In: Ecophysiology of Tropical Tree Crops ISBN 978-1-60876-392-4

Editor: Fabio DaMetta., pp. © 2009 Nova Science Publishers, Inc.

Chapter 3

CACAO: ECOPHYSIOLOGY OF GROWTH AND

PRODUCTION
Alex-Alan F. de Almeida and Raúl R. Valle

ABSTRACT
Cacao (Theobroma cacao L.), one of the world’s most important perennial crops, is
almost exclusively explored for chocolate manufacturing. Most cacao varieties belong to
three groups: Criollo, Forastero and Trinitario that vary according to morphology, genetic
and geographical origins. Cacao is cropped under the shade of forest trees or as an
unshaded monocrop. Seedlings initially show an orthotropic growth with leaf emission
relatively independent of climate. The mature phase begins with the emission of
plagiotropic branches that form the tree crown. At this stage environmental factors exert a
large influence on plant development. Growth and development of cacao are highly
dependent on temperature, which mainly affects vegetative growth, flowering and fruit
development. Soil flooding decreases leaf area, stomatal conductance and photosynthetic
rates in addition to inducing formation of lenticels and adventitious roots. For most
genotypes drought resistance is associated with osmotic adjustment. Cacao produces
caulescent flowers, which begin dehiscing in late afternoon and are completely open at
the beginning of the following morning releasing pollen to a receptive stigma. Non-
pollinated flowers abscise 24-36 h after anthesis. The percentage of flowers setting pods
ranges from 0.5 to 5%. The most important parameters determinants of yield are related
to: (i) light interception, photosynthesis and capacity of photoassimilate distribution, (ii)
maintenance respiration and (iii) pod morphology and seed formation. These events can
be modified by abiotic factors. Cacao is a shade tolerant species, in which appropriate
shading leads to relatively high photosynthetic rates, growth and seed yield. However,
heavy shade reduces seed yield and increases incidence of diseases. In fact, cacao yields
and light interception are highly correlated when nutrient availability is not limiting. High
production of unshaded cacao requires large inputs for protection and nutrition of the
crop. Annual radiation and rainfall during the dry season explains 70% of the variation in
annual seed yields.
2 Alex-Alan F. de Almeida and Raúl R. Valle

INTRODUCTION
Cacao is a preferentially alogamous tropical woody species formerly in the Sterculiaceae
family (Cuatrecasas, 1964) and reclassified in the Malvaceae family (Alverson et al., 1999).
Under natural conditions the tree can reach 20 to 25 m in height (Lachenaud et al., 1997),
whereas under cultivation height varies from 3 to 5 m. The geographical origin of cacao is
South America (Motamayor et al., 2002), where several wild populations can be found in the
Amazon and Guyanian regions. It is considered one of the most important perennial crops on
the planet, with an estimated world output of 3.46 million tons in 2008/9 (ICCO, 2009).
Cacao is predominantly grown in the tropical areas of Central and South America, Asia and
Africa (Marita et al., 2001) and is commercially exploited for seed output that is mainly
destined for chocolate manufacturing. However, derivatives and byproducts of cacao are also
transformed into cosmetics, fine beverages, jellies, ice creams and juices.
Traditionally cacao varieties in cropping systems belong to three main groups named
Criollo, Forastero and Trinitario, which differ according to morphology and genetic
characteristics and geographical origins (Bartley, 2005). Cacao Criollo was the first
domesticated cacao and has been cultivated for at least two millennia in Central and South
America (Sounigo et al., 2003). This group comprises varieties that produce fruits (pods)
containing white or pinkish seeds that yield seeds highly prized for their flavor and mainly
used in fine chocolates manufacture (Marita et al., 2001). Cacao Criollo is, however,
infrequently cultivated because of its high susceptibility to diseases (Soria, 1970). Varieties of
the Forastero group are widely cropped due to their high yields and resistance to diseases.
They produce around 80% of the world output of cacao seeds (Marita et al., 2001). This
group is subdivided into Lower and Upper Amazonian Forasteros; the former are cultivated in
the Amazon Basin and their cultivars were the first to be introduced in Africa, while the latter
are considered more genetically diversified and frequently used in breeding programs due to
their vigor, precocity and disease resistance (Iwaro et al., 2001). The first Forastero varieties
originated from the lower Amazon Basin and were cultivated mainly in Brazil and Venezuela
(Sounigo et al., 2003). The Trinitario group is considered a recent hybrid that originated from
crosses between Criollos and Lower Amazonian Forastero genotypes or intermediate types
(Motamayor, 2001).
Traditionally, cacao is cultivated under the shade of a selectively thinned forest (Lobão et
al., 2007) and represents one of the first agroforestry systems in tropical America. Cacao
agroforestry has been known since pre-Colombians times by the Mayas (Bergman, 1969). In
the Atlantic coastal forests of the states of Bahia and Espirito Santo, Brazil, around 4% of the
world and 75% of the Brazilian cacao output is obtained using a system locally called
Cabruca (Lobão et al., 2007). This system is a special kind of agroforestry in which the
understorey is drastically suppressed to introduce cacao, and the density of upperstorey trees
is reduced (Lobão et al., 2007). Cacao cultivation in this form shows much of the
sustainability attributes of the natural heterogeneous forest and is considered an efficient
agroforestry system (Figure 1) for protection of tropical soils against degradation agents
(Alvim, 1989b).
Cacao is also intercropped around the world in planned systems with other species of
economic value like Areca catechu, Cocos nucifera (Liyanage, 1985; Alvim and Nair, 1986;
Daswir and Dja’far, 1988; Abbas and Dja'far, 1989), Hevea brasiliensis, Syzigium
 Cacao: Ecophysiology of Growth and Production 3

aromaticum, Cinnamomum zeylanicum, Erythrina fusca (Alvim, 1989a; 1989b), Bactris

gasipaes (Almeida et al., 2002a) and other Amazonian species (Brito et al., 2002). Several
agroforestry systems use more than three species in planned associations (Müller and Gama
Rodrigues, 2007).

Figure 1. Some systems of cacao cropping: Cabruca, rubber x cacao and cacao without shade tree
(from left to right). Photos by the authors.

In contrast, cacao is also cropped under unshaded conditions (Figure 1). In Ghana and
Ivory Coast, for example, 50% of the total cacao farm area is under mild shade whilst an
average of 10% in Ghana and 35% in Ivory Coast is managed under no shade (Padi and
Owusu, 1998). In a shade and fertilizer trial conducted with Amazon cacao over a 20-year
period in Ghana, yield of heavily shaded plots was about half that under the non-shade
treatment (Ahenkorah et al., 1987). Despite this, the authors inferred that the economic life of
an unshaded Amelonado cacao farm in Ghana might not last for more than 15 years of
intensive cropping. Nonetheless, cacao can be produced economically with no shade if
adequate management practices including water and nutrient replenishment are utilized.
In this chapter, we provide information about the effects of biotic and abiotic factors on
growth and development of the cacao plant that could be useful in the design of strategies for
new cropping systems. This is particularly important when also considering that the cacao
planting frontiers are being expanded towards marginal lands in tropical and subtropical
areas.

EFFECTS OF ABIOTIC FACTORS ON SEEDLING GROWTH AND

DEVELOPMENT
Cacao exhibits a considerable genetic variability in morphological and physiological
traits (Daymond et al., 2002a, b). When it is multiplied via seeds, the seedling initially
displays an orthotropic growth pattern and exhibits repeated cycles of leaf flushes with spiral
phyllotaxy (Vogel, 1975). Furthermore, leaf emission occurs in a rhythmic way relatively
independent of climate, suggesting that growth rhythm is under endogenous control (Vogel,
1975). However, after achieving approximately 1.0 to 1.2 m in height, the orthotropic growth
ceases and the plant emits plagiotropic branches (Garcia and Nicolella, 1985). The number of
plagiotropic branches varies from three to five, forming what is generally named the cup or
crown of the cacao tree (Cuatrecasas, 1964). Also, there are inter- and intra-specific variations
for the orthotropic growth pattern (Batista and Alvim, 1981; Garcia and Nicolella, 1985). In
contrast to Soria (1964) and Garcia and Nicolella (1985), Batista and Alvim (1981)
demonstrated that the genetic background of orthotropic growth is influenced by
4 Alex-Alan F. de Almeida and Raúl R. Valle

environmental factors. Environmental factors exert an influence of approximately 70% on the

development of the cacao plant at this stage (Garcia, 1973).
According to Greathouse et al. (1971) and Orchard et al. (1980), the specific stages of
leaf flushing in the orthotropic growth phase are: F-1 - bud swelling: leaf initiation and
unfolding; F-2 - leaf expansion: thin leaves, with strong anthocyanin pigmentation and apical
bud dormant; I-1 - leaf expansion complete: rapid greening and apical bud dormant; I-2 -
entirely expanded dark green leaves: apical bud dormant. Immediately after the emergence of
the next leaf, leaves in the initial F-1 stage are red and are positioned almost vertically
upwards. However, with expansion (F-2 stage) the stem:petiole angle decreases to
approximately 90° such that the petiole remains horizontal and, with an angle of
approximately 90° between the leaf blade and petiole, the blade hangs vertically downwards
(Abo-Hamed et al., 1983). During the F-2 stage the leaves are very thin and flexible. This
condition is maintained during leaf expansion until they enter the period of fast greening
(beginning of stage I-1), when the stem:petiole and leaf:petiole angles increase. Up to the
moment that the leaf blade becomes completely green, it thickens and becomes rigid with a
pronounced cuticle, and both the stem:petiole and leaf blade:petiole angles increase resulting
in the petiole and blade being positioned almost vertically. Subsequently the petiole:leaf blade
angle decreases approximately 90° and the blade is positioned horizontally (lower leaves).
This position is maintained until leaf senescence (Abo-Hamed et al., 1983).
Although the duration of the several stages of the flush cycle may vary depending on the
growth conditions [cf., for example, Greathouse et al. (1971) and Vogel (1975)] the basic
development pattern is relatively fixed, that is, the beginning of the cycle is marked by
swelling of the apical bud followed by unfolding of all leaves in the flush (Abo-Hamed et al.,
1983). The end of leaf production and the growth limitation of subsequent leaves within a
flush, which are smaller in size, are probably consequences of carbohydrate limitations, as
noted by Machado and Hardwick (1988b). They proposed that a new flush occurs following
an interflush period in which the carbohydrate supply is replenished; with the return of a
favorable carbohydrate balance a new leaf flush takes place.
According to Baker et al. (1975), (i) plastid length, breadth, number of grana and lamella
index per plastid, as well as the number of lamella per granum, increase with chlorophyll
(Chl) content; (ii) the number of chloroplasts per cell remains constant (three) during leaf
development; (iii) lutein, neoxanthin, violaxanthin, α- and β-carotenes are the carotenoids
found in measurable quantities in leaves; (iii) α-carotene is present in slightly higher
quantities than β-carotene; (iv) both xanthophylls and carotenes increase linearly with chl
content during leaf development; and (v) the phase of maximum chloroplast development
occurs after the end of leaf expansion. Baker and Hardwick (1973) did not observe changes in
the molar relation of chl a:b. However, Baker and Hardwick (1974) and Baker et al. (1975)
observed that (i) the photosynthetic capacity and the chl content increased in parallel during
flushing; (ii) the maximum photosynthetic capacity per chl unit remains constant during
chloroplast development; (iii) the saturating irradiance (Is) increases with increasing chl
content; (iv) the total number of photosynthetic units increases one by one during leaf growth;
(iv) the photosynthetic efficiency increases per chl unit during leaf development; (v) the
thylakoid formation occurs in parallel with chl synthesis during chloroplast development; (vi)
the maximum phase of chloroplast development may not occur until after the end of leaf
expansion; (vii) the phase of maximum synthesis of chl coincides with the end of the linear
phase of leaf expansion; (viii) Rubisco activity increases with the content of chl and can be a
 Cacao: Ecophysiology of Growth and Production 5

limiting factor for photosynthesis; and (ix) the chl content can be used as a general indicator
of chloroplast development.
The apparent absence of green pigmentation in young cacao leaves is not related to the
delay in chloroplast development, but is due to the fact that chloroplasts are initially very
small and usually few (Whatley, 1992). The delay in greening of leaves is associated with a
late increase in the total extension of the thylakoid system inside each cell that accompanies
the subsequent increase in plastid size (Whatley, 1992). Lee et al. (1987) reported that the
changes in color shown by cacao leaves during its ontogeny resulted from varying levels of
different pigments, like anthocyanins and/or phenols, which initially mask the chl content
(Figure 2). Content of these pigments fluctuates during leaf expansion (total content of
phenols remains high and that of anthocyanins decreases with maturity).
On the hypostomatous cacao leaf surface (Hardwick et al., 1981) four main kinds of hairs
are found, the most common of which are thick and short, with a multicellular rounded head
(Abo-Hamed et al., 1983). In 4-d-old leaves, stomata are found only in the midribs, primary
lateral and minor veins; however, only those of the midribs are completely developed and can
open. Stomata of the secondary and lateral ribs open within 2-4 d, and those in inter-veinal
regions begin to develop and can open towards the end of the leaf expansion phase (Abo-
Hamed et al., 1983). According to Abo-Hamed et al. (1983) the maturation of stomata is
correlated with the presence of chl. The cuticle thickness increases at a constant rate during
leaf expansion and achieves maximum thickness during the I-1 stage. Additionally, stomata
and cuticle development follow a pattern closely correlated with the growth stages of the leaf.

Figure 2. Seedling flushing in the field and details of a flush. Note the color variability in the seedling
leaves. Photos by the authors.

Solar Radiation

Müller et al. (1992), evaluating photosynthesis in cacao leaves of different ages grown
under two irradiances, found a strong effect of light intensity during leaf development on
photosynthetic capacity of mature leaves. In cacao, shade leaves often exhibit greater total chl
concentrations per unit mass than sun leaves (Merkel et al., 1994). Ontogenetic changes in the
leaf chl a:b ratio is a controversial issue, as it has been shown to increase continuously
6 Alex-Alan F. de Almeida and Raúl R. Valle

(Merkel et al., 1994) or to be held in check (Baker and Hardwick, 1973, 1974) during leaf
development. Costa et al. (1998), evaluating the growth of cacao seedlings under different
light levels and N concentrations, verified increases in leaf lamina thickness positively
correlated with the increase of light intensity, independent of N concentration. Under shade
conditions and high N concentrations, they observed that the spongy parenchyma had a
reduced number of small cells distributed over vast intercellular spaces, while short cells
constituted the palisade parenchyma, interspersed by quite large spaces.
Seedlings grow slowly under full sunlight and some degree of shade is beneficial for their
establishment (Okali and Owusu, 1975). Thus, shelter trees have been invariably
recommended for the establishment of cacao seedlings (Alvim, 1977), which would be
gradually removed with increasing self-shade (Byrne, 1972). The slow growth of seedlings
under full sunlight is due to constraints on leaf expansion, which is most likely caused by
excessive transpiration that would induce leaf water stress (Okali and Owusu, 1975). In
contrast, shaded leaves show higher relative water content and fewer stomata per unit leaf
area than unshaded leaves. The perpendicular positioning of young leaves in the flush, the
presence of superficial hairs, the rapid synthesis of cuticle and the development of stomata,
which are all limited to the initial stages of leaf expansion, act to reduce the water loss of
developing leaves (Abo-Hamed et al., 1983).
The saturating irradiance for photosynthesis of a mature cacao leaf varies from 6%
(Baker and Hardwick, 1973) to 30% (Okali and Owusu, 1975) of full sunlight, while for a
developing leaf Is increases with increasing chl content (Baker and Hardwick, 1976). At Is,
younger leaves (PF1) in the flush showed photosynthetic rates significantly higher than those
from both the oldest leaves of PF2 (middle age leaves) and PF3 (old leaves) (Machado and
Hardwick, 1988a). Under field conditions, the total potential contribution of photosynthesis of
the oldest leaves would be remarkably reduced as a consequence of mutual shading in the
crown interior (Machado and Hardwick, 1988b).
Cacao seedlings show increasing rates of net CO 2 assimilation (A) as the
photosynthetically active radiation (PAR) increases to values in the range from near 400
(Joly, 1988) to 750 (Da Matta et al., 2001) µmol m -2 s-1, which corresponds to 20 to 35% of
PAR at full sunlight, when Is is reached. About 95% of the maximum A is obtained with half
Is. At full sunlight, A declines at light exposures above 30% of the global radiation (Okali and
Owusu, 1975; Hutcheon, 1976), in which A of many plant species usually peaks (Alvim,
1965). Baligar et al. (2008), measuring A in three cacao clones, showed that increasing PAR
from 50 to 400 µmol m–2 s–1 led to increases in A by about 50%, but further short-term
increases in PAR up to 1500 µmol m–2 s–1 had no effect on A.
For leaves developed under shade conditions, a decrease in A occurs below 20% of full
sunlight (Owusu, 1978), whereas Okali and Owusu (1975) found in shade leaves that Is
occurred at around 3-4% of full sunlight. Therefore, the varying Is should be a reflection of
differences in growth and measurement conditions.
Cunningham and Burridge (1960) by submitting cacao seedlings to full sunlight and
heavy shading (85%) showed that water and mineral nutrients were the most crucial factors
for growth promotion at full sunlight. According to Okali and Owusu (1975) nutrients may be
more important than shading for seedling establishment. Costa et al. (1998), by growing
cacao seedlings under different light levels and N doses, observed a significant increase in chl
content of mature and more shaded leaves, especially at higher N doses. The increase at all
irradiances tested occurred in the presence of up to 10 mM of N, but above that concentration
 Cacao: Ecophysiology of Growth and Production 7

chl content leveled off. Okali and Owusu (1975) and Merkel et al. (1994) found similar
results regardless of chl content and PAR for other cacao genotypes.
For cacao plantation systems it is hypothesized that once the nutritional demands are met,
the yield of the understorey crop is dependent mainly on the accessibility to solar radiation
(Cunningham and Arnold, 1962). In any case, Hutcheon (1973) evaluating the growth of the
cacao clone Scavina-6 and two other Amelonado genotypes under full sunlight, verified that
the former is more tolerant to excessive radiation than the Amelonados. Scavina-6 kept more
leaves and grew well at full exposure, whereas the leaf area of Amelonado was halved,
showing that seedling capacity to grow under full sunlight is also related to genetic factors.

Temperature

Cacao growth and development, as also occur with other tropical woody species, is
highly dependent on temperature (Raja Harun and Hardwick, 1988b; Hadley et al., 1994).
This can be observed through the reduction of photosynthetic rates at temperatures above or
below the range considered optimal (Raja Harun and Hardwick, 1988b). Low temperatures
affect mainly vegetative growth (Alvim et al., 1977), flowering (Alvim, 1988) and fruit
development (Sale, 1969).
Daymond and Hadley (2004), evaluating the effect of temperature on the initial
vegetative growth, emission of chl fluorescence and leaf chl content in seedlings of four
cacao genotypes, showed the existence of genetic variability in response to temperature stress.
In a study by Amorim and Valle (1990), in which gas exchanges were measured on 8-month-
old cacao seedlings submitted to different root temperatures (10, 20, 30 and 40°C), it was
found that at root temperatures between 20 and 30°C the water fluxes and stomatal
conductance were greater than at any other soil temperature. Consequently, the resistance to
water movement was low and A rates were higher.
Daymond and Hadley (2004) calculated cacao base temperatures (temperatures below the
point at which development ceases) that varied from 18.6 to 20.8°C. These values are
considerably higher than those obtained by Alvim (1977), who calculated a base temperature
of 9°C for Catongo based in its pod growth ability. The higher base temperatures of
Daymond and Hadley (2004) were probably due to the fact that their studies were conducted
under conditions of low irradiance, while the studies by Alvim (1977) were carried out under
field conditions.
Okali and Owusu (1975) observed that leaf temperatures of unshaded cacao seedlings are
higher than that for shaded seedlings, and the transpiration rate (E) is usually at its maximum.
The vertical orientation of developing cacao leaves decreases light interception, and, thus,
leaves can be kept cooler and water loss may be limited in comparison to mature horizontally
positioned leaves (Abo-Hamed et al., 1983). Low chl content and high anthocyanin
pigmentation of expanding leaves also contribute to maintain relatively low leaf temperature
and, together with the presence of relatively dense hair cover on the leaf surface, may
contribute to further decreased E (Abo-Hamed et al., 1983).
8 Alex-Alan F. de Almeida and Raúl R. Valle

Relative Humidity

Stomatal opening in cacao leaves is related to the air relative humidity (RH) – stomata
are kept more open at higher than lower RH (Sena Gomes et al., 1987). However, due
probably to a high cuticular transpiration, stomatal closure does not always efficiently control
water loss. In fact, some cacao genotypes do not show high stomatal resistance under water
deficit and low RH (Raja Harun and Hardwick, 1988). This is in contrast to other genotypes
where an efficient mechanism of stomatal regulation may strongly limit E under water deficit
conditions. This may be an important acclimation strategy against drought (Balasimha, 1988).
In fact, the extreme sensitivity of some cacao genotypes to low RH can be a limiting growth
factor in areas where RH is low (Sena Gomes et al., 1987). In such areas, growth may be
adversely affected as a result of stomatal closure and, consequently, a reduction of A is to be
expected. The inefficient use of water (low water-use efficiency, WUE= A/E) at low RH
probably would cause water deficit in shoots under limited soil water supply (Sena Gomes et
al., 1987). According to Baligar et al. (2008), increasing VPD leads to decreases in A with
parallel increases in E, thus ultimately lowering WUE. Furthermore, under the conditions of
their measurements, Baligar and co-workers showed an unusually poor response of gs to
VPD, which could limit the ability of cacao to grow where VPD is high. Nonetheless, high A
and high WUE of cacao in a high RH regime are consistent with many species of the humid
tropics (Alvim, 1977; Wood, 1985).

Water Supply

Both water excess (flooding) and water deficit impair cacao growth and development.
Flooding is an important barrier for the initial growth and establishment of cacao in places
subjected to periodic waterlogging. Waterlogging occurs in some cacao producing areas in
Brazil, Ghana, Nigeria and Ivory Coast, where total rainfall often largely exceeds
evapotranspiration, which creates hypoxic conditions in the soil (Sena Gomes and Kozlowski,
1986). In the state of Bahia, Brazil, the hypoxic condition occurs after heavy rainfalls in
locations with shallow soils, as well as areas with hydromorphic soils and margins of rivers
after periodic floods. Decreases in growth, leaf area, gs and A, as well as epinasty, leaf fall,
formation of hypertrophic lenticels and adventitious roots in submerged stems have been
found during the flooding period. The extent of these changes is genotype dependent (Rehem,
2006). Sena Gomes and Kozlowski (1986) reported similar results for cacao seedlings of
Catongo. Bertolde (2007) studying 35 cacao clones in the young phase found some tolerant
genotypes and demonstrated that there are no significant relationships among physiological
variables in response to flooding and heterozygosis patterns. In any case, growth inhibitions
in plants submitted to soil flooding reduce pod production due to the delay of flowering and
suppression of vegetative growth (Sena Gomes and Kozlowski, 1986). A similar result was
also found by Mariano and Monteiro (1980), who showed an approximately 60% reduction in
pod production after a flooding of some weeks.
Water deficit, in addition to impairing crop yield of adult plants, also negatively affects
growth of cacao seedlings. Almeida et al. (2002b) evaluated the effects of water stress in 5-
month-old seedlings of eight cacao genotypes grafted on the variety Cacau Comum. They
demonstrated that (i) drought resistance occurred, at least partially, through osmotic
 Cacao: Ecophysiology of Growth and Production 9

adjustment in most genotypes; (ii) the genotypes maintained values of relative water content
at around 90% with a leaf water potential (Ψw) of approximately -1.0 MPa, gradually reaching
55% at -3.5 MPa; (iii) a significant increase occurred in leaf concentration of K and P during
the dehydration process of some genotypes at Ψw of -1.5 MPa.
In cacao, leaf carbon export rate is sharply reduced with the decline of Ψ w between -0.8
and -2.0 MPa. This reduction is strictly associated with both A and the export capacity, which
is strongly reduced when A approaches zero (Deng et al., 1989). Deng et al. (1990) found
values of A close to zero at Ψw below -1.6 MPa, while average values of A of about 2.2 µmol
CO2 m-2 s-1 were found at approximately -1.5 MPa. This was also observed by Gama-
Rodrigues et al. (1995), who attributed to K fertilization a role in the mechanisms to reduce
the negative effects of severe water stress in cacao, independently of the K source.

Nutrients – K

Cacao requires large amounts of K, and approximately 700 kg ha -1 are necessary to

produce 1000 kg ha-1 of seeds per year (Thong and Ng, 1980). This element corresponds to
about 70% of the nutrients in the sap of the cacao xylem (Martins, 1976). Orchard (1978)
showed in cacao seedlings that high doses of K (5 mmol L -1) promoted increased leaf area
without affecting dry biomass production. He also showed that there is an inverse relationship
between leaf transpiration and K availability. Under field conditions cacao trees well supplied
with K may be more tolerant to the adverse effects of water stress (Bosshart and Uexkhull,
1987). According to Gama-Rodrigues et al. (1995), KCl was the K source that induced the
lowest gs, E and A, which differed significantly from the sources K 2SO4 and Kalsilite. There
were no significant effects among K sources on WUE, Ψ w and internal CO2 concentration.
The largest dose of K added to the soil, however, resulted in smaller values of A and larger
values of E, thus decreasing WUE but without affecting shoot dry biomass.

FINE ROOTS IN CACAO TREES

The capacity of a tree to absorb water and nutrients depends, among other factors, on fine
root dynamics (mortality rates and regrowth of fine roots) as well as its variation in time. The
turnover of fine roots can contribute a significant proportion of recycled nutrients in
agroforestry systems, where the competition between forest trees and crops for water and
nutrient depends on the temporal pattern of fine root regrowth (Muñoz and Beer, 2001).
Studies done under greenhouse conditions have shown that cacao seedlings, as with many
tropical woody species, have alternate phases of root and shoot growth (Taylor and Hadley
1987). Field evaluations of fine root growth in cacao trees have been very much limited to
static inventory data. For example Beer et al. (1990) assessed biomass only at a single
harvest. Although Kummerow et al. (1981, 1982) accomplished a pioneer study with cacao
root dynamics, their methodology involved sequential coring, which can lead to significant
over or underestimation of root productivity (Anderson and Ingram, 1993).
Biomass of active fine roots (≤ 2 mm at 0-10 cm soil depth) of 11-yr-old cacao plants
shaded by Erythrina glauca was reported as 0.8 Mg ha-1 (Kummerow et al., 1981) and 0.4 Mg
ha-1 for roots ≤ 1 mm (Kummerow et al., 1982). In cacao plantations shaded by E.
poeppigiana or Cordia alliodora the biomass of cacao fine roots, on average, was
10 Alex-Alan F. de Almeida and Raúl R. Valle

approximately 1.0 Mg ha-1 and changed little during the year (Muñoz and Beer, 2001). The
maximum value found at the beginning of the rainy season was 1.85 Mg ha -1 in the cacao-C.
alliodora system compared to 1.20 Mg ha-1 for cacao-E. poeppigiana. According to Muñoz
and Beer (2001), if the annual turnover of fine roots were constant and close to 1.0 Mg ha -1 in
both systems, the input of nutrients would be 23-24 (N), 2 (P), 14-16 (K), 7-11 (Ca) and 3-10
kg ha-1 yr-1, which would represent 6-13% and 3-6% respectively of the total input of nutrients
in the organic matter for the C. alliodora and E. poeppigiana systems.
Growth and longevity measurements of fine roots (diameter < 1 mm) performed by Silva
and Kummerow (1998) directly in the soil-litter interface, in a Brazilian cacao plantation
shaded with E. fusca, revealed that the elongation rates of fine roots were, on average, 3.7 and
1.8 mm d-1 for ramifications of first and second order, respectively. They also found that the
average period of functional life of these roots was 3 d, with a range of 1 to 10 d, and that
growth of fine roots was closely correlated with rain frequency. Data reported by Silva and
Kummerow (1998), Kummerow et al. (1982) and Medeiros et al. (1987) suggest that life
spans of fine roots are short, but, in compensation, they have high turnover rates.
According to Muñoz and Beer (2001), the low fine root production between October and
January in Costa Rica is probably the result of internal distribution of carbon to pod
production during the six months that it takes for fruits, harvested in December-January, to
develop and mature. Furthermore, these authors reported that partial pruning of cacao in
December might also reduce cacao fine root production during the following months due to
reduction in photoassimilate production and internal competition between flushing and root
renewal (Muñoz and Beer, 2001). A negative correlation between number of new leaves per
flush and number of growing root tips was reported by Kummerow et al. (1982). The
observations of Muñoz and Beer (2001) in Costa Rica are consistent with observations in
Brazil that the number of active cacao root tips (0-10 cm soil depth) was lowest in the driest
month and highest in the wettest month (Kummerow et al., 1982). Therefore, this temporary
increase in fine root production seems to be part of the normal phenological cycle of cacao
(Muñoz and Beer, 2001) since formation of new root tips in cacao grown in nutrient solution
was reported only during the rainy month of July in Costa Rica (Rodriguez et al., 1963).
Additionally, the results regarding fine root dynamics of shaded cacao plantations suggest
that a greater proportion of nutrients in organic or inorganic fertilizers is absorbed by cacao at
the beginning, rather than at the end, of the rainy season (Muñoz and Beer, 2001).
Experiments (B.V. Leite, 2007; unpublished results) conducted with 35-month-old cacao
plants in a semi-arid region of Bahia, Brazil, in which irrigation was performed by placing
one or two plastic pipes at 0.30 m from the trunk with drip emitters 0.50 m apart, revealed
that in plants grown in plots with two irrigation pipes, the longitudinal root length (along the
planting line) was about 1.70 m and the perpendicular (to the planting line) root length was
about 1.0 m for both sides of the trunk. For the cacao plots with only one pipe the longitudinal
root length (along the planting line) was, on average, 2.0 m. Perpendicularly to the planting
line a clear root displacement towards the dripping pipe was observed. At the opposite side,
without a pipe, the roots were finer, fewer and smaller in length. Clearly, the greater
concentration of radicels was observed near the irrigation pipe. In the two-pipe plots the tap
roots reached a depth of 1.3 m against 1.0 m for the one pipe irrigation plot, indicating that
the growth limitation was due to the size of the wetted soil area (Figure 3).
 Cacao: Ecophysiology of Growth and Production 11

Figure 3. Root system of cacao plants grown in plots with one (left) and two (right) irrigation pipe lines.
Note the skewed distribution of roots in the plant grown in plots irrigated with one pipe line. Photos by
the authors.

The wetted area, after 2 h of irrigation, reached 0.7 m in depth and 0.6 m diameter. In this
system 3% of the roots were concentrated between 0-0.1 m, 80% between 0.1 to 0.6 m, and
17% below 0.6 m depth. For the one-pipe plot the wetted area was 0.4 m in depth and 0.5 m
diameter after 2 h of irrigation. In this system 9% of the roots were distributed between 0-0.1
m, 56% between 0.1-0.4 m, and 35% below 0.5 m depth, and these were skewed to the side of
the irrigation pipe. The distribution of fine, medium and coarse roots followed the same
patterns, that is, in the two-pipe plots, they were distributed along the two sides of the pipes.
The fine roots were at a higher percentage while the medium and coarse roots presented about
the same percentage. In the one-pipe plots the percentage of fine roots was higher than
medium and coarse roots. However, fine roots were generally low in number as well as
distributed preferentially on the side of the irrigation pipe. From these preliminary results
(B.V. Leite, 2007, unpublished data) it can be concluded that the behavior of the plant as a
whole, and the root system in particular, is different from the traditional cropping areas in
which soil moisture depends on rainfall. In the traditional cacao growing areas the fine root
system is concentrated in the first 0.2 m soil depth, and most fine roots are located in the first
0.10 m (Leite and Cadima Zevallos, 1991).

THE FLUSHING CYCLE

Formation of a cacao branch is characterized by alternate periods of growth and
quiescence, i.e. the flush cycle (Greathouse et al., 1971). During flush activity a variable
number of leaves is produced in fast succession, after which the leaves rapidly expand (Baker
and Hardwick, 1973). Under field conditions, the flushing cycle is activated usually by
environmental changes (Almeida et al., 1987). However, the cycle persists under constant
12 Alex-Alan F. de Almeida and Raúl R. Valle

conditions (Orchard et al., 1980) that indicates there is an endogenous form of control along
the flush cycle. Studies reported by Alvim et al. (1974b), Orchard et al. (1981) and Abo-
Hamed et al. (1981) showed that the alternating growth and quiescence periods of the apical
bud of the cacao branch (the flush cycle) is controlled by a regular cycle of accumulation and
depletion of growth promoter and inhibitor compounds in the bud. Instead of in situ synthesis,
transport of abscisic acid (ABA) (Alvim et al., 1974b) and cytokinins (Orchard et al., 1981)
towards the apical bud has been proposed.

Growth Promoters and Inhibitors

The highest levels of auxin and cytokinin are found in young, expanding cacao leaves of
a new flush (Orchard et al., 1981). This suggests that these leaves could be acting either as
sinks, importing auxin and cytokinins from other parts of the plant, or as auxin and cytokinin
sources, from where these promoters are exported to the apical bud and other plant parts. The
auxin and cytokinin levels in young leaves close to the apex could also reflect the increase in
transport of these compounds to the higher parts of the flush, including the apex (Abo-Hamed
et al., 1984).
Analyses of new and previous flushes of cacao plants under controlled environment
(Orchard et al., 1980) or under water deficit conditions (Alvim et al., 1974b) showed high leaf
levels of ABA when the apex is in the quiescence phase of the cycle. In agreement with the
distribution of [14C]ABA in cacao, it is suggested that ABA is exported from the leaf, via
phloem, to areas of low ABA concentrations such as buds and young leaves, which is
consistent with source/sink relationships. ABA induces branch quiescence or, alternatively,
helps to maintain the branch in the non-active state. Appearance of radioactivity in mature
leaves can be due to the lateral movement of [ 14C]ABA in the xylem due to transpiration flow
(Abo-Hamed et al., 1981).
Cacao seedlings under optimal irrigation conditions also show flush growth, which is
related with changes in the flush ABA levels (Orchard et al., 1980). These observations
suggest that, despite adequate water supply, the seedlings suffer from an internal water deficit
when the water uptake rate by the root system is lower than the rate of losses by leaf
transpiration. This would cause the accumulation of ABA in the apex and, consequently, the
imposition of apical bud inactivity at least in part of the interflush period. This hypothesis was
confirmed by Abo-Hamed et al. (1983) after relating physical parameters (leaf blade position,
population of hairs, cuticle thickness and number of stomata) of leaves of a new flush with
transpiration and, therefore, with the water balance of cocoa seedlings in different stages of
the flush cycle.
The decrease in diffusive resistance and increase of water loss at the I-1 stage coincide
with the period of higher ABA levels in leaves of the new as well as previous older flushes
(Orchard et al., 1980). This formation of ABA after stage F-2 could be the result of water
stress created by leaf transpiration exceeding the capacity of water uptake. The rate of
maximum water loss of a new flush is only reached when the leaves are fully expanded,
stomata completely developed and the leaves horizontally positioned (Abo-Hamed et al.,
1983).
Iserentant (1976), comparing the effects of removal of all leaves of cacao seedlings in
stages equivalent to F-1 and I-2 of the flushing cycle, verified the break of apical dormancy
after defoliation in both stages. Iserentant (1976) concluded that quiescence in seedlings was
 Cacao: Ecophysiology of Growth and Production 13

due to correlative inhibition instead of a true dormancy. Similar results were found by Abo-
Hamed et al. (1981) when only leaves in stages F-2 and I-2 of a new flush were removed.
This suggests that there are no reversible or irreversible stages in cacao seedlings (Abo-
Hamed et al., 1981), as described by Vogel (1975) for cacao trees grown under field
conditions. However, even in plantlets, it seems that two physiological stages exist in the
period of bud growth inhibition (Abo-Hamed et al., 1981).

Nutrient X ABA Interactions

Vogel (1975) described the initial phase of quiescence as a reversible phase that occurs
during F-2 (Abo-Hamed et al., 1981). It appears that this phase in cacao seedlings can be
initiated by the competition for nutrients between the apex and the developing leaves of the
flush. According to Orchard et al. (1980), F-2 leaves of a new flush contain low levels of
ABA and, consequently, are not the source of this compound. Since the removal of leaves in
the flush stimulates the apical bud to produce new leaves, this stimulus could result in the
removal of a powerful drain for nutrients instead of the removal of an ABA source (Abo-
Hamed et al., 1981). Only the second phase, previously known as the irreversible phase
(dormancy in I-1 and I-2), is controlled by the transport of ABA from mature leaves to the
apex. In this case, the role of ABA seems to be in maintaining rather than initiating apical bud
quiescence (Abo-Hamed et al., 1981).
Removal of new flush leaves at the I-2 stage causes reduction in the quiescence period of
the apex (Abo-Hamed et al., 1981). Since in this period the new flush leaves contain high
ABA levels (Orchard et al., 1980), probably the inhibitory effect of new flush leaves in I-2 is
due to ABA and its transport to the apex (Abo-Hamed et al., 1981). The removal of previous
flush leaves at the I-1 stage also promotes reduction in the apex dormancy period. These
mature leaves are sinks for nutrients (Abo-Hamed et al., 1981) and also contain high ABA
levels (Orchard et al., 1980) and, therefore, they could serve as an ABA source for the apex,
like mature leaves of new flushes (Abo-Hamed et al., 1981).
Application of ABA always extends the inactive period of growth. Alternatively, the
application of growth promoters (GA3 or zeatin) reduces the duration of the subsequent
quiescence stage of the apex. Probably, endogenous inhibitors in leaves are overcome by the
effects of growth promoters that shorten the quiescence period. On the other hand, the fact
that GA3 applied to new flush leaves in the F-2 stage does not promote a reduction effect on
the duration of the apex dormancy period reinforces the hypothesis that the apical dominance,
at this stage, is due to competition by nutrients instead of the presence of hormonal inhibitors
in the leaves (Abo-Hamed et al., 1981).

Source–Sink Relationships

Developing leaves display a low photosynthetic capacity (Baker and Hardwick, 1973)
and are considered sinks for photoassimilates during most of the expansion phase. During this
period, leaf contents of fructose, glucose and sucrose increase, presumably due to imports
from older leaves. However, when leaf expansion ends, glucose and fructose pools decline
and sucrose concentration increases, which reflects the end of photoassimilate import and the
beginning of endogenous sucrose synthesis (Baker and Hardwick, 1975). The simultaneous
14 Alex-Alan F. de Almeida and Raúl R. Valle

development of several leaves within a flush constitutes, however, a very large carbohydrate
consumption event during a period of 10 to 15 d that exceeds the available current
photosynthate (Machado and Hardwick, 1988b). Although there is little competitive influence
among developing leaves within the flush, leaves that develop at the end of a normal flush,
are smaller in size (Machado and Hardwick, 1988a). This strongly suggests that the
subsequent leaves in a normal flush grow under the increasing deficit of a presumable factor
that supports growth, such as carbohydrate, which may negatively influence the potential for
full size growth. In the interflush the stock of carbohydrates is reestablished, and, with the
return of a favorable carbohydrate balance, the appearance of a new flush can occur
(Machado and Hardwick, 1988b).
F-2 leaves of a new flush also represent a powerful sink for nutrients, which would be in
competition with the shoot meristem (Abo-Hamed et al., 1981). Chemical composition of
cacao leaves showed significant decrease in total N levels of mature leaves during the
development of a new flush (Santana and Igue, 1979). This decline was attributed to an
increase in N mobilization from mature to expanding leaves. At this stage in the flushing
cycle, there is a large quantity of endogenous auxin (Orchard et al., 1981) and a high rate of
14
C translocated out of the labeled leaf as compared with other stages (Sleigh, 1981).
Additionally, at the beginning of F-2, the amount of carbohydrate translocated from mature
leaves to the developing flush exceeds (positive balance) that required for leaf growth. This
excess can be stored in the newly formed stem and used for its expansion, mainly in mid F-2,
when the leaves of the flush constitute a very strong carbohydrate drain, and the amount
required for leaf growth considerably increases (Machado and Hardwick, 1988b).
In the I-2 stage, new flush leaves are completely expanded, show maximum
photosynthetic capacity, and are capable of photoassimilate export (Baker and Hardwick,
1975). At the end of this stage, cytokinin, imported from the root, accumulates in the apical
bud and helps to activate the breaking of bud quiescence in the transition to stage F-1 (Alvim
et al., 1974b; Orchard et al., 1981).
Observations reported by Sleigh (1981) and later confirmed by Machado (1986) showed
that, at the F-2 stage, expanding leaves import considerable amounts of 14C that is assimilated
by the mature leaves. Unexpectedly, if a high proportion of photoassimilates is removed from
mature leaves during the F-2 stage, a corresponding increase of photosynthesis does not
occur.
This suggests that the requirement of carbon for growth, at least for leaves, does not
directly control photosynthesis of mature cacao leaves (Machado and Hardwick, 1988b).
According to Baker and Hardwick (1973), A of older leaves, with 90% more chl, is only 10-
20% higher than that of leaves at the end of flushing. Consequently, the photosynthetic
efficiency of older leaves is significantly smaller. Therefore, in leaves of cacao, the decline in
photosynthetic efficiency occurs during the period between the end of flushing and onset of
senescence.

Climate

Information exists about the periodicity of cacao leaf production in relation to changes of
climatic elements, especially the effect of temperature (Couprie, 1972), solar radiation
(Snoeck, 1979), and rainfall (Alvim and Alvim, 1978). Almeida et al. (1987), carried out a
study in Ilhéus, Bahia state, Brazil, that investigated the relationship between actual
 Cacao: Ecophysiology of Growth and Production 15

evapotranspiration and reference evapotranspiration to quantify the effects of water deficit on

cacao flushing. The actual-to-reference evapotranspiration ratio was used in the original
expression of degree-day in order to evaluate the concomitant effects of temperature and
water on flushing. Nominating this weighted function penalized degree-days (PDD), Almeida
and co-workers verified that the weighted function represented approximately 80% of the
existing interactions between flushing and weather factors. The high degree of significance of
this function is due to the fact that PDD represents the joint effects of energy and water. For
the conditions of Bahia the PDD that best explained flushing was that calculated three weeks
before leaf emission. In Bahia, the main flushing period, characterized by relatively short
periods of vegetative rest, is from September to February. During those months, the energy
curve (averages day-1) is in ascension and coincides with the period of more intense flushing
(Almeida et al., 1987). Furthermore, in that area, the mean air temperature at the 8 th week,
combined with the solar radiation and insolation of the 9 th week, exert great influence on
flushing. In contrast, the period of decreasing energy (solar radiation, temperature and PDD)
coincides with those months in which no flushing occurs (Almeida et al., 1987).
The seasonal influence of temperature on phenological events is evident in the flowering,
fruiting and pod growth profiles found by Cazorla et al. (1989), who also observed low
flushing during June to September. This period is characterized by relatively low
temperatures. Therefore, the absence of flushing and flowering in south Bahia is intimately
related to the decrease of fruiting, which culminates with an almost total absence of pods
from January to March of the following year (Leite and Valle, 2000). In Bahia, cacao usually
stops flushing from March to July and remains in a relative vegetative rest afterwards (Alvim,
1977). Therefore, the period of minimum flushing varies from 90 to 120 d – June to
September (Alvim et al., 1974a).
Alvim et al. (1974a), Alvim (1977) and Almeida et al. (1987) studied the influence of
rainfall and water balance on flushing within defined months of the year. According to Alvim
(1977), in most cacao producing areas, the beginning of leaf emission follows the first rains
after a dry period. For Sale (1968), rainfall or irrigation, after a dry period, is a necessary
condition to elicit flushing. However, under field conditions, flushing begins in response to an
increase of plant Ψw, but not necessarily to rainfall, after a period of humidity stress (Almeida
et al., 1987). Once apical bud growth resumes, rainfall plays a decisive role on leaf expansion.
This may be one of the possible reasons to best explain the vigorous flushing associated with
rainfall that follows a period of humidity stress (Alvim, 1977).
Almeida et al. (1987) showed that during the vegetative inactive period, the changes in
soil moisture availability and, consequently, the sequence of water deficits and excesses, do
not seem to play a direct role in the phenomenon of bud dormancy break. However, once the
apical bud dormancy is broken, water availability seems to exert great influence on the
intensity of leaf emission. Therefore, flushing becomes relatively intense when the period of
water deficit (storage of water in the soil below 50 mm) is followed by rains that replenish
soil water content (Almeida et al., 1987). Similar results were found by Sale (1968), Alvim
(1977) and Machado and Alvim (1981).
Physiological responses of cacao are affected when the soil water content is below 60-
70% of the maximum available soil water capacity (Alvim, 1960). Therefore, the
development of water deficit during the dry season is considered an important factor in the
control of flushing in field-grown cacao plants (Alvim et al., 1974a). At the beginning of a
new flush cycle the transition from inactivity to apical growth of a branch occurs immediately
16 Alex-Alan F. de Almeida and Raúl R. Valle

following the decrease of water deficit. This is exhibited through leaf shedding at the
beginning of the rainy season and the consequent reduction of the high shoot ABA levels in
the preceding period of water deficit (Alvim et al., 1974b). In contrast, the effects on the
apical bud activity, triggered by dry periods followed by rain, seem to occur through
modifications of the ABA:cytokinin balance. Increase in the ABA:cytokynin ratio inhibits
flushing (Alvim et al., 1974b). Removal of mature leaves produces modifications in that
balance as well. Therefore, after defoliation, ABA content in the apical bud decreases, while
cytokinins content increases considerably (Alvim and Alvim, 1976).
In tropical climates temperature oscillations are relatively small compared to temperate
climates. Nonetheless, trees grow intermittently within seasonal climatic sequences, and
flushing is considered the beginning of a growth cycle (Alvim, 1977). For cacao grown in
southeastern Bahia, the flowering, fruiting and pod growth patterns show low flushing in June
to September (Cazorla et al., 1989), a period characterized by relatively lower temperatures
(Leite and Valle, 2000). According to Almeida et al. (1987), temperatures above a reference
value (around 23°C) coincide with periods of increased flushing. Similar responses were
observed under controlled conditions by Sale (1968), when cacao plants submitted to 23.3,
26.6 and 30°C emitted leaves respectively at 95, 36 and 20 d intervals. Therefore, it seems
that high temperatures accelerate flushing initiation.

Role of Endogenous Factors on Cacao Flowering

Cacao produces caulescent flowers that arise from meristematic tissues above leaf scars
on the woody stems that are at least 2-yr-old (Figure 4). The floral meristem or cushion
produces flowers throughout the life span of the tree (Aneja et al., 1999). Cacao flowers
contain different polyphenolic compounds such as: (i) hydroxycinnamic acids, found in in the
periphery of the organs, except in the ovary where parenchyma cells are present and the
ovules occur; (ii) tannins; and (iii) anthocyanins. Both tannins and anthocyanins are located in
the epidermis of different floral parts. Anthocyanins are mainly restricted to ornamentation in
petals and staminoids (Alemanno et al., 2003). These external locations of phenolic
compounds may function as chemical protective barriers against damages caused by pests and
diseases (Alemanno et al., 2003).
In each flower the sepals begin to dehisce in late afternoon, and the flower is completely
open at the beginning of the following morning. During this period the anthers liberate pollen
and the stigma is receptive (Aneja et al., 1999). Although cacao flowers are hermaphroditic,
the anthers are covered by the lower half of the petals. Without the intervention of insects,
mainly Forcipomyia sp. (Diptera: Ceratopogonidae), pollination is impaired, since the viscous
pollen does not come into contact spontaneously with the stigma (Dias et al., 1997). In Bahia,
Brazil, high populations of this insect are found from March to July, which coincides with the
increased pollination that results in the main harvest.
 Cacao: Ecophysiology of Growth and Production 17

Figure 4. Counterclockwise: flowering on the trunk and branches of an adult tree, flowering in a 9-
month-old seedling in the field and details of a flower cushion. Photos by the authors.

The fly populations are low during the high temperature months. This is reflected in the
low and sporadic productivity that follows this fruit setting period (K. Nakayama, 2007;
unpublished results). If the flowers are not pollinated, they abscise 24-36 h after anthesis,
without visible signals of senescence (Aneja et al., 1999; Hasenstein and Zavada, 2001). If
pollination is accomplished and fertilization occurs, the ovary increases in size, the pedicel
enlarges, and the corolla wilts and deteriorates (Aneja et al., 1999).
The genus Theobroma represents an atypical example of an ovarian auto-incompatibility
system that is distinct from most vegetal incompatibility systems (de Nettancourt, 1977).
Incompatibility in cacao was first reported by Pound (1932), who noted the occurrence of
changes from self-incompatibility to self-compatibility during certain periods of the year. The
incompatibility system in cacao is complex, since it includes different degrees of cross and
self-incompatibility of some genotypes as well as a high level of self-pollination (Lanaud et
al., 1987).
Although there is no causal link between expression of enzymatic activity and
incompatibility, Warren and Sunnite (1995) argued that enzymes like isocitrate
dehydrogenase (EC 1.1.1.41 and 1.1.1.42), malate dehydrogenase (EC 1.1.1.37) and acid
phosphatase (EC 3.1.3.2) are indicative of incompatibility systems. In contrast, Aneja et al.
(1992, 1994) showed that the pollen grains of the self-incompatible clone ‘IMC-30’ did not
germinate at low CO2 partial pressure. However, ‘IMC-30’ pollen grains germinated under
high CO2 concentration, which helped overcoming incompatibility in self-pollinated flowers.
The main effect of overcoming incompatibility systems in self-incompatible genotypes occurs
when self-incompatible and compatible (foreign) pollen are mixed, which permits self-
18 Alex-Alan F. de Almeida and Raúl R. Valle

fertilization in self-incompatible trees (Glendinning, 1960; Bartley, 1969; Lanaud et al. 1987).
These results imply that, in Theobroma, a substance other than CO2 may condition the
maternal plant self-fertilization (Hasenstein and Zavada, 2001).
The percentage of flowers that sets pods in cacao is usually very low and varies from 0.5
to 5% (Aneja et al., 1999). This is partly due to the fact that the effective rate of self-
fertilization in auto-incompatible trees is low, whereas in self-compatible trees self-
fertilization can reach up to 43% (Yamada and Guries, 1998). The success of fruit (pod) set
can depend on two factors, the degree of pollen compatibility and the number of pollen grains
deposited on the stigma (Lanaud et al., 1987). There is some evidence that a high number of
pollen grains is beneficial for overall pod set (Hasenstein and Zavada, 2001). Moreover, a
high proportion of pollinations results from the visit of a single pollinator (Yamada and
Guries, 1998). At anthesis floral auxin concentration is low and the retention of flowers, as
well as fertilization and possibly pod development, depend on an external source or signal for
satisfactory pollen germination or fertilization, as found by Hasenstein and Zavada (2001).
They also noted an inverse correlation between the number of flowers per plant and floral
auxin content and suggested that the genetic control of self-incompatibility in cacao may be
modulated by the hormonal content of the flower. Hasenstien and Zavada (2001) concluded
that this variation in floral hormonal content could explain the great number of pods set in
some instances, even after incompatible pollinations.
A better understanding of the factors involved in floral abscission can be of great
horticultural value (Aneja et al., 1999). Abscission of flowers in cacao occurs within 2-3 d
after anthesis in the absence of pollination or fertilization. Even after manual pollination, only
a small percentage of pollinated flowers develops and produces pods (Hasenstein and Zavada,
2001). Although there is no recognizable abscission area in the flower pedicel at anthesis,
within 24 h starch grains appear in the pedicel cortex in the area of the incipient abscission
zone, and cell separation occurs in the epidermis at 2-3 mm acropetally from the point of
pedicel attachment to the flower pad (Aneja et al., 1999). Flower abscission due to
incompatible pollination begins before the pollinic tube reaches the ovules (Aneja et al.,
1994; Baker et al., 1997).
It has been observed that auxin concentration of developing floral buds is high and
depends on the number of flowers formed. However, when the buds mature, the auxin level
decreases during the initiation of the abscission process (Hasenstein and Zavada, 2001), while
ABA levels of the flower increase significantly before abscission (Aneja et al., 1999). In
contrast, most of the increase in ethylene production occurs simultaneously with abscission
(Aneja et al., 1999). There is no abscission in the absence of ABA even when ethylene is
present. Therefore, in cacao flowers, ABA seems to be the primary inductor of abscission,
while ethylene, although accelerating abscission does not seem to be required for the process
(Aneja et al., 1999). The increase of indoleacetic acid (IAA) levels in the cacao ovary, that
follows compatible pollination and fertilization (Baker et al., 1997), or application of
naphthalene acetic acid, can overcome the effect of ABA in the formation of the abscission
area and reduce floral senescence (Aneja et al., 1999).
Hormonal changes after compatible and incompatible pollination indicates a strong
increase of endogenous auxin after compatible pollination and a strong increase of ethylene
after incompatible pollination (Baker et al., 1997). These hormonal responses occur before
the interaction of the pollin tube with the ovule and suggest that the incompatibility system
may be influenced by auxin (Hasenstein and Zavada, 2001). According to Aneja et al. (1994),
 Cacao: Ecophysiology of Growth and Production 19

the ability of CO2 to overcome abscission and the subsequent self-compatibility can be the
result of minimization of ethylene effects. However, it is conceivable that high auxin levels
also override the abscission signal, affecting or controlling the auto-incompatibility response
in Theobroma (Hasenstein and Zavada, 2001).
Since flower production represents a considerable plant investment (Valle et al., 1990),
the number of flowers can affect the auxin partition to each flower. This may explain the low
concentration of endogenous auxin in clones with high flower productions (Hasenstein and
Zavada, (2001). Auxin concentration is significantly higher in cacao clones with low
production of flowers. Producers of a high number of flowers can show a 20-fold
improvement in pod set, in spite of the existence of a poor relationship between pod and
flower number. In addition, there is a strong negative correlation between the number of
flowers produced and the endogenous content of IAA, while the content of ABA did not
differ between the two sets (Hasenstein and Zavada, 2001).

EFFECTS OF ABIOTIC FACTORS ON YIELD

The adult stage of cacao is reached when the tree enters reproductive growth. Pod
production is small during the first years and increases as the tree matures. In general, most of
the pod yield is found on the trunk and branches (Figure 5). Bartley (1970), studying the yield
of cacao per tree, proposed to initiate selection of potentially high-yielding cacao trees in the
8th yr of planting. Since with young trees fruit bearing begins approximately at the 4 th yr, it
has been suggested that only at the 8 th or 10th year after planting do the trees express their
productivity potential (Dias and Kageyama, 1997).
Also, measurements of diameter and stem height, average crown diameter and leaf area
of cacao trees between 1- to 3-yr-old have been used as variables to estimate the yield
capacity of cacao (Batista and Alvim, 1981; Garcia and Nicolella, 1985). However, the same
abiotic stresses that reduce pod production can also indirectly decrease yield by increasing the
duration of the juvenile period, which varies widely within cacao genotypes. Flowering, for
instance, in some cacao hybrids can begin 18 months after planting in the field, while in other
genotypes flowering may be initiated after 3 to 5 yr (Sena Gomes and Kozlowski, 1986).
The most important determining biotic parameters of cacao yield are related to (i) light
interception, photosynthesis and capacity of distributing photoassimilate, (ii) maintenance
respiration, and (iii) pod morphology and seed formation (Zuidema et al., 2005). These three
items are crucial to seed yield and were examined in initial studies concerning cacao
production (Yapp and Hadley, 1994). On the other hand, these biotic parameters can be
modified by abiotic factors, such as those inhibiting, for instance, initiation of floral buds
(Alvim, 1977) or those that, after flowering and pod set, influence pod development (Wood,
1985). Climatic changes exert outstanding effects in the flush cycle, an important determinant
factor of yield (Almeida et al., 1987).
In the definition of an appropriate climate for cacao cultivation in forest understorey
conditions, it is well known that their microclimatic variables are completely different from
the standard weather conditions outside that environment (Bonaparte and Ampofo, 1977).
20 Alex-Alan F. de Almeida and Raúl R. Valle

Figure 5. Most cacao pods are beared in the trunk. Cacao pods of two different genotypes. Photos by
the authors.

Also, there are indications that the structure and dynamics of the cacao agroecosystem
exert some influence in the microclimate (Beer, 1987). Air temperature, RH and rainfall are
correlated and affect the seasonality of microclimatic conditions inside a cacao plantation
(Miranda et al., 1994). The magnitude of the meteorological parameters varies less in a hot
week than in a cooler week. The predominantly clear summer (average of 9 h daily sunshine)
and the cloudy winter (average of 5 h daily sunshine) regulate, to a great extent, the behavior
of microclimatic parameters above and inside a cacao plantation.

Temperature

According to Daymond and Hadley (2004), temperature is one of the main limiting
factors for cacao production, since temperature stress affects the seasonal variation in seed
yield. In a cacao plantation, the difference in temperature measured outside and inside canopy
layers was around 2 ± 0.5°C, although higher values have been registered during summer
weeks. The undercanopy maximum temperatures are characterized by drops varying from 1
to 3°C and temperature differences inside the canopy vary from 2.5 to 4°C, while the above
canopy maximum temperature decreased from 1.5 to 3.5°C (Miranda et al., 1994). When
measuring leaf temperature in the upper canopy layer of a non-shaded cacao tree on a sunny
day, Valle et al. (1987) verified that it was always 2°C above the leaf temperatures of the
middle and lower canopy layers. Leaves of the upper canopy layer showed higher differences
between leaf and air temperature and, therefore, water vapor deficit, in comparison to lower
layers.

Water and Light

Several factors influence leaf dynamics, among them water and light availability, which
largely vary between locations and cropping systems (Zuidema et al., 2005). Additionally,
wind conditions during the dry season reduce cacao yield due to water loss and reduction of
leaf area (Alvim, 1977; Leite et al., 1980). Southeastern Bahia is a region with an average
annual temperature of 23.5°C, annual total rainfall of approximately 1700 mm, no
 Cacao: Ecophysiology of Growth and Production 21

pronounced dry season, and trees are shaded with E. fusca and/or E. poeppigiana. Here the
relative light intensity received by cacao trees cultivated varies between 30 and 100% of full
sunlight and 4 to 10% radiation at soil level. The average leaf area index and the extinction
coefficient are about 3.9 and 0.61, respectively (Miyaji et al., 1997a). Leaf longevity varies
with position inside the canopy and/or incident irradiance, time of emergence and height in
relation to the soil level. According to Miyaji et al. (1997b) for Catongo trees of about 7-yr-
old the leaf duration at heights greater than 220 cm and above 60% full sunlight is 181 d,
which is half the duration of leaves at heights between 0-150 cm and 5 to 20% sunlight.
Factors like soil water deficit and leaf respiration, due to high air temperature and high
radiation can be causes for the intensive leaf fall in the middle and upper canopy layers of the
cacao crown (Alvim and Alvim, 1978).
The cacao yield is strongly related to rainfall in the dry season. It also depends on soil
type and consequently water retention capacity (Zuidema et al., 2005). According to Leite and
Cadima Zevallos (1991), the elevation and residence of the water table lead to increase and
maintenance of humidity levels above the subsurface of the soil, a zone in which the cacao
root system concentrates in traditional cacao cropping areas. To reach high levels of cacao
yield in south Bahia, the monthly rainfall totals should be equal or greater than 200 mm,
while the permanence of the water table should be more than 30 d (Leite and Valle, 2000).
The fluctuation of the water table can be considered a climatic variable (derived from rainfall)
that can influence cacao production as a form of natural sub-irrigation (Leite and Valle,
2000).
Using a simulation model, Zuidema et al (2005) showed that annual solar radiation and
rainfall during the dry season explained 70% of the variations in annual seed yields obtained
for 30 sites along the tropical cacao producing areas of the world. Drought promotes
decreases in leaf area (Orchard and Saltos, 1988), A (Hutcheon, 1977), flowering (Sale, 1970)
and yield (Khan et al., 1988). Changes in carbon distribution and export will depend on the
severity and duration of the stress, as well as its occurrence in relation to flowering, pod set
and pod maturation (Deng et al., 1989). There are positive correlations between yield and
rainfall during months that precede harvest (Alvim, 1978), where rain distribution is more
important than its magnitude (Atanda, 1972). Alvim (1988) considered that harvest can be
influenced by the rain distribution that occurred six months prior. This relationship seems
valid for countries such as Brazil (southeastern Bahia), Ghana, Ivory Costa and Malaysia. In
Bahia, the agricultural cacao calendar year begins in April and extends to March. The yield is
divided into two distinct periods, minor harvest (temporão) from April to August and main
harvest (safra) from September to March. The relative importance of each harvest depends on
rain distribution (Alvim, 1973). Furthermore, cacao pods take approximately from six to
seven months to complete their development (Almeida and Valle, 1995). This explains the
importance of higher rainfall and longer phreatic residence of the water table at the soil
subsurface in one year for high production in the following year (Leite and Valle, 1990; Dias
and Kageyama, 1997).
According to Valle et al. (1987), data modeled for transpiration showed that, on a cloudy
day, a non-shaded cacao tree could transpire about 45 L (1.2 L m -2 d-1). This assumes that all
leaves are transpiring at the same rate. A shaded cacao tree, in contrast, could transpire about
26 L (0.8 L m-2 d-1). However, on a sunny day both trees showed a substantial difference in
transpiration when compared with a cloudy day. The daily E was estimated to be 90 L tree -1
(2.4 L m-2 d-1) for the unshaded tree and 40 L tree -1 (1.2 L m-2 d-1) for the shaded one when
22 Alex-Alan F. de Almeida and Raúl R. Valle

100% of its leaf area was exposed to incident irradiance (Valle et al., 1987). Estimates of
daily E on a cloudy day for both tree conditions and for the shaded canopy layers on a sunny
day were of the same order of magnitude of those calculated by Cadima and Alvim (1983).
However, for unshaded trees, on a sunny day, Valle et al. (1987) showed an average E more
than twice the estimated by Cadima and Alvim (1983). From the estimations of LAI (Miyaji
et al., 1997a) and radiation (Valle et al., 1987), it is concluded that more than 60% of the
water can be transpired by the upper canopy layers of unshaded cacao trees. In the shaded
trees less than 5% of the absorbed water might be transpired by the upper layers, which is
caused by the greater irradiance received in the lower layers than in the top layers due to the
shade tree that limits light penetration to the upper layer even at lower solar angles.

Shade vs Unshaded Cropping

In its natural habitat wild cacao grows under shade trees, but with a low pod production
(Murray and Nichols, 1966), since the shade trees compete with cacao for water, nutrients and
light (Bonaparte, 1975). In fact, many people believe that shade is indispensable for cacao
cultivation (Owusu, 1978) and, therefore, earlier cacao plantations were conducted under
shade regardless of yield (Urquhart, 1961).
Evans and Murray (1953) in Trinidad, Alvim (1959) in Costa Rica and Cunningham and
Lamb (1959) in Ghana were the pioneers to critically examine the role of shade and light in
cacao cultivation. In a shade and manurial experiment in Ghana strong reductions in yield
were observed due to shading (Ahenkorah et al., 1974). In fact, cacao grown at full sunlight
produced more pods than when grown under shade (Owusu, 1978). Control of both growth
and density of cacao trees, and the increase in aeration and light penetration throughout the
canopy are fundamental cultural practices necessary to promote increased seed production
(Vernon and Sunderam, 1972). More recently the comprehensive work of Ahenkorah et al.
(1987) further elucidates the importance of appropriate fertilization in unshaded cacao
plantations.
Studies on light interception and photosynthesis show that seed production is light
limited, probably due to external and internal shade (self-shading) in the cacao stand and light
extinction inside the canopy (Yapp and Hadley, 1994). In fact, light intensities below 1800 h
yr-1 suppress flower production with a significant negative effect on pod yield (Asomaning et
al., 1971). Heavy shade not only reduces seed yield (Zuidema et al., 2005), because of low
photosynthate production (Alvim, 1977; Ng, 1982), but also increases the incidence of
diseases (Alvim, 1977). On the other hand, cacao is a shade tolerant species (Guers, 1985), in
which appropriate shading could lead to adequate photosynthetic rates, growth and seed yield.
Shading also alleviates effects of unfavorable ecological factors, such as wind velocity and
excessive evapotranspiration (Miyaji et al., 1997a; Beer et al., 1998) and, consequently,
decreases in both humidity stresses and extreme air and soil temperatures during the dry
season are to be expected (Anim-Kwapong, 2003). This is essential for survival and
establishment of cacao seedlings in dry and seasonally humid environments (Beer, 1987),
since the seedlings are highly susceptible to dry periods (or soil moisture stress) (Leite et al.,
1980). Furthermore, shelter trees used for shade greatly contribute to the formation of soil
organic matter, carbon sequestration, nutrient recycling and maintenance of biodiversity
(Lobão et al., 2007; Müller and Gama Rodrigues, 2007). Specifically in regions with low
 Cacao: Ecophysiology of Growth and Production 23

access to inorganic fertilizers, the multistrata plantation is used to maintain soil fertility with
the subsequent increase in nutrient availability for cacao (Isaac et al., 2007).
When nutrient availability is not a growth-limiting factor, there is a positive correlation
between cacao yields and light (Bonaparte, 1975; Ahenkorah et al., 1987). Vernon (1967)
concluded that the relationship between cacao yield and available light was approximately
linear from 30 to 60% full sunlight. However, when modeled from 0 to 100%, a quadratic
model showed better adjustment than the simple linear model, suggesting that some degree of
shading is desirable.
The higher production of non-shaded cacao results in a shorter productive lifespan, a
larger demand for fertilizers (Owusu, 1978), a decrease in the incidence of cherelle wilt
(Asomaning et al., 1971) and larger investments (Ahenkorah et al., 1987). The subsequent
decline in productivity (Bonaparte, 1973) is attributed to high losses of exchangeable bases in
the soil, attack of insects and diseases, excessive leaf transpiration and increase in soil
evaporation (Ahenkorah et al., 1974). The inherent high risk of unshaded cacao cultivation
was illustrated by the economic analyses of Cunningham (1963), the extra expenditure and
work associated with clear-felling and growing unshaded cacao with great amounts of
fertilizers would probably be justifiable only if yields of more than three tons are obtained.
However, most of the cacao experiments involving shade and fertilizers show that shade
reduces the responses to fertilizer applications (Alvim, 1977), and such conditions are rarely
economically justifiable (Beer, 1987). This shows that cacao production is positively related
to light intensity (Vernon, 1967), but this relationship depends primarily on the availability of
nutrients in the soil (Figure 6).

Figure 6. Cacao grown under different shade and fertilizer regimens in Ghana. From Ahenkorah et al.
(1974) with permission of the American Society for Horticultural Science.

In commercial plantations it is hard to understand and detail the potential productivity in

relation to pod-set and yield, since the cacao canopies are under several shade levels (Alvim,
1977). With shade management practices cacao production can increase during the first
decade and thereafter stabilizes (Rosenberg and Marcotte, 2005). Cacao trees can produce
24 Alex-Alan F. de Almeida and Raúl R. Valle

during 30 to 80 yr (Stevenson, 1987); however, yield subsequently decreases (Rosenberg and

Marcotte, 2005). Furthermore, it is probable that the physiology of trees under heavy shade
would be modified, because shaded trees can be more efficient in photosynthesis and leaf
dynamics (rates of production and abscission), as, for example, the increase in leaf lifespan
(Miyaji et al., 1997a). According to Owusu (1978), there is a close relationship between these
physiological processes and carbohydrates in cacao. Improved growth, increase of flushing,
enhanced flowering and increased yield under unshaded conditions can, however, be
explained in terms of increase in carbohydrate production (Asomaning et al., 1971). This is
exemplified by the internal competition between flowers and pods (Alvim, 1954), in spite of
the small total energy expenditure for flowering (Valle et al., 1990). Furthermore, Alvim
(1977) considers that besides the competition within pods, the annual variation of pod set is
also subject to competition with other vegetative events, such as flushing and root and
cambium growth.
CONCLUSION
The expectation for the next 10 yr, without considering India and China, is a consumption
increase of nearly one millions tons of cacao seeds. The big question is from where would the
extra demand come? A strong possibility is an increase in productivity in West African
countries, but this would be at the expense of significant resource inputs and occur over a
long period. Actually, the chocolate industry foresees difficulties regarding future supplies.
Today the world faces a new situation with the search for new renewable energy sources. As
an example, Malaysia has replaced cacao plantations with oil palm, and Indonesia has
become less and less interested in producing cacao.
According to the World Cocoa Foundation (www.worldcocoafoundation.org) there is a
tendency towards an increase in the consumption of chocolate with high cacao content (43%)
in Europe and USA. Also, there is a world trend in the proliferation of small companies that
process cacao and make chocolate. Their search for cacao is remarkable. There is also a
consensus that in countries with a better level of development, such as Brazil, the future will
be the investment in quality and production of differentiated cacao. Therefore, the efforts in
Bahia to produce fine cacao is in reality a world tendency for other locations with similar
environments, thus, constituting a great opportunity for expansion in this market.
Furthermore, the markets under greatest expansion are those of the emerging nations.
Therefore, expectations are strong for the inclusion of China and India in the chocolate
consumption market. However, if the markets show a promising tendency, more and more of
the cacao farmers will feel the burden of high costs of resource inputs, mainly for control of
diseases and insects that are steadily increasing. Several pests and diseases attack cacao, with
some estimates putting losses as high as 30 to 40% of global production
(www.icco.org/about/pest.aspx). In this sense farmers seek new lands that offer an escape
from known diseases, and in this respect we may highlight the importance of studying the
behavior of the crop under semi-arid conditions as mentioned earlier. The knowledge of the
effects of factors that affect the cacao growth and development under such conditions should
even make it possible to engineer new practices for the traditional cropping areas.