40 cm. Most of the fallen P. juliflora did not have a well formed tap root, and the soils had an excess of water.

In more humid regions, P. juliflora tends to develop more its vegetative system, producing more wood. Planted in flat ground, with an underdeveloped root system, sizable height and heavy branches, P. juliflora tends to fall when hit by strong winds. Figures 1 and 2 show P. juliflora root system after 12 months of out planting. In both cases a main root can be observed with over one meter in length. The extension of the fasciculate roots corresponds approximately to crown diameter.

Figure 1

Figure 1. P. juliflora root at 12 months of age.
(Seedlings raised in styroblock)

Figure 2

Figure 2. P. juliflora root at 12 months of age.
(Seedlings raised in plastic containers)

Silva and Lima (1985), analyzing the quality of P. juliflora seedlings produced in different types of container, found differences in growth of plants in the nursery, without bearing on survival rates. No significant differences in height or root length were found 12 months after out-planting, as shown in Table 1. Another factor observed is that the seedlings produced in bottomless containers and in paraffin-treated paper tubes show higher biomass production.

Height, Root Length, Canopy Size and Biomass Obtained from P. juliflora Seedlings Raised in Different Containers

TreatmentNursery (2 months) Field (12 months)
Root Size
Fertile pot20.9a231.381.22.06a1,048
Plastic bag21.8a301.581.32.32a2,644
Layered strata16.7ab241.731.32.55a3,830
Styroblock8.0  b131.321.31.57  b   955
Laminated paper12.7  b261.53   1.42.30a4,350

Figures followed by the same letter do not differ from each other as per Duncan test at 5% probability level.
Source: Silva & Lima (1985).

Seedlings produced by vegetative propagation

P. juliflora cuttings, once they grow shoots and roots, are placed in containers with a soil-manure mix. The method for inducing rooting in P. juliflora cuttings was described by Souza and Nascimento (1984) and by Nascimento et al. (1985). Once the seedlings become established, they are removed from greenhouse (35° C temperature and 80% relative humidity) and placed in the nursery, where the only controlled factor is soil moisture, through irrigation.

For final outplanting, seedlings raised from cuttings undergo the same process as seedlings raised from seeds.

Planting Techniques

The planting system with P. juliflora is related to the size of the undertaking. Afforestation operations carried out by forest companies generally include land clearing, either manually or mechanically, followed by plowing and levelling, with seedlings planted at regular spacing in square stands. Plantings made by small farmers normally involve placing the seedlings in straight lines around plots, pastures, alongside roads and in small stands, where only the spot itself where the plant will be located is cleared, normally some 2 m around the pit, which is 30 cm × 30 cm.

Planting is carried out at the onset of the rainy season, so that the seedlings find freshly moist soils and can count on all the water from the coming rainfall. Pit planting involves placing the seedling some 15 cm below ground level, so that runoff water accumulates in the depression and improves seepage and water takeup.

Researchers from the Rio Grande do Norte Research Agency (emparn) have conducted promising research on planting with P. juliflora pseudo-cuttings, obtained from seedlings pruned some 10 to 15 cm above and below the collar, respectively. The main objective of this system is to reduce transportation and labor costs, without negative bearing on plantation development. This process has the further advantage of making possible the use of seedlings which have stayed too long in the nursery, as a result of postponing outplanting due to lack of rains, and normally considered useless on account of the size and extension attained by the root system.

Irrigated planting

When planting is carried out during the dry season, the plants are irrigated from a water truck provided with hoses. The truck moves slowly along the plant rows, so as to enable the operators to control the amount of water being provided to each plant. Two to three liters are supplied per plant every 10 days, as long as considered necessary.

Another alternative being examined by cpatsa is the use of permeable mud pots buried near the plants. They keep the soil moist near the roots. These pots are about 40 cm tall and have around 10-liter capacity, being replenished every 30 days, assuring plant survival and development until the onset of the rainy season. With large-scale plantings, however, this system is impractical due to the high cost involved.

In situ rainwater catchment

The dry season at the Brazilian Northeast lasts about 9 to 10 months, and the rainfall regime is irregular. Soil preparation for plantings should include the use of techniques to improve rainwater retention and infiltration.

cpatsa is presently assessing technically and economically in situ rainwater catchment systems for plantings of some evergreen species, including P. juliflora. Soil texture, structure and porosity, and depth reached by the plants' root system are essential features to consider when planning such systems.

The system consists of modifying the soil surface, so that the area between plant rows acts as a catchment area. It is necessary to form a slope to increase water runoff and to direct it towards the area containing the roots. Furrows are made following the contour and with a minimum gradient (Silva and Porto, 1982).

This system showed good results in terms of plant survival and development at a cpatsa experimental plot with P. juliflora. Furthermore, it permits the establishment of edible or fodder crops at the water catchment furrows, in the space between the plants, during the year of establishment, without affecting P. juliflora development negatively.


Spacing for P. juliflora plantations depends on the product output aimed at (timber or fodder) and on site characteristics. Alves and Campos (1985) have analyzed the various options for intercropping with P. juliflora, determining the optimum spacing for each case in terms of good area use and absence of negative interference with P. juliflora development.

With 3 × 2 m spacing aimed at fuelwood production, timber volumes produced in Petrolina were 7.2 and 15.5 m3/ha at 3 and 5 years of age, respectively (Lima, 1985). As from the third year, height growth rate starts to slow down as a result of competition, being suggested therefore to clearcut in the fifth year to obtain fuelwood. Gomes (1961) recommends 5 × 5 m as the smallest spacing for P. juliflora plantations in the Northeast.

Wider spacings, in excess of 10 × 10 m, enable greater canopy development and, consequently, higher fruit output. Trees with an average of 100 m2 vital space at the Bebedouro experimental station, Petrolina, produced pods at a mean rate of 78 kg/tree/year.


A common practice among the region's farmers is to plant Opuntia ficus var. indica in association with P. juliflora. Alves and Campos (1985) report on some alternative options for intercropping with P. juliflora tried at the Pendência farm, Paraíba.

For shading of Opuntia planted at 1 × 1-m spacing, these authors recommend a 5 × 5-m spacing for P. juliflora. In association with edible crops during the plantation establishment stage, they recommend spacings of 10 × 10 m and 2 × 1 m for corn (Zea mays) and macassar bean (Vigna unguiculata) in alternate rows. With buffel grass (Cenchrus ciliaris), they recommend leaving a 2-m-diameter area around P. juliflora free of this grass. Ribaski (1986) also recommends a clearing not smaller than 1 meter in diameter around P. juliflora, as the competition between both crops is strong, with up to 10% P. juliflora mortality when associated with buffel grass if this clearing is not provided.

When planting P. juliflora in pastures of C. ciliaris with free cattle grazing in the area, a protecting fence must be erected around the tree in addition to the clearing described above. Ribaski (1986) reported 60% mortality of P. juliflora associated with buffel grass due to damage caused by cattle. The remaining trees exhibited smaller height and diameter gain than those protected behind fences.


Considering P. juliflora rusticity, the Brazilian Forest Development Institute (idbf) has not encouraged fertilization in projects subsidized by the state. However, institutions and companies participating in afforestation campaigns involving farmers generally recommend the use of fertilizer.

Manure is the fertilizer recommended, as the semi-arid region soils are poor in organic matter. Generally speaking, 1 kg of manure is used per pit. cpatsa is carrying out trials with both chemical and organic fertilizers for P. juliflora, considering plant survival and growth rates. Results show better height growth and crown development in plants receiving 5 kg of manure per planting pit.

Cultural Treatments

As any agricultural crop, P. juliflora requires certain minimum care to become established, develop and produce well.

From nursery operations onward, weed eradication is necessary. After final out-planting, weeding three times per year is a must during the first two years to insure firm P. juliflora establishment in this region. It is not necessary to perform an in-depth clearing, and a 2-m clear space around the tree is enough to permit unimpeded P. juliflora growth, which then has no need to compete for water and nutrients. When weeding is carried out, it is advisable to leave the removed weeds lying on the ground, to check soil moisture loss through evaporation.

Pests and disease

In addition to ants, the appearance of other insects is controlled with chemicals at the nursery stage. In the field, damage caused by Oncideres spp. has been reported frequently.

Lima (1982) observed damage in 66% of the trees planted, as from the third year after field planting. Spread of this insect is controlled by burning all fallen and sawn-off branches, where the eggs have been deposited and larvae start their development.

In the nursery, Santos and Silva (1983) found that P. juliflora is a susceptible host to the nematode Meloidogyne javanica (Trub 1985) Chitwood 1946, which does not affect seedling survival rates. Nematode-bearing seedlings were planted in the field and exhibited normal development. Nematodes are controlled in the nursery by fumigating the soil.

The presence of the grasshopper Striphra robusta Mello-Leitão has been reported in P. juliflora plantations. Although it reduces foliar area in the trees, no serious damage has been observed. Recently, bee atacks to P. juliflora fruit have been detected at the final ripening stage, consuming the entire fruit pulp prior to its falling to the ground. The attack did not occur on all individuals in the population, suggesting that those attacked had very high sugar concentration in the pods. No form of control has been tested against this insect.


P. juliflora must grow unimpeded when the plantation purpose is stake and fuelwood production, without pruning. A P. juliflora plant, in regular plantations with 3 × 2-m spacing, has an average of 6 forkings below dbh (Lima, 1982). In commercial plantations whose objective is to produce fuelwood, pruning is expensive and not advisable.

At Brazilian Northeast towns, it is common to plant P. juliflora alongside streets and roads. In this case, pruning is recommended. It consists of cutting off secondary branches until the plant reaches about 1.8 m height, at which point three or four branches are left, forming the base of the tree's canopy.

Pruning for canopy shaping is also common, whereby branch tips are cut off until the desired shape is obtained. Pruning is performed at the onset of the rainy season. Drastic pruning cases are not uncommon, where the canopy or the lateral branches are reduced in size to keep them from surrounding or reaching power lines. The tree recovers normally, being fully sprouted by the time the dry season starts again.

No records exist regarding fruit productivity research on P. juliflora in which pruning has been performed either to guide the stem or to shape the canopy, in comparison with free-growing individuals. Studies conducted at cpatsa found a correlation between pod production and canopy size, plant nutritive status and flower pollination efficiency. Relative humidity and insects can have bearing on increase or decrease of P. juliflora pod output.


P. juliflora regeneration in the Northeast occurs naturally, and plants can be seen growing on river banks, alluvial soils and barren lands where animals graze freely. Animals feed on the fallen pods and then disseminate the seeds encapsulated in their droppings. After the rains, seeds start to germinate and, if conditions are favorable, they become seedlings and, later, trees.

Under adverse fertility and moisture conditions, regeneration will hardly take place as in fertile humid lands. A pod contains an average of 18 seeds and even if all of them manage to emerge intact out of the animal's digestive tract and then germinate, the resulting seedlings will be grazed by the animals themselves when faced with the scantiness of browse or pasture in these areas during the dry season.

To prevent undesired P. juliflora propagation in pastures or subsistence farming lands, it is advisable to feed the animals ground pods, either alone or combined with other fodder, so that the seeds are totally destroyed and plants will not proliferate through seeds embedded in animal droppings.

Stumps issue new shoots after clearcut felling, more intensely so in young trees. When firewood production is desired, no pruning of these regrowth is advisable, although ideally no more than two shoots should be left per stump.


alves, a. q. and campos, s. v., 1984: “A importancia prática do consórcio da algarobeira - Prosopis juliflora (Sw) dc com plantas forrageiras e culturas de subsistencia,” Silvicultura, 10 (37): 43–6.

carvalho, r. f., 1976: “Alguns dados fenológicos de 100 espécies florestais, ornamentais e frutíferas, nativas ou introduzidas na eflex de Saltinho,” pe. Bras. Flor. 7 (25): 42–4.

franco, a. a., 1982: “Fixação de N atmosférico em Prosopis juliflora (Sw) dc,” In: Proceedings, I. Simpósio Brasileiro sobre Algaroba, Natal, 1982, emparn, pp. 319–329. (emparn. Documents 7)

gomes, p., 1961: “A algarobeira,” Rio de Janeiro, Ministério da Agricultura, Serviço de Informação Agrícola, 49 p. (Series sia, 865)

lima, p. c. f., 1982: “Comportamento silvicultural da Leucaena leucocephala (lam) de Wit comparado a Prosopis juliflora (Sw) dc e Eucalyptus alba Reinw ex Blume em Petrolina-pe, região semi-árida do Brasil,” Curitiba, Universidade Federal do Paraná, 96 p. (M. Sc. Thesis)

lima, p. c. f., 1985: “Tree productivity in the semiarid zone of Brazil,” Petrolina-pe, embrapa-cpatsa, 15 p. (Paper presented at the Symposium Establishment and Productivity of Tree Plantings in Semiarid Regions, Kingsville, usa).

moraes, g. j.; ramalho, f. s.; souza, s. m. de; silva, c. m. m. de S. and lima, p. c. f., 1981: “Insetos associados a sementes de forrageiras e essencias florestais no tropico semi-árido do Brasil,” Petrolina, embrapa-cpatsa, 2 p., (embrapa-cpatsa, Research in progress, 11).

nascimento, c. e. de S.; lima, p. c. f. and silva, h. d. da, 1985: “Influencia do número de gemas no enraizamento de estacas da algaroba,” Petrolina, embrapa-cpatsa, 3 p., (embrapa-cpatsa. Research in progress, 39).

ribaski, j., 1986: “Estudo do establecimento da algaroba plantada em área cultivada com capim buffel,” Petrolina, embrapa-cpatsa, 4 p., (embrapa-cpatsa, Research in progress, 47).

ribaski, j., 1986: “Sobrevivencia e desenvolvimento da algaroba, plantada com e sem proteção, em área de capim buffel sob pastejo,” Petrolina, embrapa-cpatsa, 4 p., (embrapa-cpatsa, Research in progress, 48).

santos, j. m. dos, and silva, h. d. da, 1983: “Susceptibilidade de espécies florestais á Meloidogyne javanica na região semi-arida do Nordeste,” In: Simpósio iufro em melhoramento genético e productividade de espécies florestais de rapido crescimento, Auguas de São Pedro, Silvicultura, 8 (3): 379–9.

silva, h. d. da, and lima, p. c. f., 1985: “Tipos de maceta para la producción de plantas de algarroba,” In: Proceedings, Encuentro Regional de ciid, América Latina y el Caribe, 2, Santiago, Chile, “Forestación en zonas aridas y semi-áridas”, Santiago, ciid/infor, pp. 97–104.

silva, a. de S. and porto, e. r., 1982: “Utilização e conservação dos recursos hídricos em áreas rurais do trópico semi-árido do Brasil; tecnologias de baixo custo,” Petrolina-pe, embrapa-cpatsa, 128 p., illust., (embrapa-cpatsa, Documentos, 14).

souza, s.m. de; lima, p. c. f. and araujo, m. de s., 1983: “Sementes de algaroba: métodos e custos de beneficiamento,” R. Bras. Sem., Brasília, 5 (3): 51–61.

souza, s. m. de, and nascimento, c. e. de e., 1984: “Propagação vegetativa de algaroba através de estaquia,” Petrolina, embrapa-cpatsa, 3 p., (embrapa-cpatsa, Research in progress, 27).

Genetic Improvement of P. juliflora at the Brazilian Northeast

Ismael Eleotério Pires
Forester, M. Sc.
Assistant Professor, Forestry Department, Agricultural Sciences Center
Universidade Federal de Vicosa Minas Gerais


The widespread distribution of Prosopis juliflora (Sw) dc in the Northeast results from its advantageous position in comparison to other potential species, in terms of silvicultural performance and multiple use possibilities, such as wood or fodder production, or as support for apiculture.

A factor raising concern among specialists, however, is the fact that the seeds used for the plantations have been obtained within the region, without any genetic controls on their production, thus jeopardizing the future of the resulting populations.

This paper aims, therefore, at bringing attention to genetic base aspects of the Northeast's P. juliflora populations, in an attempt to define exploitation strategies for the existing genetic variability, with the least inbreeding effect possible.

Genetic Base of the Populations and Improvement Prospects

Assuming that the four trees surviving introduction constituted the basis for the formation of all populations existing at present, a narrow genetic base may be presumed, a fact which restricts the adoption of genetic improvement programs through successive generations. This situation was pointed out by Pires and Kageyama (1985), and later by Pires et al. (1986), who found low genetic variability in open pollinated P. juliflora progenies, at a population located in Soledade, State of Paraíba, for the characteristics of height, diameter at base and mean crown diameter, evidenced by heritability equal to or very close to zero, at ages 18 and 36 months, respectively.

Figure 1

Figure 1. Schematic P. juliflora expansion in the Brazilian Northeast.

The irregular expansion of P. juliflora in the Northeast (Fig. 1) led to the establishment of uneven-aged populations with varying degrees of kinship among and within the populations, entailing a high degree of phenotypic variability not derived from the genetic variability often expected from phenotypical assessments. This is in line with Falconer (1981), who mentioned the increase in variation within endogamic matrices in the first generations, due to the phenomenon of segregation, becoming uniform after a certain number of generations as a function of the resulting homozygosis, the speed of which is linked to the number of genes controlling each characteristic (Fig. 2).

Figure 2

Figure 2. Variance distribution in a population of related individuals (Falconer, 1981), where Vt = total genetic variance; Vb = variance between families; Vw = variance within families.

It is worth pointing out that the wide spacings used in P. juliflora plantings, namely 10 × 10 m, reduce the effect of selection by competition, as they permit the survival of individuals with different levels of inbreeding within the same population, leading to the high phenotypic variability observed. On the other hand, the adoption of tighter spacings would cause the death of the more depressed plants, as a result of the increase in the level of competition, entailing a faster rise in the mean level of inbreeding through the survival of increasingly related individuals.

Selection efforts with materials of narrow genetic base, as those from the plantations discussed here, would jeopardize the future of an improvement program, particularly with the application of selection among families, there being even a likelihood of obtaining negative results, as Venkovsky (1978) states, as a result of the genetic oscillation effect pertaining to materials of narrow genetic base.

Besides the aspects of selection gains, there is also the vulnerability factor, commonly found in materials of inadequate genetic base, i.e. materials of narrow genetic base, such as monoclonal plantations, which are prone to catastrophes caused by the attack of a pest or the occurrence of disease, pronounced water stress, etc.

It is concluded, therefore, that the establishment of a genetic improvement program for P. juliflora started from the populations existing at present in the Northeast calls for great caution, until more genetic information is acquired.

P. juliflora Propagation in the Northeast

Propagation through seeds

Considering the restricted genetic base hypothesis, the use of seeds from only one population must be avoided, in order to reduce the vulnerability of the resulting population.

Keeping in mind the argumentation of Namkoong et al. (1983), that the populations vegetating in dissimilar sites segregate differently, preference must be given to the use of seeds from different populations, not least because, in such case, it is likely that genetically differentiated materials from different generations will be used, thereby lowering the risks entailed by inbreeding and the vulnerability of the material.

The use of seeds obtained generation after generation must be avoided for regular plantations, so as to diminish the speed of endogamic depression, being it necessary to select the seed-supplying populations.

Vegetative propagation

For these P. juliflora populations, vegetative propagation constitutes an excellent tool for the multiplication of superior phenotypes, without aggravating the inbreeding levels which are inevitable in propagation through seeds in materials with restricted genetic base, until safe strategies for genetic improvement can be devised.

This type of reproduction could be adopted even at commercial scale, as P. juliflora presents satisfactory rooting indices in stakes from coppicing, in accordance with studies carried out by Souza and Nascimento (1984) and Nascimento et al. (1985). There is a need only of more accurate data on the performance of the root system of materials raised from cuttings rooted under dry conditions.

Strategies Recommended

Genetic studies

More detailed studies are necessary on hybrid identification and genetic structure of P. juliflora populations existing in the Northeast, through the establishment of provenance and progeny trials, considering the populations as provenances, at different sites, so as to count on more accurate data regarding the levels of genetic variability among and within populations, as well as the effects of genotype × environment interaction. The estimate of genetic patterns, in these trials, is essential for defining the likelihood for success of a genetic improvement program.

Seed orchard

The application of massal selection of superior phenotypes is recommended, among and within populations, aimed at establishing a clonal seed orchard for commercial seed supply.

This strategy would enable the crossbreeding of genotypes from different populations, making it possible to obtain a probable heterosis in the populations resulting from these crossbreedings. If done appropriately, this would insure assessment at clonal level, with possibilities of applying selection among clones, both for timber and for pod production purposes.

Expansion of the genetic base

The introduction of materials from the natural range of the species for crossbreeding with species from Northeast populations, taking advantage of traits possibly present in the local breeds, would be the safest alternative for the establishment of a genetic improvement program for P. juliflora in the Brazilian Northeast.

Introduction of new species

The genus Prosopis comprises 44 species (National Academy of Sciences, 1979), among which several, in addition to P. juliflora, show potential for the semi-arid conditions of the Brazilian Northeast. An introduction program for those species would be advisable, in the line of the program initiated by the Brazilian Agriculture and Livestock Research Center (embrapa), through the Agriculture and Livestock Research Center for the Semi-Arid Tropic (cpatsa), in Petrolina, State of Pernambuco.


azevedo. g., 1982: “Como e porque a algaroba foi introduzida no Nordeste,” In: Proceedings, Simpósio Brasileiro sobre Algaroba, 1, Natal, 5–7 de October 1982, Natal, emparn, Vol. 1, pp. 300–306.

falconer. d.s.: “Introdução a genética quantitativa,” Viçosa, Imprensa Universitária, 279 p.

namkoong, g.; barnes, r.d. and burley, j., 1983: “Tree breeding strategies and international cooperation,” In: Silvicultura, São Paulo, 8(32): 721–23.

nascimento, c.e. de s.; lima, p.c.f. and silva, h.d., 1985: “Influencia do número de gemas no enraizamento de estacas de algaroba,” (research in progress), embrapa, Petrolina, (39): 1–3.

national academy of sciences, 1979: “Tropical legumes: resources for the future,” Washington, u.s.a., 331 p.

pires. i.e.; andrade, g. and araujo, m. de s., 1986: “Variação genética para características de crescimento em progenies de algaroba,” typewritten report, 14 p.

souza. s.m. de, and nascimento, c.e. de s., 1984: “Propagação vegetativa de algaroba attravés de estaquia,” (research in progress), embrapa, Petrolina, (27): 1–3.

vencovsky, r., 1978: “heranca quantitativa,” In: melhoramento e produção do milho no Brasil," paterniani, e. (coord.), Piracicaba, Fundação Cargill, pp. 128–129.

Progeny Trials with P. juliflora at Seedling Stage

V. A. Rodrigues
Forester, Professor, Forestry Department
Health and Technology Center, Universidade Federal da Paraíba

J. A. Silva
Forester, Professor, Forestry Department
Health and Technology Center, Universidade Federal da Paraíba


Progeny trials at nursery level are among the most suitable techniques for improving forest species, offering advantage to both nursery and afforestation managers in their efforts aimed at obtaining genetic and silvicultural improvement in future populations.

With this technique, the selection of parent trees for seed collection can be made basing on their progenies, and a rigorous phenotypical quality control at nursery level can be immediately performed, while in the medium and long term better parent tree selection can be attained, arriving eventually at elite trees with follow-up field research.

This trial with P. juliflora (Sw) dc progenies was conducted at the Forestry Department nursery, Universidade Federal da Paraíba, and had the objectives of improving experimenting techniques with progenies, promoting student interest in forest research and seeking improvement options for the species under study.

Literature Review

According to Shimizu, Kageyama and Higa (1982), a population's genetic structure and genetic parameters can be ascertained from seedling stage. For field trials, there should be no less than ten plants per plot in the incomplete blocks layout, carrying out the number of replications permitted by each case or multiples thereof, depending on seedling availability and the area available for planting.

These trials must be established at least on three different locations, with the purpose of determining the genotype × environment interaction. Genetic structure and parameters can also be ascertained for progeny tests aimed at producing improved seeds. However, for better estimate accuracy, it is advisable to obtain genetic material for the tests through sampling and not through selection.

According to Pires and Kageyama (1985), in light of the vast areas unsuitable for agriculture or pastures, forestry appears to be a recommendable option. Consideration must be given to the introduction of versatile species able to produce timber and fodder, as well as to control erosion, support apiculture, provide fuelwood and charcoal, etc. Consequently, research studies aimed at identifying the most suitable species and provenances are essential.

Souza and Tenório (1982) state that P. juliflora is an excellent xerophyte tree species, with exceptional qualities.

The two basic characteristics on which selection was based were that parent trees had no apparent disease, and that a good number of fruits were present in their canopies. The purpose was improving these qualities with progeny tests at seedling stage.

According to Lima (1984), seeds can be sown directly in plastic, tin or wicker pots. Sowing depth must be enough to cover the seed. One or two seeds can be placed per pot; in such case, the weaker one will be culled after germination, placing the remaining seedling squarely in the center of the pot.

45 to 60 days after sowing, the seedlings are ready for final out planting. Planting must be made in pits at the onset of the rainy season, with the purpose of securing the establishment of the plants without having to provide irrigation.

According to Shimizu, Kageyama and Higa (1982), the technique of progeny trials in genetic improvement programs raises great interest among forest researchers and managers, as they permit a genotypical assessment of trees selected for production of improved seeds and for other purposes. Trials aimed solely at assessing the benefits obtained by the use of the material selected can be simple, without having to worry about starting the trials with a large number of individuals per progeny. Thus, for a location with reasonable homogeneity, a complete randomized block layout is recommended, allotting not fewer than four plants per plot and with eight replications as a minimum. Trials aimed at production of improved seeds through clearcut felling and transformation into stands raised from seedlings must be established with randomized blocks, plots having a minimum of four plants and the maximum number possible of replications.


The trial was established and carried out at the forest nursery of the Forestry Department of the Universidade Federal da Paraíba, Campus vii, Patos, Paraíba.

Fifteen parent trees were selected phenotypically on the basis of pod production observed. The first seven treatments were carried out at the Picuí region, and the 8 remaining treatments at the Patos region, Campus vii, ufpb. Ripe pods were collected and processed manually, taking care not to mix seeds from different progenitors. The seeds came from open pollination (half-siblings). The validity of this method is based on the assumption that all matrices contributed equally with their pollen and that they were equally receptive to fecundation during the same period (Shimizu, Kageyama and Higa, 1982).

The most vigorous seeds extracted from the indehiscent pods were selected for the trial. Prior to sowing, all seeds were soaked in 80° C water for 5 seconds —separated by progenitor—, according to Porter, as quoted by Popinigis (1985). The treatments to break dormancy were aimed at accelerating, uniforming and increasing seed germination rate.

The layout used was randomized blocks, with three blocks and 15 treatments each, blocks i and ii being under open environment conditions and treatment III in rustic greenhouse (75% shading).

Two seeds were sown per pot, consisting of 11 × 18-cm plastic bags, at 1-cm depth. The substratum used was sifted mulch, establishing lots with 25 pots each, with simple fencing around each plot.

Germination took place 3 to 6 days after sowing, and one week later a thinning was carried out to leave the most vigorous seedling in each pot; the pots where no germination occurred were broken up.

30 days after germination, height was measured for all seedlings. Observation, daily irrigation and cultural treatments were performed throughout the trial. Finally, 50 days after germination, data were gathered, considering the following parameters: total plant height, stem diameter, main root length, dry weight of aerial portion, root dry weight. Samples from each treatment and nine seedlings were placed for 24 hours in a kiln at 80° C, to ascertain root and aerial portion dry weight.


Progenitor No.Location

The basic considerations for selecting the above parent trees (progenitors) were: good bearing of fruit, absence of pests and disease, and good crown development.

With the data from the five characteristics outlined above, variance analyses, heritability analyses (characteristics 1,2, and 3) and Turkey tests were performed.

Results and Discussion

Assessment of the Characteristics of Prosopis juliflora (Sw) DC Progenies 50 days after Germination in Nursery Conditions. Patos, Paraíba, 1986.

Mean value of the characteristics
Treat. No.12345
0634.18 c30.99 a0.25 a4.08 a6.22 c
0735.42 c25.77 a0.26 a5.50 a3.98 bc
08    39.25 abc25.88 a0.28 a4.82 a8.92 a
0932.24 c28.49 a  0.24 ab4.55 a6.42 a
1035.47 c27.66 a  0.25 ab6.37 a  4.37 bc
1420.86 d22.89 a0.15 b4.67 a  3.60 bc
15   26.31 bd20.55 a  0.22 ab3.00 a3.49 b
5128.92 c22.66 a  0.20 ab4.44 a6.24 c
52    39.83 abc28.22 a0.26 a7.13 a4.59 c
53  41.08 ac28.00 a0.26 a7.17 a4.95 c
54    38.13 abc34.44 a0.27 a4.43 a8.94 a
5529.72 c28.33 a  0.22 ab4.34 a8.06 c
5632.59 c26.77 a  0.24 ab4.65 a7.29 a
5729.42 c23.44 a  0.21 ab5.77 a  3.97 bc
58  43.08 ac31.22 a0.29 a4.98 a9.40 a

* Figures followed by the same letter do not differ from each other as per Tukey's test, at 5% probability level.

Variance Analysis for Characteristic 1, Total Seedling Height in cm, Taken from Table 1

Cause of variationG.L.S.Q.Q.M.F
Block  2  230.08 
Treatment141549.98110.714.75 **
Residue28  652.58  23.31 
Total442432.04        CV% 14.3 

** Significant at 1% probability level.
CV%: Variation coefficient

For the characteristic of total seedling height (characteristic No. 1), significant differences were observed between the mean values of treatments at a level of 1 % probability. Treatments Nos. 8, 52, 53, 54 and 58 differ from treatment No. 14, and treatments 53 and 58 differ from treatment 15; the remaining treatments, as per Tukey test at 5% probability, do not differ from one another.

The heritability coefficient, at the level of familiy mean value for the above characteristic, was 0.7984.

Variance Analysis for Characteristic 2, Main Root Length in cm, Taken from Table 1

Cause of variationG.L.S.Q.Q.M.F
Block  2    31.65 
Treatment14  566.5040.461.36 N.S.
Residue28  831.8429.71 
Total441429.99      CV% 20.2 

N.S. Not significant.
CV% Variation coefficient.

For the main root length variable (2), no significant differences were observed among the mean values of the 15 treatments as per Turkey test at 5% probability.

Heritability coefficient at family mean value level for the above characteristic was 0.2657.

Variation Analysis for Characteristic 3, Seedling Collar Diameter in cm, Taken from Table 1

Cause of variationG.L.S.Q.Q.M.F
Block  20.00709 
Treatment14  0.52930.003783.43**
Total440.09100                CV% 13.82 

** Significant at 1% probability level
CV% variation coefficient.

For the seedling stem diameter characteristic (3), significant differences were found between the mean values for the treatments at 1 % probability; treatments 6, 7, 8, 52, 53, 54, and 58 differ from treatment 14, while, as per Turkey test, the remaining treatments do not differ among themselves at 5 % probability level.

Heritability coefficient at familiy mean level for the above characteristic was 0.7089.

Variance Analysis for Characteristic 4, seedling root system dry weight in grams, Taken from Table 1

Cause of variationG.L.S.Q.Q.M.F
Block  2  2.94 
Treatment1454.893.9211.70 N.S.
Total44122.3803    CV% 30 

N.S. Not significant.
CV% Variation coefficient.

For the seedling root system dry weight (4), no significant differences were found as per Tukey test among the 15 treatments at 5% probability.

Variance Analysis for Characteristic 5, seedling aerial portion dry weight in grams, Taken from Table 1

Cause of variationG.L.S.Q.Q.M.F
Block  2    1.45 
Treatment14182.7813.065.91 **
Residue28  61.79  2.21 
Total44246.02           CV% 24.64 

** Significant at 1% probability level.
CV% variation coefficient.

For seedling aerial portion dry weight (5), significant differences were observed among the means of the treatments at 5% probability; treatments 8, 54, and 58 differed from treatments 7, 10, 14, 15 and 57; treatment 55 differs from treatment 15 the same way as treatment 58 differs from treatment 52. The remaining treatments, as per Tukey test, do not differ from one another at 5% probability level.

A general analysis of characteristics 1, 3 and 5, where significant differences were observed, showed that treatments 8, 52, 53, 54 and 58 are similar, while the superiority with regard to treatments 14 and 15 was evident.

It was also observed that treatments 2 and 4, i.e. main root length and main root weight, did not show significant differences between their mean values.

Analysis of Progeny Performance

Prosopis juliflora (Sw) DC 50 days after Germination under Nursery Conditions; Blocks I and II in the open; Block III in Rustic Greenhouse (75% shading)

Mean values of the 15 treatments per block
CharacteristicsBlock I and IIBlock III
1. Total seedling height32.26 cm36.77 cm
2. Main root length26.58 cm27.91 cm
3. Collar diameter  0.24 cm  0.24 cm
4. Root system dry weight4.87 g5.38 g
5. Shoot dry weight6.03 g6.02 g

A small difference in performance may be observed for characteristics 1, 2, and 4 in relation to the environmental conditions under which their trial was conducted, considering that all blocks received the same treatment. Therefore, a certain advantage can be attributed to the more amenable rustic greenhouse conditions.

It may also be observed that the mean value for collar diameter coincides exactly in both environments. The interesting fact is that seedling height in greenhouse is greater, while its respective weight is smaller, when compared to outdoor environment. This is attributed to the seedlings growing less and having greater foliage outdoors, while in greenhouse the opposite occurred: longer growth but fewer leaves.


The progeny trial at seedling stage does not provide conclusive data, due to the rather short period during which data were gathered for the characteristics of interest (50 days), mainly as regards genetic improvement forecasts from a genotypical selection of parent trees.

Some suggestions can be put forth: the number of parent trees must be increased considerably, so as not to narrow the genetic base; research must be controlled at the field stage in order to perform a reliable analysis concerning genetic heritability of the progenies and, later, make a genotypical selection of progenitor trees for the purposes sought; a more detailed recording of parent tree characteristics must be made in order to improve one or more characteristics.

At nursery level, the relevance of progeny phenotypical selection on good results must be stressed; and at field level, the importance of the comparative base must be stressed regarding genetic parameters.

Good selection of parent trees can improve significantly pod production in afforestation efforts. Here, P. juliflora pods strewn on the ground in a Brazilian plantation.


lima, p.c.f., 1984: “Algaroba, uma das alternativas para o Nordeste,” Brasil Florestal No. 58, ibdf, 65 p.

pires, i.e. and kageyama, p.y., 1985: “Caracterização da base genética de uma população de algaroba - Prosopis juliflora (Sw) dc - existente na região de Soledade-pb,ibdf, Piracicaba, esalq/usp, 65 p.

popinigis, f., 1985: “Fisiologia das sementes,” Brasília, 289 p.

shimizu, j.y.; kageyama, p.y. and higa, a.r., 1982: “Procedimentos e recomendações para estudos de progenies de essencias florestais,” Curitiba, embrapa/urpfcs.

souza, r.s. and tenorio, z., 1982: “Potencialidades da algaroba no Nordeste,” Simpósio Brasileiro sobre Algaroba, Natal, 470 p.

Pretreatment of Acacia and Prosopis Seed by Two Mechanical Methods

Finn Stubsgaard
Danida Forest Seed Centre
Krogerupvej 3 A
DK-3059 Humlebaek, Denmark


Two simple mechanical pretreatment methods for Acacia and Prosopis seeds are presented. Large amounts of seeds can be pretreated in a short time, either on a small scale (as in a laboratory) or on a large scale (as in large planting programs). Viable seeds germinate well after these pretreatments.

fao has funded this research and supplied the seeds.

Morphology and Structure of Acacia and Prosopis Seeds

Most Acacia and Prosopis seeds are flat. On each of the two flat sides there is often a light oval line, the pleurogram. The area within the pleurogram is slightly raised and is called the areole. The hilum is seen as a small bump.

Figure 1

Figure 1. Outer seed structure seen from above.

Figure 2

Figure 2. Longitudinal section through a seed.

Problems with Mechanical Scarification

Scarification Methods

Impaction Scarifier

The scarifier works by slinging the seeds against the wall of a concrete cylinder. The instrument, the “Seed Gun,” is home-made and is shown in Figure 3.

Figure 3
Cover and bottom of cylinder is made of plywood. The concrete cylinder is a section of drainage tube.

Figure 3. A sectional view of Seed Gun Mk 1.

The present model comprises a rotating pipe (length 26 cm and diameter 4 cm) inside the concrete cylinder (30 cm in diameter). The pipe is driven by an electric drill with electronic speed regulation (e.g. a Bosch 1159.7 55W). A manually driven model should have 1:20 gearing between the handle and the pipe. The drill should operate at 800–1000 revolutions per minute. This speed corresponds to a speed of 11–15 m/s, with which the seed should hit the concrete wall of the cylinder.

The calculations are as follows:

The dimensions of the instrument are not critical. When the inside diameter of the concrete cylinder is α meter and the rotating pipe is 4 cm shorter than the diameter of the cylinder, engine revolutions are calculated as follows:

To find the optimal number of revolutions for a seed batch, some seeds are treated while the speed is increased until white spots appear on a few (approx. 5%) of the seeds, which means that they have either lost part of their seedcoat or have been broken. At this speed the majority of the seeds will have acquired microscopic cracks in the outer layers of the seedcoat. It is essential that the seeds encounter only one impact big enough to form cracks. A second impact would damage the seed as the protective structure of the seedcoat is already damaged by cracks from the first impact. A batch of 1 kg can run through the “Seed Gun” in 10 seconds.

Preliminary results for germination after impaction

Germination tests were carried out to assess the effect of scarification in the “Seed Gun.” As controls were used seeds scarified by holes made in the seedcoat with a dentist drill (this method gives, with few exceptions, germination equal to seed viability measured by the tetrazolium test).

The seed gun proved generally best for Prosopis seed, with little variation in effect for revolutions between 800 and 1,100 r.p.m.

For six Prosopis species tested, the effectiveness of the pretreatment —measured as the ratio germination % after seed-gun treatment/germination % after treatment with dentist drill— was as follows:

Prosopis chilensis1.0
P. cineraria0.7
P. flexuosa1.1
P. glandulosa torreyana1.0
P. tamarugo0.7
Prosopis sp.0.9

For Acacia seed the effectiveness was generally not so high and varied with speed. In general, 1100 r.p.m. gave better results than 800 r.p.m., probably because some species have a thick seedcoat resembling a crash helmet, with inner layers absorbing the impact. The results are the following:

 800 r.p.m1,100 r.p.m.
Acacia caven0.6
Acacia farnesiana0.4
Acacia nilotica indica0.30.8
Acacia nilotica tormentosa0.20.9
Acacia senegal1.00.8
Acacia tortilis0.6

Mechanically preatreated seeds should be sown within two weeks after the treatment until further investigations have clarified for how long, and under which conditions, pretreated seeds can be left unsown.

There were no signs of increasing proportion of abnormal seedlings after impaction.

In conclusion, it may be stated that for the above 6 Prosopis and 5 Acacia species, reasonable results can be expected with large-scale application of impact scarification (at 15 m/sec).

Burning or heat scarification

The problem of breaking the seedcoat of some Acacia species led to further studies of burning as a pretreatment method.

Partial burning was tried by Lunden and Kinch (1957, Agronomy Journal 49, p. 151–153) on Alpha Clover. The seeds slided or jumped down a hot plate. The fraction of impermeable seed was reduced by 70%.

For a more intensive use, it was thought that a commercially available “glow-brander” (see Fig. 4) could be used.

When the point of the glow-brander (or point of any red-glowing iron) touches the seedcoat for ½ second, a small pfft sound is heard and a small brown hole and some fine cracks are made in the outer impenetrable layers of the seedcoat.

Most Acacia and Prosopis seeds are flat and will lie on one areole showing the other one on top. Burning is done within the areole. This will prevent any damage to the radicle and save time finding the end opposite the radicle.

Small burnt marks may appear on the cotyledons, but further growth is apparently not affected.

In order to prevent damaging the seed unnecessarily with heat penetrating the seedcoat, the glow-brander should (1) work at the highest possible temperature, (2) have as small a point as possible and (3) touch the seedcoat for as short a time as possible. 100 seeds can be treated in a couple of minutes.

Figure 4

Figure 4. The “glow-brander”.

Acacia Preliminary results for germination after burning

For Acacia the burning method gives the same germination results as holes made manually in the seedcoat (drilling, chipping, filing, etc.), i.e. a germination percentage very near the highest possible.

The conclusion is that if seed is in scarce supply in research work and in the laboratory, “point-burning” is a simple and effective method of pretreatment. The “glow-brander” is now used for standard pretreatment for all hard-coated seeds at the danida Forest Seed Centre.

Previous Page Top of Page Next Page