Relevance of Banana Seeds in Archaeology

In this contribution, the importance of banana seeds collected during archaeological investigations is indicated. Because fully-formed seeds are not ordinarily produced in bananas cultivated for food, the archaeological relevance of banana seeds may initially appear to be limited. However, there are a number of contexts in which the recovery and identification of seeds can be important for understanding the initial domestication and dispersal of bananas by people. In this respect, the possible existence of naturalized species and/or subspecies is hereby reported. Several innovative hypotheses are advanced based on botanical considerations, which may have profound consequences for the reconstruction of the prehistory of banana domestication and the involved regions, and which archaeology can assist in confirming, modulating or refuting. The “Center of primary Musa diversity” is not a rigid concept The center of primary diversity in the genus Musa is large and very complex (Figure 1). It was the grand achievement of Cheesman to classify and to describe most of the numerous taxa in a series of articles that appeared in Kew Bulletin (Cheesman 1947, 1948, 1949, 1950). He was obliged to coin the botanically uncommon term of “Section” for the highest order of differentiation in the genus. This classification has since been widely adopted with satisfaction and extended mainly by Simmonds through the addition of newly found species (De Langhe et al. 2009). Any banana plant with fully-developed seeds that is found under natural conditions in the center of primary diversity is usually considered to belong to a wild Musa species. However, such an assumption ignores the possibility that seedy banana plants may have been ‘naturalized’ (see De Langhe et al. 2009 for an explanation of the terminology) Published: July 30, 2009 Ethnobotany Research & Applications 272 www.ethnobotanyjournal.org/vol7/i1547-3465-07-271.pdf pathogens, chewing insects and other predators. Various polyphenols in the mesotesta would protect against fungi. An opening in the coat, through which the radicle and hypocotyle of the embryo will emerge at germination, is protected by a hard plug (operculum). It has been shown that the seeds remain unharmed – although with decreased germination capacity – after digestion by animals that consumed the ripe seedy fruits (Simmonds 1959b). Depending upon the species, banana seeds in the soil will normally undergo a relatively long period of dormancy, of which the mechanism is still not well understood. However, it has been estimated that most seeds lose germination capacity after one year (Simmonds 1959b). As a consequence of the above, seeds that remain in various burial environments should persist for an indeterminate time due to the highly resistant coat. It is therefore remarkable that seeds are not readily found in the deeper soil profiles that interest archaeologists. In part, the paucity of archaeobotanical records of banana seeds reflects the lack of systematic archaeobotanical investigation in parts of Island Southeast Asia and especially Melanesia, and the inadequacies of suitable modern reference collections against which archaeological specimens are compared (Tim Denham pers. comm. 2008). Due to more or less advanced stages of female and male infertility, most edible bananas will never produce viable seeds under cultivation or when naturalized. However, when the female flowers are abundantly pollinated by artificial means, some varieties can produce a few mature seeds per bunch, which is how genetic improvement became possible. Therefore, some of these seedy fruits may plausibly contain one or more mature seeds in natural conditions when the plants are in the proximity of wild banana plants. Such cases have indeed been noticed and are even rather common for the ‘Pisang Awak’ subgroup, belonging to the hybrid ABB genome-group (Simmonds 1959a). In conclusion, although the occurrence of seeds from edible bananas in archaeological contexts should be rare, their presence should be considered for older portions of the stratigraphy that correspond to the periods when the first semi-seeded edible diploids developed. Consequently, archaeological research should theoretically observe a decrease in banana seeds upwards through the stratigraphy at sites within the primary Musa diversity center; namely in places where wild and cultiwild plants have progressively been replaced by semi-seedy diploids and later by sterile, seed-suppressed diploid and triploid cultivars.