TRADITIONAL CROP REVISED: YIELD AND QUALITY OF PALM-TREE KALE, GROWN AS A MECHANISED INDUSTRIAL CROP, AS A FUNCTION OF CUTTING HEIGHT

Palm-tree kale is a local crop grown in some areas of the northern Apennines, Italy. The food industry is presently interested in recovering traditional recipes: besides the processing aspects, a critical factor for raw-material production is complete automation. Previous research demonstrated the possibility of growing palm-tree kale at high density, with early simultaneous harvest. Biomass partitioning and quality as a function of cutting height were further investigated. An autumn crop was cut at ground level on three dates. The plant parts at 0-3, 3-6 and over 6 cm, or 0-6, 6-12 and above 12 cm, respectively, for the first and the other harvests, were separated in leaf blades, petioles and stalks, and analysed for dry matter, fibre and nitrates. High yield was attained already at the second harvest, with about 62% represented by the upper plant parts. At low cutting heights the yield rose, but with a higher incidence of the basal fractions. Fibre and nitrate contents were higher in the basal parts, where petioles and stalks prevailed. As a result, cutting at greater height increased quality. Early harvest yielded sufficient biomass, unless cutting was done at ground level. Nitrate content increased with plant ageing, but not relevantly in the upper fractions. The basal parts also had high fibre contents, not compatible with acceptable quality. A harvest stage optimising yield and quality can be identified. INTRODUCTION Palm-tree kale (Brassica oleracea L., ssp. acephala DC, var. sabellica L.) is a minor form of Brassica oleracea, grown in some areas of the northern Apennines, (central Italy). Although described as a rare and recently rediscovered crop (Hammer et al., 1990; Hanelt, 1998), its cultivation is well established in the native areas. Traditional cultivation areas, uses and cropping systems were recently described, together with the opportunities for new ways of exploitation (D’Antuono and Neri, 1998a, b). The perspectives for traditional Portuguese Brassica landraces were discussed by Monteiro and Dias (1996), who pointed out that their success as “new” products depends on consumer attitude, and that the quality of up to date products, such as frozen greens, strictly depends on flavour and texture, in turn determined by plant age and size. Palmtree kale recently attracted the interest of the food industry for its potential as frozen vegetable. The results of previous research showed its good adaptation to intensive, highdensity cropping systems, similar to industrial spinach, requiring a 60-80-day cycle in autumn as well as in spring. Nitrogen was a key factor for yield and quality, especially with respect to leaf colour (D’Antuono and Neri, 1998a). Efficient mechanical harvest requires a minimum biomass yield, with specific quality characteristics for subsequent processing. This research was a further contribution to the evaluation of palm-tree kale as an intensive crop, particularly aimed at: a) analysing the possibility of scaling up production from plot to commercial level; b) checking the findings of previous experiments; and c) examining yield and quality variation as a function of cutting height. MATERIALS AND METHODS A field trial was established at Cesena (northeast Italy). A 3500 m plot was sown Proc. IS on Sust. Use Of Plant Biodiv. Eds. E. Düzyaman &Y.Tüzel Acta Hort. 598, ISHS 2003 124 with a SAIS (Società anonima italiana sementi) palm-tree kale selection. Sowing was done by a mechanical drill, at a target density of 250 plants.m on 11 September 1998. Fertilisation was only applied as abundant manure, in the form of wastes from a vegetable processing plant. No weed control, pesticide or insecticide was needed. The experimental treatments were three cutting heights and dates. Dates were 27 October, 11 and 24 November 1998. Cutting heights were 0, 3, 6 cm, on 27 October, and 0, 6 12 cm, on the two other dates. Cutting heights were simulated by cutting at ground level three randomly selected, 6-m plots per thesis, and then subdividing a representative subsample in fractions corresponding to the three cutting heights. The following traits were determined: total above-ground biomass; its partitioning in the fractions obtained from the different cutting heights (respectively 0-3, 3-6, above 6 cm or 0-6, 6-12, above 12 cm, according to cutting date); dry-matter content; biomass partitioning into stalks, leaf petioles, green and yellow laminas, for each height fraction; nitrate and crude fibre of each fraction; specific area and chlorophyll of green blades. Nitrates were extracted from 1 g dry powdered sample with a 71:10:5:14 aluminium sulphate:sulphaminic acid:boric acid:silver sulphate extracting solution, and analysed by an ion-selective electrode. Crude fibre was determined according to Weende (AOAC, 1975). Specific leaf area was calculated from the ratio between blade area (∆-T electronic area measurement system) and dry weight. Chlorophyll was determined by grinding a sample of 50 g leaf under liquid nitrogen; a 0.5 g subsample was extracted with 80:20 acetone:water solution, and kept in a refrigerator in the dark for 12 hours. After filtration, the sample was read with a spectrophotometer at 664 and 647 nm. All data were subjected to an analysis of variance. RESULTS Total biomass (Table 1) markedly increased from the first (27 October) to the second (11 November) sampling, without further relevant increase at the third harvest. In all cases, most biomass (>60 %) was present in the upper plant part. On average, leaf laminas represented 24-28 % of the total fresh biomass, with only slightly significant interaction between cutting height and sampling date. However, big differences were observed in the composition of the three cutting-height fractions (Table 2). In fact, the upper fraction mainly consisted of leaf laminas (62-65 %), with the remaining biomass represented by petioles and only a small fraction of stems (<5 %). The proportion of laminas in the upper fraction did not depend on the cutting date, as indicated by the absence of significant "cutting date x height" interaction. The lower and intermediate fractions, however, were mainly represented by stems and petioles, with only a low amount of blades. The percentage of yellow leaves was low in all fractions. Dry-matter content (Table 3) followed a characteristic pattern. In fact, it decreased from the basal to the middle fraction, where less lignified stem parts prevailed, and attained the highest value in the upper part, because of the higher blade percentage. Dry matter markedly increased from the second to the third sampling, as a result of decreasing temperatures during late autumn, as also previously observed (D’Antuono and Neri, 1998a). Nitrate content (Table 3) was very high and markedly rose from the first to the third sampling date, except for the upper fraction, where it attained a lower level in the second sampling date, as indicated by the significant "cutting date x height" interaction. Crude fibre (Table 3) was also affected by the interaction of cutting date and height. In fact, fibre content of the bottom layer increased with delayed harvests, as a result of stem lignification; that of the middle fraction remained almost constant, whereas that of the upper parts markedly decreased from the second to the third sampling, as a result of leaf expansion, which reduced the ratio of vascular to parenchymatous leaf tissues. Specific leaf area and total chlorophyll of leaf laminas (Fig. 1) followed a similar pattern, as a function of the cutting date. In fact, both increased from the first to the