The cultivated Mauka is a herbaceous plant, perennial, erect when young and decumbent or prostrate and open, to maturity. Propagated from seed, the primary root is axonomorphic, thickened, branched from an early age, with small fine rootlets on its surface that have an absorption function 2223. In contrast, the plants originated from vegetative propagation ‒ basal stems and cuttings ‒ have a fascicle of roots (without primary root differentiation) thickened that emerge from the crown or base of the plant 24. At maturity, regardless of the origin, the roots show abundant lenticels grouped in circular rows.

The thin stems (Ø = 0.5 to 3 cm at the base) are green or purple-green and can reach lengths of 80 to 140 cm, with ramifications of the first, second and third order. Their number per plant varies from 5 to 17, depending on whether the plant originates from seed or vegetative propagation. The leaf nodes are prominent and the internodes up to 16 cm long. Generally, only one of the two axillary buds at each node (opposite leaves) develops while the other atrophies. Just below ground level the base of the stems thicken and form short, bulging and globose corm-like internodes with depressed nodes.

These corms constitute the principal source of propagules for vegetative propagation as typically used by the farmers. At the base of the stems (neck of the plant) a crown is formed from which the roots emerge downwards and the stems upwards. The opposite, cordate-ovate or elliptical leaves are light green or dark green with or without purple pigmentation 25. Before flowering the leaves can reach 12 cm in length and 9 cm in width 12426.

The inflorescences are aggregated cymes with the terminal inflorescence unit a cyme with 3 to 5 densely grouped flowers. The pentamerous and sympetalous flower buds are yellow, brown or purple, 3‒4 mm long while 5‒6 mm in diameter at anthesis. The synsepalous calyx consists of five plicate, green, sepals; they are persistent covering the gynoecium at it grows and at fruit maturity almost completely, i.e. anthocarpous. The perianth is formed by five sympetalous tepals, lilac, white or white coloured with slight violaceous pigmentation. It has three free stamens and anthers with two reniform thecae containing 25‒30 pollen grains each. The style is longer than the stamens and curved at the apex. The stigma is capitate-papillate, and the ovary is superior, monocarpellary, unilocular, central placentation, single ovule.

The flower opens only once, and in conditions of the valley of Cajamarca, the anthesis starts between 05:30 and 06:00 and terminates at midday between 12:00 and 13:00s. Its reproduction type is fundamentally autogamous, with 6‒25% cross-pollination. Cross-pollination is entomophilous, mainly carried out by species of the genera Syrphus Fabricius, 1775, Allograpta Osten Sacken, 1875, Apis Linnaeus, 1758 and a Microlepidoptera of the Cosmopterigidae family 27.

The fruit is an elliptical or ovoid achene, 3 mm long × 1.7 mm Ø; grey, black or brown in colour, wrapped by the fleshy anthocarp of the persistent calyx and covered by glandular gum containing hairs most abundant at the apex. The gum allows the fruit to adhere to any surface when detached from the plant and be transported by animals and humans, i.e. an adaptation to epizoic dispersal. At maturation, it is common to find insects, and even birds or mice trapped between the plants. The achenes are the unit of dispersal. The seeds have white or crystalline endosperm and have epigeal germination. One hundred seeds weigh 0.6‒1.2 g 1242628.

The Mauka plant, a perennial geophyte that – in a similar way to other tuberous species – has evolved in the Andean environment, marked by a wet-temperate season alternating with a dry and cold season. In its natural environment, the abundant reserves stored in its roots and the base of its stems enables it to survive under adverse growth conditions. In addition, the thickened basal stems, i.e. corms, serve as a means of regeneration 2930.

Its cultivation in Bolivia, Ecuador and Peru probably dates from the pre-Inca period. Findings and collections, since 1965, indicate that in the three countries it is cultivated at altitudes of 2300–3500 m a.s.l. 12021313233. In Peru, these altitudes correspond to the Quechua region 33. In this region, the average daily temperature is 13.5 °C, with maximum temperatures of 25 °C and minimums of 5 °C, and the average annual rainfall is 680 mm. However, the area of ​​greatest distribution is the same as that of starchy maize (Zea mays L.) cultivars (above 2200 m a.s.l. and below 3000 m a.s.l.), with which it is frequently associated.

It responds better to soils of frank and sandy-loam texture, with a pH between 6.8 to 7.2, with a good proportion of organic matter (≥ 3%). Cultivated in association with crops such as potatoes (Solanum spp. incl. S. tuberosum L.), maize and vegetables it is competitive. Its foliage grows and expands rapidly, due to its decumbent nature, thus compromising the growth potential other species. Due to this behaviour, it is common for small-scale farmers to plant Mauka along the edge of the main crops 72635. Nevertheless, it does not behave like an invasive plant 363738. It produces better in full sun than in shade, as seen in a study carried out by INIAP (Instituto Nacional de Investigaciones Agropecuarias), Ecuador, which showed that experimental plots under shade trees yielded up to 35% less than plots in full sun 39.

Five distinct landraces (I, II, III, IV and V) have been characterized by Seminario and Valderrama 21; one originating in Southern Peru (Puno) and the rest from northern Peru (Cajamarca and La Libertad) ‒ with one originating from a cross occurring ex-situ. Another landrace was observed by Gendall 19 in Puno, distinguished by the vivid magenta colour of its subterranean stem cortex. In an unpublished thesis by Alvarez-Mamani 33, two landraces were registered in Chullín (La Paz Department, northern Bolivia) ‒ Yuraq Mauk’a (white) and Kellu Mauk’a (yellow). It is not yet known whether these are distinct from those landraces seen in Peru, because the descriptors used are not identical and, since the material is not conserved ex-situ, no direct comparison has yet been made.

The plant has shown high adaptability to environments outside the Andes, and in recent decades has been successfully introduced in the Czech Republic and Belgium 40 and the state of Illinois, USA 1038 and in the UK (Owen Glyn Smith, pers. comm.).

Mauka is, as mentioned above, propagated both by seeds and vegetatively (Figure 1). The easiest method of vegetative propagation is by using the basal corms. These corms are thickened ‒ with abundant reserves and with two buds at each node ‒ and can be planted directly in the field. This is the most common approach used by traditional Andean farmers. In Puno and Cajamarca, farmers usually separate the propagules and leave them out in the sun for a period of time, anywhere between a few days and two weeks, before planting. It has been suggested that this practice mitigates against fungal attack from soil residues and may accelerate sprouting 19. The plant is also propagated by aerial stem cuttings, rooted in a suitable substrate before being transplanted to the field. Whether the stem cuttings originate from the base, the middle section or the terminal section makes no difference to quality or yield. In general, these cuttings are ready to be transplanted to the field following 45‒60 days 28. During the rainy season the direct planting of cuttings in the field, i.e. without previous rooting, can be successful. Franco and Rodríguez 41 reported that cuttings planted directly in the field had a 71% to 97% survival rate.

The advantage of propagation by cuttings is that the multiplication rate is higher than with underground stems. An evaluation carried out using the purple landrace, at seven months of age, found that each mother plant produces 266 cuttings, with four nodes each (eight buds); meanwhile 100 plants are needed to plant one hectare at the density of 0.90 m × 0.50 m, i.e. 22 222 plants ha^-17. However, the results do vary depending on the substrate used and the time of year.

Under greenhouse conditions, Alva 42 studied the effect of three substrates on the propagation of the Mauka by stem cuttings. The substrates were tested using three landraces (I, II and III), in completely randomised block design and with three repetitions. The test lasted for 60 days, at which time the cuttings were ready to be transplanted to the field. The following traits were evaluated: number of cuttings with shoot, number of rooted cuttings with shoot, height of shoot (cm), number of roots per cutting, length of root (cm) and number of leaf pairs per cutting. No significant statistical differences were found for the landrace interaction by substrate type, for the variables studied. The best landrace was light green (III), in which 83% of rooted shoots were obtained and the best substrate was sand with 72% of rooted cuttings. Each shoot produced 12 to 20 leaves (Table 2a–b).Sexual propagation implies regeneration from seed, which is produced in abundance by the plant 743. This seed is non-dormant, orthodox, and retains its viability for several years. Tests with seeds kept in a closed container at 18 °C showed an 88.5% germination rate after three years 28. The characteristics of seed lots of three Mauka landraces and the germination (in Petri dish on paper) at six months post-harvest were similar; all showed high germination rates. Similarly, planting seeds of the same lots in soil, sand and humus (at the ratio 1: 1: 1) showed similar seedling emergence and number of days to first pair of true leaves (Table 3). This uniformity in morphological and physiological characteristics (absence of dormancy, uniform and rapid germination, although the crop has retained the auto-dispersal of their seeds) indicate that these landraces have been strongly affected by the domestication syndrome in that they no longer behave as wild material 44.

A first noticeable difference between plants raised from seed and vegetative propagule (corms or stem cuttings), refers to the origin of the tuberose root, which is the harvestable part. In the first case, the tuberous root is the result of the lengthening and thickening of the radicle and hypocotyl, which form a single tuberous axis (Figure 2). In this aspect, it is different from other roots such as beetroot (Beta vulgaris L.), where the reserve organ corresponds to the hypocotyl only 45. Its thickening due to accumulation of storage reserves (Figure 3) occurs due to the activity of secondary changes that produce vascular tissues and reserve parenchyma that are located successively, in concentric rings, at a rate of approximately one per month 23, similar to what happens in turnip (Brassica rapa L.) and beetroot 4546. In each primary root thus described, four to five secondary roots are equally thickened. At 270 days this set of roots has accumulated 88% of their reserves and 330 days have accumulated the total reserves (Figure 4).

On the other hand, in plants from propagule, the tuberous roots are adventitious and numerous, originating from the corm, i.e. the crown or base of the stems. Whether from seed or propagule, in plants with this type of tuberous roots, i.e. without independent fibrous roots, of which the primary function is to absorb the photosynthates which are destined to the apical meristems of the fibrous roots and the stems are translocated through the tuberous roots. In the same way, both water and mineral nutrients must be transferred through the tuberous root to the leaves 46.

The Cajamarca field experiments were carried out at the National University of Cajamarca ^UNC, located at 7° 29'45'' S and 78° 10'12'' W, at 2 656 m a.s.l. with an average daily temperature of 14 °C, 65% of RH and an annual rainfall of 650 mm. These experiments involved the comparison of three landraces, raised from both seeds and propagules (Table 4). The results indicated that these plants are similar with respect to growth period, maximum leaf area index and harvest index. However, they differ in terms of the type and linear measurements of the tuberous roots, number and weight of tuberous roots, number of stems, maximum leaf area, crown weight and root yield. The lower yield exhibited by plants raised from seed is notable, and might be explained by the lower number of stems and lower leaf area. These characteristics indicate that Mauka propagated from seed should be planted at higher densities than plants propagated vegetatively, in order to maximize yield per ha^-1.

Weeding is necessary during the first months following planting. The frequency of weed removal depends on the conditions during the cropping season, i.e. wet or dry, and the conditions at the site. Removing weeds as they appear, before they become competitive to the crop, is generally recommended. After four to five months the Mauka plants should cover the furrows between ridges and outcompete the weeds on their own accord.

Observing Mauka’s growth habits under various experimental planting arrangements since the late 1980s (again conducted at the UNC experimental fields) indicates that hilling, i.e. the placement of additional soil around the base of each plant several months into its growth cycle, is not of fundamental importance for the production of Mauka roots, although local farmers have reported that the practice is beneficial to cultivation 19. Because the roots develop in a vertical or slightly inclined direction – as seen other Andean root crops, e.g. maca (Lepidium meyenii Walp.), yacon (Smallanthussonchifolius^Poepp. H.Rob.) and arracacha (Arracaciaxanthorrhiza Bancr.) ‒ it appears not to influence how the tuberous roots originate and grow around the crown or main stem. In this sense, they differ from crops classified as tubers, such as oca (Oxalis tuberosa Molina) or mashua (Tropaeolum tuberosum Ruíz & Pav.), where hilling is an indispensable task because it ensures that the stolons develops tubers, as opposed to just producing aerial stems.

Hilling up soil around the base of the Mauka plants instead appears to stimulate the growth in number and thickness of the underground parts, at the expense of the growth and thickening of the roots themselves ‒ which are the principal product. This effect is seen in the first drawing of a Mauka plant cultivated in Bolivia, published by Rea and León 1, which illustrates both the underground corms and a single tuberous root (Figure 5). Hilling is therefore recommended for maximizing the production of underground vegetative propagules, i.e. corms, and to protect plants during the cold season, but it is not advised if root production is the priority (Figure 5). As already indicated, in places with abundant rainfall and flat land good results are obtained by planting on ridges, raised 20–30 cm.

Mauka is a crop grown by smallholder farmers at high altitudes in the Andes, where its cultivation is subject to and synchronised with the rainy season, i.e. September to April. As such it tends to be irrigated only occasionally, as and when the main crops are irrigated in those areas that have access to water for this purpose.

Mauka's water consumption behavior has not yet been compared across different climatic and soil conditions. Further inquiry into these aspects would be highly relevant for the development of the crop.

With regards to stress, the greatest risk to Mauka plants is probably excess water, which can cause root rot. This can occur as a result of abundant rainfall combined with poorly drained soils 710. Albeit, Kritzer et al. 38 observed that in field trials conducted in southern Illinois the two landraces tested did respond differently. Local farmers describe Mauka as a naturally healthy plant which tends not to succumb to pests and diseases 18. A field trial at the UNC ‒ where four Mauka landraces were planted in monoculture plots of 2 000 m2 ‒ found that, whilst stem-boring and leaf feeding insects were observable on leaves and shoots, they did not damage the plants in a way that was economically significant 7. The Mauka mosaic virus (Mirabilis potyvirus^Mir), transmitted by aphids whose indicator plant is quinoa (Chenopodium quinoa Willd.), has been identified by Lizárraga 48. However, only one case of probable attack by this virus has ever been recorded ‒ during the agricultural season 1994–1995. This manifested as chlorotic stains on the leaves, but only in a few plants of the Mauka germplasm collection maintained ex-situ at the UNC 7. Mauka is a host to nematodes of the genera Pratylenchus and Paratylenchus. In pre and post-culture evaluations, an index of population growth greater than 1 was found 39.

Low temperatures can also induce stress in Mauka plants. In the Cajamarca Valley, where temperatures fall below -2 °C, frost has been known to damage the entire leaf area. However, sustained by its underground reserve and abundance of buds originating from the underground corms ‒ with the addition of weekly irrigation ‒ it can typically recover within 30 days.

Experimental studies in several locations in several locations of Peru, Ecuador and Bolivia (Table 9) indicate that the production potential of tuberous roots is high, but at the same time highly variable (12 to 137 t ha^-1) due to the influence of management and environmental conditions. From the data reported, it can be inferred that yields of 30 to 58 t ha^-1 are typical for the Andean region.

However, the observations of five Peruvian landraces over several plantings and years (Table 10) show important differences in the yield components and the yield of roots and foliage (stems plus leaves), which implies the possibility of selecting materials for specific purposes and the need to test agronomic treatments for crop improvement.

In addition to tuberous roots, the Mauka plant also yields leaves and stems (Figure 8), which are traditionally used as feed for both smaller animals, e.g. guinea pigs and rabbits, as well as larger livestock, e.g. cattle and sheep ‒ however, the practice is not universal amongst farmers. 9262958. Some identify Mauka as an important forage plant, and several farmers have noted that grazing livestock prefer Mauka to other plants. In Corongo (Ancash), the practice of cutting green material from Mirabilis prostrata (Ruíz & Pav.) Heimerl Mauka’s wild relative (proliferating at the border of a field) ‒ as fodder for pigs, has been observed. In Chachapoyas, where the practice of cultivating Mauka now seems to have disappeared, one farmer recounted the historical use of Mauka roots as pig feed 19.

An evaluation carried out over three agricultural seasons, which assessed production levels for 40 different accessions of Mauka (5 plants per accession), produced forage yields which averaged 1.6 ± 0.6 kg plant^-120 - which is equivalent to 33.3 tons of green fodder ha^-1.

In order to assess forage production, Bazan et al. 9 planted a plot of 557 m2 with Mauka plants (purple landrace), with 0.8 m spacing between rows and 0.5 m between plants.. The yield per plant (Table 11) corresponds to 76 t ha^-1 of green fodder with a dry matter percentage of 22%. This yield is superior to the yield obtained both in experimental trials with lucerne (Medicago sativa L.) 5960 and Napier grass (Pennisetum purpureum Schumach.) from a Taiwanese pasture 61.

Using these studies as a starting point, further research on forage production from Mauka can be carried out compared to the production levels of other forage species, as suggested by Kritzer Van Zant 1038, taking into consideration the fact that yields will inevitably vary according to crop husbandry, environmental conditions, the time of harvest and ‒ most likely also ‒ landrace.

The roots, dense clustered with low water content, are resistant to handling. Post-harvest they are seldom attacked by pathogens, and their tissues show no changes of the kind that might inhibit consumption, as with other roots, such as arracacha (Arracaciaxanthorrhiza), yuca/manioc (Manihot esculenta) and yacón (Smallanthussonchifolius). Fresh Mauka roots are naturally astringent; a characteristic which is more pronounced in certain landraces. . This is traditionally removed in two ways: (a) by changing the water once or twice during cooking, and (b) by exposing the roots to the sun for a period of 2‒10 days ‒ a practice known as ‘soleando’, which also sweetens the roots by increasing the concentration of sugars 119.

The roots can be kept outdoors for several weeks without deterioration, and, if they are left to dehydrate in a dry environment, can remain unchanged for many years (Figure 9, left). They can also be preserved using the parboiling technique, following the procedure used to precook arracacha 62 ‒ by washing the root, cooking them for five minutes, slicing, sun drying and packing them (Figure 9, right, top). Another way is through natural drying, used for yuca/manioc, which entails washing and cutting the raw roots, drying them in the sun, and packaging 63. In a modification of this last technique the roots are cut into thin slices ​​ (Figure 9, right, bottom).

Both pre-cooked and raw products can be used in savoury and sweet preparations, or they can be ground to obtain flour or starch for various uses.

Farmers wash the roots and peel them either prior to cooking by scraping the thin skin off with a knife, or after cooking ‒ usually by hand. They typically consume the roots boiled as an accompaniment to other food, in the same way that yucca/manioc is eaten Mauka is also served unaccompanied; boiled and washed down with coffee for breakfast or lunch, in place of bread or potatoes. Other dishes made with Mauka mentioned by farmers include ‘mazamorra’ (a sweet pudding), ‘picante de cuy’ (spicy stew made with guinea pig), soups and broths, and ‘pachamanca’ or ‘huatia’ (an earthen oven which entails burying the roots in the ground with hot coals) 19.

The International Potato Center (CIP) and the Swiss Agency for Development and Cooperation (COSUDE) are gratefully acknowledged for funding research via the Andean Roots and Tubers Biodiversity Program on Mauka and other tuber crops during the decade from 1993 to 2003. Thanks to the Universidad Técnica del Norte-Ecuador for the space given for the implementation of conservation fields of miso at the Pradera Biological Knowledge Center. This will allow carrying out future research related to the crop. We also give warm thanks to the farmers who have shared their knowledge with the authors and other researchers over the years.