This topic together with others was written for and published in the Espresso Coffee – Book edited by Andrea Illy, Rinanto-nio Viani & Furio Suggi Liverani (1995, 2005 Elsevier).

Now, the content of the entire book has been made freely accessible by the Trieste Coffee Cluster under MediaWiki called Coffeepedia.

For those who may take and distribute Coffeepedia contents it will make sense to (a) cite the source, (b) provide the authors' affiliations, and (c) contact the copyright owner of illustrations and photos in order – in spite of everything what already happened – not to offend against taste, decency and law. Therefore, see here soon the affiliations and contact details of the authors.

The time period between flower opening (anthesis) and the fully ripe fruit is species-specific and varies considerably among the coffee species. It depends further on the genotype and on climatic and cultural conditions.

The economically important species C. arabica and C. canephora require 6–8 and 9–11 months for maturation, respectively (Guerreiro Filho, 1992).

As might be expected, all flower organs contain the purine alkaloid caffeine, with highest concentrations in the stamens. Amazingly, the latter accumulate, besides traces of theobromine, easily detectable amounts of theophylline, indicating an alternative biosynthetic pathway in the male part of the flower with theophylline as the direct precursor of caffeine. In leaves and seeds, caffeine biosynthesis was shown to proceed via theobromine. In analogy to citrus plants, where the highest caffeine concentration has been found in the protein-rich pollen (Kretschmar and Baumann, 1999), we may assume a preferential PuA allocation also to coffee pollen grains. However, the related analyses have not yet been done. Bees are, in contrast to many insects, not only amazingly tolerant against caffeine and other phytochemicals (Detzel and Wink, 1993), but rather, after caffeine uptake, have an improved performance such as a boost in ovipostion by the young queen, an enhanced activity of the bees outside the hive, and an improved defence by bees against hornets at the hive entrance (reviewed in Kretschmar and Baumann, 1999).

When the blossom falls from the coffee tree the persisting ovary develops into the young green coffee fruit (Figure 3 and Figure 4).

Fruits always signify high investment costs and, therefore, to defend them against predators the coffee plant pursues several linked strategies.

First, the very young and green fruits are not showy but rather inapparently arranged in clusters in the leaf axil (Figure 3). Secondly, both chlorogenic acids and purine alkaloids are highly concentrated, and, thirdly, the development of the real endosperm is postponed until mechanical protection works. This last point is a most remarkable feature of the coffee fruit development. Within 3–4 months after anthesis the still green fruit reaches a size suggesting readiness for maturation. When cut across, two greenish beans, already typically rolled up, can be recognized.

However, the appearances are deceptive: the fruit is far from being mature, the (generally) two beans are false perisperm beans made up of mother tissue (Carvalho et al., 1969). At the adaxial pole (towards the fruit stalk) of each bean one can see the beginnings of endosperm development: a whitish, very soft tissue (also called liquid endosperm) starts to invade into and resorb the perisperm bean. Recent studies show that perisperm metabolites such as sugars and organic acids are most likely acquired by the endosperm (Rogers et al., 1999b). We may assume that this process is similar to the invasion of the cotyledons into the endosperm during germination described above: the metabolites shift from one tissue to the other, whereby they have to pass through the so-called apoplast, i.e. the extracellular space between peri and endosperm during seed development, and between endosperm and cotyledons during germination. However, the endosperm is more than a simple blot of the perisperm, since it owns high biosynthetic activities. In conclusion, and philosophically speaking, in coffee the way to the next generation is characterized by transitions in which the metabolites are shuffled around twice.

During this invasion the inner layer (endocarp) of the fruit wall (pericarp) noticeably and increasingly solidifies and later results in the parch layer described above. The mechanical defence of the endosperm itself is remarkably increased by the formation of thick cell walls containing, besides cellulose, the so-called hemicelluloses, i.e. arabinogalactan and galactomannan (Bradbury, 2001).

Hemicelluloses are highly complex polysaccharides primarily renowned for giving an amazing degree of hardness to palm seeds (cf. date, Phoenix dactylifera; vegetable ivory, Phytelephas macrocarpa). It remains to be mentioned that the coffee perisperm finally atrophies into the thin seed coat, the silverskin, that falls off during roasting. Very soon after anthesis the pericarp contains an absolute amount of caffeine kept unchanged until ripeness. However, the initially high (>2%) caffeine concentration drops by dilution to around 0.2% during the further growth and maturation processes, culminating in the transformation of the fruit wall (pericarp) into three distinct layers which serve for fruit dispersal: the tough endocarp protects the seed from digesting enzyme activities in the gut of the frugivores such as birds or mammals; the fleshy, sugarcontaining (Urbaneja et al., 1996) middle layer (mesocarp) softened by enzymes (Golden et al., 1993) acts as a reward, while the vivid coloration by anthocyanins (Barboza and Ramirez-Martinez, 1991) of the outermost layer (exocarp) is to attract the dispersing animal.

We should not close this section without relating our thoughts about biochemical ecology to a practical question of our daily life: how does the espresso bean get its caffeine? Though numerous publications on caffeine biosynthesis exist (for a comprehensive review see Ashihara and Crozier, 1999), this problem has never been addressed and therefore we can only speculate about it. Clearly, the endosperm has a certain biosynthetic capacity for caffeine. But is this all? Are there contributions of other sources? The perisperm provides around one-third of the seed caffeine as estimated from the caffeine content of the perisperm bean (see Figure 10.5 in Sondahl and Baumann, 2001). The leaves are not directly contributing to seed caffeine, but the pericarp may be a valuable source, as studies with labelled caffeine have shown (Keller et al., 1972; Sondahl and Baumann, 2001).

Obviously, caffeine migrates from the fruit wall into the developing seed, most likely due to the high concentration of chlorogenic acids allocated to the perisperm/endosperm. Unfortunately, the extent of this caffeine transport is unknown. Conceivably, this fraction depends on both the fruit developmental time and the chlorogenic acids allocations, and is correspondingly larger in a slowripening species with a high ratio of seed to pericarp chlorogenic acids.

Again, synthesis, transport and accumulation of chlorogenic acids eventually determine where and how much caffeine is to be allocated in the seed. In conclusion, perisperm and pericarp are certainly important sources of the seed caffeine, whereas the leaves, the pericarp and perhaps also the greenish perisperm may provide most of the chlorogenic acids crucial to gather and firmly fix the caffeine to the coffee bean. However, the degree of contribution from each side (maternal tissues versus endosperm and embryo) is not yet known. Additional studies on the developmental biology of the coffee seed (Marraccini et al., 2001a, 2001b) as well as reciprocal crosses between coffee species differing in their caffeine and chlorogenic acids content will cast some light into the espresso’s darkness!

Coffee plant (C. arabica) flowering and fruiting. (©Beatrice Häsler). The drawing shows a side branch with flowers at various stages: wilting and falling off, in fresh bloom, or buds (from the base to the apex). Above, flowers are illustrated in detail and enlarged. The yellow stamens are inserted in the throat of the corolla. Similarly, various fruit stages are shown: like the flowers they are arranged in composite clusters at the leaf axil.

from flower to fruit