Contents - Previous - Next

This is the old United Nations University website. Visit the new site at

3 Response and stress tolerance of caatinga trees to various water conditions

A better knowledge of response rates, which are expressed by a and ß in equations (2) and (3), and stress tolerance of trees to water conditions in the tropical semi-arid region is important to an understanding of the adaptation of the caatinga trees to their natural environment. For this purpose, the relationship between annual growth rate of trees and water storage and water deficit were studied. Two hundred and fifty samples from sixteen tree species were collected at five sites, Pocinhos, Patos, Mossoro, Quixada, and Petrolina (see fig. 13.1).

Annual tree growth is expressed by the relative ring width, which is called relative growth ratio in order to eliminate the influence of the microenvironment, such as topography and, soils, and in order to allow comparison of samples with each other, is defined to be:

Gr= D r/R (1)

where D r = annual tree ring width, R = radius of tree sample in direction of measurement.

Tsuchiya (1990) expresses the relationship between tree age and tree radius as a linear equation throughout the entire growth period for the two samples of the dominant tree species: jurema-preta and catingueira, 20 and 13 years old, respectively. The Thornthwaite and Mather method is employed to calculate the water balance, assuming the water-holding capacity of the soil as 50 mm/m, (Tsuchiya, 1990).

Water storage and water deficit must be considered for the period of tree growth rather than the calendar year or the water year. Tree growth is found once per year in the rainy season, usually from October to April or May. Thus, the tree growth year is defined by the period from the appearance of water storage to just before the next appearance. However, if water storage is absent throughout a given year, the next year begins only when the water deficit records the minimum value in the year.

Figure 13.2-1 Relationships between Gr and Ws and Wd (faveleira).

Figures 13.2-1 and 13.2-2 exemplify the relationships between the relative growth ratio (Gr) and annual water storage (Ws), and that between relative growth ratio and annual water deficit (Wd) for faveleira and pau-branco, respectively. These relationships are expressed by the following equations:

Figure 13.2-2 Relationships between Gr and Ws and Wd (pau-branco).

Gr = a Ws + ß (2)
Gr = p Wd + q (3)

>a, ß, p, q: parameters <

In these equations, the parameters a and ß represent the response rate of trees to the annual water storage and annual water deficit, respectively.

Assuming the magnitude of stress tolerance equals water deficit when the relative growth ratio is zero, the stress tolerance to water deficit (Tws) is expressed by the following equation:

Table 13.4 Values of a, p, q, and Tws*

Local names of trees a p q Tws*
1 Burra-leiteira 0.494** - 0.082 124 1,220
2 Faveleira 0.389** - 0.056* 140 1,790
3 Aroeira 0.336** - 0.052* 103 1,920
4 Jurema-preta 0.321** - 0.053** 141 1,890
5 Angico 0.288** - 0.049* 98 2,040
6 Cablo-de-negro 0.332** - 0.086** 203 1,160
7 Jurema-branca 0.258** - 0.084** 162 1,190
8 Pinhão 0.256** - 0.058** 146 1,720
9 Catingueira 0.230** - 0.056** 126 1,790
10 Pau-branco 0.119** - 0.049** 85 2,040
11 Imburana 0.104** - 0.044** 105 2,270
12 Sabiã 0.112* - 0.041* 123 2,440
13 Jucã 0.068** - 0.027** 70 3,700
14 Pau-de-casca 0.250** - 0.027 73 3,700
15 Espinheiro 0.133* - 0.022 90 4,550
16 Pereiro 0.105* - 0.021* 65 4.760

Note: **significant at the 1% level; *significant at the 5% level.

Tws = Wd = |-q/p| (4)

However, the stress tolerance must be standardized as the relative stress tolerance by the common value of q which corresponds to the magnitude of G(r) when the water deficit is zero. Therefore, if 0.1 mm is adopted for the common value of q, the relative stress tolerance Tws is expressed by the following equation:

Tws = |-q0.1/p| (5)

Values of Tws for sixteen species are also shown in table 13.4. Next, the relationship between two independent parameters and Tws is found (fig. 13.3).

Sixteen tree species are classified into three groups. Group A includes tree species which have low stress tolerance although the response rate is large; group B varies little from group A in forms of the stress tolerance, but the response rate is larger; and group C includes trees whose response is similar to group B. but whose tolerance is the highest of all.

When comparing these results with the floristic composition of caatinga stands at the five sites, it can be estimated that the dominant species transit successively from group A to group B. and from group B to group C. Trees of group A are dominant species in the primary stage of plant succession, which includes pioneer species. Trees of group C are dominant in the mature stage and trees of group B belong to the transition stage between group A and group C.

Figure 13.3 Relationship between a and Tws*.

4 Utilization of the caatinga trees

Trees of the caatinga are utilized for a variety of purposes: browsing by livestock, production of firewood and charcoal, fencing, and medicine. There is abundant evidence that thoughtless removal of trees results in land degradation. Therefore, if the caatinga is to continue as a viable ecosystem, practices based on a knowledge of the ecological characteristics of trees are essential. Response rate and stress tolerance of trees as a function of soil water conditions are important factors in assuring the continuity of the caatinga. Partition rates of chemical elements among trees, soil, and bedrock are also important to an understanding of the transfer of nutrients from soil to the trees (Masuda et al., 1989). Moreover, greater knowledge of the utilization of the caatinga trees aids in proper forest management.

Dietary selection by grazing animals such as cattle, goats, and sheep, as well as the human utilization of caatinga trees, were surveyed near Petrolina, Bahia, in 1990 and a survey of firewood utilization for bakeries was conducted for Campina Grande and Patos, both in Paraíba, in 1986 (Nishizawa and Pinto, 1988).

Both questionnaires and interview surveys with farmers, ranchers, and bakery operators were employed. Table 13.5 shows dietary selection by grazing animals, as well as human utilization, of 36 species of caatinga trees near Petrolina. Seven species - arapiraca, calumbi, juazeiro, Jucá, juremabranca, jurema-preta, and mororo - are most preferred by cattle, goats, and sheep. In comparison, seven species burra-leiteira, Cabelo-de-negro, Joãomole, pereiro, quipa, and rompe-jibão - are normally rejected by the same livestock.

Table 13.6 shows the results of three surveys on dietary selection by livestock. Hardesty et al. (1988) surveyed five species at two sites located in Sobral and Quixada, and Saito and Maruyama (1988) conducted a survey of eleven species at São Joao do Cariri.

Catingueira is classified as extremely palatable (EP) and somewhat preferred (+ +) by the three types of animals as stated both by Hardesty et al. (1988) and our survey in 1990, mentioned above. However, Saito and Maruyama identified catingueira as a non-edible tree for goats and sheep. In spite of this finding we observed goats eating very young foliage of catingueira at Fazenda Varginha, and cattle at Fazenda Nogueira. Catingueira is one of the first trees to produce new leaf growth with the arrival of seasonal rains, and the foliage is eagerly sought by livestock. As the rains begin, within a few days the foliage releases a strong pungent smell which livestock find offensive, and consequently they ignore the leaves until they dry and fall at the start of the dry season. This characteristic ensures that the catingueira foliage is available for animal consumption during the dry season (Hardesty et al., 1988). Pfeister and Malechek (1986) state that catingueira is relatively unpalatable as green foliage and that the dry, nutritious tree leaves provide the bulk of the late dry season foliage. Thus, animal preference for catingueira depends on the season. Jurema-preta, one of the most common trees in the early succession of the caatinga stand, is identified by Hardesty et al. and Nishizawa et al., respectively, as extremely palatable (EP) and somewhat preferred by livestock. However, Saito and Maruyama (1988) identified jurema-preta as a non-edible tree for cattle. Marmeleiro is a principal species preferred by livestock, according to the survey by Nishizawa et al., but Pfeister et al. (1988) state that livestock mainly browse on it during relatively less severe dry seasons.

Table 13.5 Utilization of caatinga trees

    Human utilization
Dietary selection by livestock Firewood  
Local name of tree Cattle Goats Sheep Lumber Charcoal Bakeries Brick-yards Fencing
Angico Man- + + + + + + + + + -
Arapiraca da- +++ +++ +++ ++ +++ ++ -
Aroeira caru + + + + + + + + + -
Bolusu Malva + + + + + + + + + -
Brauna Mar- + + + + + + + + + -
Burra-leiteira me- -   -   -   -
Cabelo-de-negro leir- -   -   -   -
Calumbi o-br- + + +   + + +   + + +   -
Carqueijo anco + +   + +   + +   -
Canafistula Maria- + +   + +   + +   -
Catingueira aura + +   + +   + +   -
Espinheiro Mo- + + +   + +   + +   -
Faveleira fum- + +   + +   + +   + +
Imburana (Umburana) bo + +   + +   + +   -
João-mole   _   _   _   +
Juazeiro   + + +   + + +   + + +   -
Jucá   +++   +++   +++   -
Jurema-branca   + + +   + + +   + + +   -
Jurema-preta   + + +   + + +   + + +   -
Mororó +++ +++ +++ - - - - ++
Pau-branco + + + - + + + + + +
Pau-de-casca + + + - + + + + + +
Peretro - - - - + + + + ++
Pinhão - - - - - - - -
Quebra-facão + + + + + + + + - + + + + +
Quipá - - - - - - - -
Rompe-jibão - - - - - - + -
São-joão + + - - + + + + + +
Sobiá ++ +++ +++ - ? + + ?
Uricuri + + + + + + + - - - - -
Xiouexique + + + - - - - -

Key: Dietary selection: + + +: highly preferred;; + +: somewhat preferred;+: slightly preferred; -: never preferred. Utilization: + +: most utilized; +: utilized; -: not utilized.

Table 13.6 Comparison of three surveys on dietary selection by livestock

Local Name Name of researcher Dietary selection
Cattle Goats Sheep
Angico Hardesty et al. (H)      
Nishizawa et al. (N) + + + + + +
Saito et al. (S) NE NE NE
Brauna Hardesty et al. (H)      
Nishizawa et al. (N) + + +
Saito et al. (S) E E NE
Catingueira Hardesty et al. (H) EP EP EP
Nishizawa et al. (N) + + + + + +
Saito et al. (S) E NE NE
Jurema-branca Hardesty et al. (H)      
Nishizawa et al. (N) + + + + + + + +
Saito et al. (S) NE E E
Jurema-preta Hardesty et al. (H) EP EP EP
Nishizawa et al. (N) + + + + + + + +
Saito et al. (S) NE E E
Malva Hardesty et al. (H)      
Nishizawa et al. (N) + + + + + + +
Saito et al. (S) NE E E
Mandacaru Hardesty et al. (H)      
Nishizawa et al. (N) + + +
Saito et al. (S) E NE NE
Marmeleiro Hardesty et al. (H)      
Nishizawa et al. (N) + + + + + +
Saito et al. (S) E E E
Pau-branco Hardesty et al. (H) P NP NP
Nishizawa et al. (N) + + +
Saito et al. (S)      
Pinhão Hardesty et al. (H)      
Nishizawa et al. (N) - - -
Saito et al. (S) NE NE NE
Sobiã Hardesty et al. (H)   EP EP
Nishizawa et al. (N) + + + + + + +
Saito et al. (S) E E E
Xiquexique Hardesty et al. (H)      
Nishizawa et al. (N) + + +
Saito et al. (S) E E E


Key: (Hardesty et al.): (Nishizawa et al.): (Saito et al.):
EP: extremely palatable + + +: highly preferred E: edible
P: palatable + +: somewhat preferred NE: non-edible
NP: non-palatable +: slightly preferred  
  -: never preferred

In the semi-arid interior of North-East Brazil, both temporal and spatial variations in rainfall cause pronounced fluctuations in the quality, quantity, and distribution of available foliage as well as changes in the floristic composition with successive plant stages. Therefore, the dietary selection by livestock is different from place to place and from season to season.

Near Petrolina, seventeen species are identified as being used for charcoal production, twenty-one for firewood at bakeries, twenty for firewood at brick and tile manufacturers, and seventeen for fencing. Fifteen species of caatinga trees are commonly used for all of the these purposes (table 13.5).

Almost all the bakeries in Campina Grande and Patos, except three that also employ supplementary electric ovens, use firewood to bake bread. Total consumption of firewood only for bakeries is approximately 600 tons per month in Campina Grande (population 300,000) and 50 tons per month in Patos (population 70,000).

Trees for firewood use by bakeries mainly consist of two species, jurema-preta and catingueira. There is a little difference in the composition for firewood of these two species used by bakeries in Campina Grande and Patos (table 13.7). In Campina Grande, catingeira is prevalent, followed by jurema-preta. On the other hand, in Patos, jurema-preta predominates, followed by catingueiro.

Almost all the bakeries in Campina Grande purchase firewood through brokers from the Cariri, where human population is sparse and thus human impact on the caatinga has been comparatively small. However, the bakeries in Patos obtain their firewood directly from the environs of the town, and thus human impact is relatively greater.

Table 13.7 Tree species for firewood used by bakeries in C' pine Grande and Patos

  Campina Grande Patos
Tree species Frequency % Frequency %
Jurema-preta 51 34 12 33
Catingueira 69 46 10 28
Pereiro 19 13 3 8
Aroeira 4 3 4 11
Juazeiro 3 2 3 8
Other species 3 2 4 11

While catingueira and jurema-preta are also used by the bakeries in Petrolina, other species, such as angico and calumbi, are generally used. However, there are some differences between species preferred by bakeries in the region around Petrolina, on the São Francisco River in Pernambuco, and the interior of Paraíba State.

Dominant trees for firewood used by bakeries also vary from place to place, as in the case of dietary selection by livestock. Although fencing materials of wood are now being replaced by concrete, several types of traditional fences are still constructed from caatinga trees (Barros, 1985; Ribaski, 1986).

Although many of the trees from the semi-arid interior of North-East Brazil are used as stated above, not all are useful species, nor are they always available in any given region, because floristic composition of the caatinga stand varies with stages of plant succession and the magnitude of human impact.

5 Deforestation associated with increased firewood consumption and charcoal production

Utilization of caatinga trees is a matter of increasing concern, mainly because of the increasing population and changing lifestyles. For example, recently new types of bread are being introduced in North-East Brazil, which indicates a change in the traditional dietary patterns. This change has caused an increase in the number of bakeries and deforestation of the caatinga for firewood in baking.

Along with changes in dietary patterns, the construction materials used in housing have also been gradually upgraded. Even in the rural areas, houses with mud walls and thatched roofs are changing to brick walls and tiled roofs. Therefore, consumption of firewood at brick yards and tile factories has also increased.

More accurate estimates of deforestation due to firewood and charcoal consumption are necessary in order to understand better the associated problems, as well as to suggest appropriate methods of sustainable development and sound management of the caatinga.

The rate of deforestation per month can be expressed as follows:

Adf = Mdf/rtrsPf (6)

where Mdf = consumption of firewood (kg/month); Pf = mass of trunks and branches (kg/100 Km²); rt = utilisation rate of trunks and branches; rs = rate of mass of useful species for firewood to total mass of the stand.

Further, Pf is expressed as follows:

Pf = N(D)Wt(D) (7)

where N(D) is the tree density in a unit of the stand, which is usually expressed by the number of trees with breast height having diameter D in 100 m², and Wt(D) is the weight of the trunks and branches of trees with breast height diameter D.

Hayashi (1981, 1988) has proposed the following empirical equations of N(D) and Wt(D):

N(D) = 217 exp.( - 0.42D) (8)

Wt(D) = 0.206D2.233 (9)

Accordingly, total mass of trunks and branches with the magnitude of D from D1 to D2 per 100m² is calculated with the following equation:


Although the diameter of a firewood log is seldom larger than 10 cm, it ranges roughly from 5 cm to 8 cm. Assuming that D1 and D2 are 5 cm and 8 cm, respectively, the total mass is calculated at approximately 690 kg per 100m² by equation (10).

The magnitudes of rt and rs in equation (6) vary with the floristic composition, which differs by the stage of plant succession and the magnitude of human impact.

Consequently, assuming further that two pairs of rt and rs are as follows:

rt = 0.6, rs = 0.6

in the first pair of parameters,

rt = 0.4. rs = 0.4

in the second pair of parameters.

Adf in equation (6) is expressed as follows: in the first case

Adf = (1/248) Mdf (11)

and in the second case

Adf = (1/110) Mdf (12)

Inserting the firewood consumption of 6.0 x 105 kg per month by bakeries in Campina Grande into equations (11) and (12) yields the values of 2.4 x 105 m² per month and 5.5 x 105 m² per month respectively, as Adf.

The deforestation area per year is approximately 2.9 Km² and 6.6 Km² in the first and second pair of parameters, respectively.

If the consumption of bread per person for all of North-East Brazil, with a population of about forty million, is assumed to be at the same rate as in Campo Grande, and the firewood for bakeries is also assumed to be from the caatinga stand, then the total deforestation area for firewood from bakeries alone would be 312 Km² for the first parameter and 715 Km² for the second.

The deforestation of caatinga trees caused by charcoal production is also a great concern of ours. Deforestation area can be estimated for charcoal production by employing equation (6), employed above with firewood for bakeries.

Generally, one unit mass of charcoal is produced from about two unit mass of freshly cut green caatinga wood (Hayashi, 1988). Equation (6) for charcoal production is then converted to the following:

Adf = Mdf/rtrsPf
= 2Mdc/rtrsPf (13)

where Mdc is charcoal production (kg/year).

Charcoal production in North-East Brazil is roughly 500,000 tons per year (IBGE, 1991). Therefore, about 1,000,000 tons of green wood is consumed in charcoal production.

Employing equations (11) and (12), the deforestation area per year is approximately 4,000 Km² and 9,000 Km² in the first and second pair of parameters, respectively.

Although no scientific data is available for firewood use at brick and tile factories, rough estimates from factory managers suggest that the total area deforested for brick and tile production is about 1,000 to 1,500 Km² per year.

Combining all the above uses, the total deforestation area of caatinga trees associated with firewood and charcoal production in the semi-arid interior of North-East Brazil is estimated in the range of 5,300 Km² to 11,200 Km² per year.

6 Conclusions and a proposal

Response rate and stress tolerance, which indicate the adaptabilities of caatinga trees to the semi-arid interior, are ascertained by measuring tree growth of 16 plant species. Moreover, the characteristics of their adaptabilities are classified into three tree groups, A, B. and C, associated with plant succession after clear-cutting.

Trees for animal grazing and human utilization are found not only in the mature stage (Group C) of plant succession but also in the primary stage (Group A) and the transitional stage (Group B). Trees belonging to Group A, such as aroeira and jurema-preta, require about fifteen years to reach maturity. These two, and other tree species, reach sufficient size for firewood and charcoal production in less than ten years. Trees in Group B normally have sufficient growth for firewood and charcoal production within fifteen years.

The area covered by the potential caatinga stand is about 930,000 Km². It is assumed today that only about one-fourth of the potential caatinga stand still remains. Therefore, an area of about 230,000 Km² is still available for firewood and charcoal production.

As stated above, the most useful caatinga trees provide sufficient firewood and charcoal production for a continuous recycling period of fifteen years. Thus, sustained use of the caatinga stand allows for regrowth of trees, if careful forest rehabilitation programmes are implemented. Moreover, it would be theoretically possible to supply about 400,000 more tons of charcoal from the caatinga area than is presently consumed.


Contents - Previous - Next