«Andean roots and tubers: Ahipa, arracacha, maca and yacon M. Hermann and J. Heller, editors Promoting the conservation and use of underutilized and ...»
Accessions JJV06 (A, B, C, D, F) and ECU1154 (E, G, H).
106 Arracacha (Arracacia xanthorrhiza Bancroft)
4.3 Reproductive biology
4.3.1 Flower induction Whereas flowering arracacha can be observed only occasionally in the Andes (Bukasov 1930; Hodge 1954; Higuita 1970; Bristol 1988), flowering is frequent in Brazil, especially in the southern states of Minas Gerais, São Paulo and Paraná over 900 m altitude (Zanin and Casali 1984b). It is important to remember that arracacha production in this region is south of the 20th parallel, that is, south of the latitudinal range of Andean arracacha production. Seasonal variations in temperature and daylength in southern Brazil are therefore much more pronounced, and have been related to the seasonal pattern of arracacha flowering. Zanin and Casali (1984b) proposed that the low temperatures and/or short days at midyear induce flowering in commercial plantings from July to October. (Higuita (1970) also briefly mentions "low temperatures" as the cause for occasional flowering in Colombia.) To test this hypothesis, Bajaña (1994) conducted a two-factorial greenhouse trial varying night temperatures (5-8°C vs. 12-15°C) and photoperiod (10 hours vs. 15 hours). Of the 10 genotypes used, only three flowered. Flowering responses were in general weak (only 15 out of a total of 160 plants). The treatments with short days and low night temperatures did not have statistically significant effects on flowering. In a collateral experiment, the same author removed the storage roots from mature plants and left the crowns to dry until they lost about half of their weight. Eighty percent of the replanted crowns flowered. Recent experiments with the same clone (from San José de Minas, Ecuador) showed that dehydration of mature plant crowns (20-30% weight loss) and subsequent culturing of the crowns induces, in over 90% of plants tested, the formation of nearly as many inflorescences as there are cormels in a plant (Hermann, 1996, unpublished results; see Fig. 6). These results confirm observations of Dr F.F.
Santos in Brazil (1994, pers. comm.), who has repeatedly seen vigorous flowering responses in arracacha plantations that were subjected to drought, such as in abandoned fields in Goiás. Farmers in Ecuador also report increased flowering after spells of dry weather. Since drought (in Brazil) is often associated with short photoperiods and low temperatures, the effects of these factors cannot be separated. On the basis of present evidence, however, it is dehydration rather than vernalization that induces flowering in arracacha. In this context, it is perhaps noteworthy that the recently found putative ancestor of arracacha flowers after extended periods of drought (in the Ecuadorean Andes; see Section 184.108.40.206).
By contrast, flowering in Old World apioid cultigens such as celery, carrots and parsley is induced by vernalization.
Flowering also seems to depend on genotype, as shown by Bajaña (1994) and observed by Plasencia and Huertas (1986) and in field collections in Brasilia by Dr F.F. Santos (1994, pers. comm.). The preliminary conclusion from these observations is that flowering is more easily induced in Ecuadorean accessions than in Peruvian and Brazilian cultivars.
Promoting the conservation and use of underutilized and neglected crops. 21.
4.3.2 Breeding system As described in Section 4.2.2, the styles of the perfect flowers in a given umbel become receptive several days before the first stamens shed pollen. This phenomenon, known as protogyny, hinders self-fertilization and thus promotes outcrossing. Seed progenies of arracacha have been found to segregate considerably (Dr V. Casali, 1991, pers. comm.), which suggests a high degree of heterozygosity of arracacha cultivars and is consistent with the putative outbreeding nature of the species. The fact that spontaneous seed set can also be observed where only one clone is grown, as in Brazil, or in commercial plantations elsewhere, indicates that arracacha is sexually self-compatible, or at least is so for certain cultivars. In light of this preliminary evidence, arracacha appears most likely to be a facultative outbreeder.
4.3.3 Seed formation, storage and germination The 'seed' of arracacha is shown in Figure 10. It is an achene or mericarp, a dry oneseeded fruit resulting from a schizocarp. At maturity, which is reached 8-10 weeks after pollination, the fruits are not shed but remain connected to a carpophore. Seed set is impaired or absent in sites that have comparatively high temperatures (Tacna, Peru; Brasilia) (Hermann, unpublished field observations). This is possibly due to the heat-induced internal breakdown of anther tissue which prevents the thecae from dehiscing and releasing pollen. During periods of daily peak temperatures above Fig. 10. Mature arracacha fruits (accession AMM5161). Length: 8-10 mm.
108 Arracacha (Arracacia xanthorrhiza Bancroft) 35°C and less than 30% relative humidity (as observed in Quito greenhouses), the anthers shrivel before they can shed pollen.
The seed of arracacha and other Arracacia species is orthodox, meaning that it can be dried to a low moisture content, which allows storage at temperatures below freezing point. Germination of recently harvested seed is rarely higher than 30%. The effects of plant origin, vernalization and seed treatment with chemicals and fungicides on germination have been studied to some extent, but no conclusive treatment to enhance germination has been identified (Sediyama 1988; Sediyama et al. 1990a, 1990b, 1991a).
Larger seed, though, has significantly higher germination (Sediyama et al. 1991b).
4.4 Plant propagation Plant propagation is the single most important issue in growing arracacha profitably.
First, this is because the propagules can be produced on-farm, year after year, without degeneration of stocks. This, combined with robustness and nutrient efficiency, makes arracacha quite attractive to small farmers, who do not need to obtain credit to buy seed. Second, as will be seen below, root productivity depends greatly on the preparation of the propagule.
As outlined in Section 4.2.1, the cormel is used traditionally, and exclusively, as the propagule. Depending on age and development, the cormel has a few to several dozen buds, each of which has the capacity to sprout and form a new shoot, which Fig. 11. Arracacha propagules. Propagules in lower row develop into more productive plants (with higher harvest index).
Promoting the conservation and use of underutilized and neglected crops. 21. 109 will swell at its base into a new cormel. A large or entire cormel, if used as the propagule, will therefore grow into a plant with many shoots (as in Fig. 3), whereas a propagule consisting of only the apical part of a cormel will result in a plant with fewer shoots, less foliage, smaller aerial parts and a higher proportion of total dry matter being allocated to the economic product, the storage roots (Casali et al. 1984).
This relationship, which is rarely appreciated in the literature, is of the utmost importance for growing arracacha successfully.
Arracacha propagules can be taken at any stage of plant development. Senescent crowns from a mature crop are not really dormant (unlike those of Arracacia andina; see Section 6.2.1) and cormels taken from them will immediately root and sprout given appropriate conditions of temperature and moisture. To prepare the propagule, the cormel-offshoot is detached from the rootstock, and its leaves are trimmed back to leave only a few centimetres length of the petiole. Then the proximal two-thirds or threequarters of the cormel are cut off in a slant cut, as seen in the upper row of Fig. 11. In the Andes, it is customary to cut a cross into the surface of the cut. This is believed to result in better spacing and a more equal lateral distribution of the storage roots. Before the propagules are planted, they are left for a few days to permit the cut surface to dry.
The more observant farmers are acutely aware of the importance of a good propagule with a minimum of eyes, yet one that has sufficient reserves to support post-planting dehydration and stress. As can be seen in Fig. 12, plants resulting from
the right propagule can be tremendously productive. Such plants have as few as 10 shoots but no more than about 20, whereas more than 40 shoots (and cormels) are normally found on unproductive plants with 'hypertrophic' rootstocks. Table 4 provides data on dry matter partitioning of arracacha in three locations in highland Ecuador. Ironically, it is the small farmers of San José de Minas without access to 'technology' and 'advice' from research institutes who raise the most productive plants, with a harvest index of 82%. This is most remarkable for a root crop, especially for one that has never been 'improved' by plant breeding.
Table 4 also shows the extreme variation in dry matter partitioning of arracacha that can occur in different sites under different practices. This aspect must be observed in germplasm evaluation trials to ensure that genetic differences in plant architecture are measured and not phenomena that result from poor propagation techniques.
The multiplication rate of arracacha, that is, the number of propagules obtained from a mature plant through the conventional field method, depends on the number of cormels. In the Brazilian clone, the number of cormels per plant varies between 15 and 40 (Zanin 1985). Briceño (1977), evaluating 10 Bolivian, 12 Colombian and 4 Ecuadorean accessions, found the multiplication rates of this diverse material to
range between 9 and 48 after one year of crop duration (mean and standard deviation:
27±11; calculated from data in Briceño 1977). This rate, however, can be greatly increased by common horticultural methods such as bud cuttings, removal and rooting of sprouts from the planted cormel and other methods.
Table 4. Yield and relative dry matter partitioning (DMP) of mature arracacha plants in three locations in the Equatorial Andes
Conditions of cultivation: San José de Minas: from commercial production by small farmers at 1960 m asl, crop 14 months old (source: Hermann, unpublished data); Tumbaco: germplasm field
collection (accessions ECU1161, ECU1179, ECU1181), 2400 m asl, crop 20 months old (source:
Hermann, unpublished data); Cumbayá Nestlé experimental station, 2500 m asl, crop 10 months old, derived from data in Raffauf and lzquierdo 1994; all localities near equator. (Example: "DMP rootstock (%)" is the fraction of total plant dry matter accounted for by the rootstock.) n.d.= no data.
Promoting the conservation and use of underutilized and neglected crops. 21.
Tissue culture of arracacha from meristems or protoplasts has been pioneered by Brazilian authors (Reis et al. 1989; Pessôa and Esquibel 1991; Pessôa et al. 1991a, 1991b,
1994) but it is of no commercial relevance to date. Recent research (Landázuri 1996) identified promising culture media for shoot-tip culture and micropropagation. A protocol is now available to propagate the plant in vitro at a rate of six to one per 8week cycle by using an MS medium supplemented with 3% sucrose, 56 ppm BAP (6-benzylamino purine) and 0.05 ppm ANA (a-naphthaleneacetic acid).
4.5 Crop husbandry
4.5.1 Planting Generally, the unrooted propagule is planted at the onset of the rainy season. A very promising development in Brazil, promoted by extensionists and increasingly adopted by farmers, is the rooting of propagules and subsequent transplanting to the field. This improves early plant development and, as a classical horticultural technique, brings about a number of benefits, the most significant being more homogeneous plant canopies, reduced crop duration and higher yields (Câmara 1992, 1993). According to farmers in Goiás I have interviewed recently, pre-rooting of propagules takes about 45-50 days. It is conveniently achieved on small plots with improved soil substrate. At planting, transplants can be selected and subsequent crop establishment is rapid. As a consequence, the crop can be harvested 6-7 months after planting compared with a crop duration of 8-10 months for unrooted planting material. Planting densities vary between 15 000 and 30 000 plants/ha (Higuita 1970; Santos et al. 1993). Santos et al. recommend a spacing of 0.8 m between rows and 0.4 m within rows. This seems to be a common practice for growers in Brazil.
Popular belief in Northern Peru has it that during the planting season one should avoid sleeping with crossed legs as this will cause the storage roots to become twisted and result in deformed arracachas at harvest.
The reason for the slow initial plant development of arracacha is unknown, but possibly the early allocation of dry matter to storage structures (cormels and rootstock) causes reduced leaf area replication. Because initial plant growth is slow, arracacha is, in the traditional cropping systems of the Andes, often intercropped with maize, which matures after 5-6 months. Weeding, either mechanical or chemical (in Brazil), is needed during early crop development. Once the arracacha canopy closes, however, arracacha can form high leaf area indices and thus suppress weed growth.
Hilling is traditionally done in the Andes and in parts of Brazil. Farmers believe that this stimulates storage root formation. A recent experiment, however, has shown that hilled plants have reduced harvest indices and lower yields (Raffauf and Izquierdo 1994). Hilling (as well as deep planting) can lead to elongated and hypertrophic rootstocks. Farmers in Brazil are discouraged from continuing this practice because of doubtful benefits and the labour costs involved (Dr F.F. Santos, 1996, pers. comm.). Some caution is perhaps indicated to not prematurely dismiss hilling as inappropriate.
Arracacha (Arracacia xanthorrhiza Bancroft) 4.5.2 Fertilization In the Andes, fertilizers are often not or only sparingly applied to arracacha. Higuita (1970) probably has Andean soils of poor fertility in mind when he recommends for Colombia 500-600 kg of compound fertilizer (N-P-K = 10-30-10 or 12-24-10). It is not clear whether this recommendation is based on experiments, but the suggested quantities do not emphasize nitrogen. In Brazil, arracacha is typically grown from residual nutrients left over from a preceding potato crop (Santos 1993). Santos et al.