Physiological causes of yield variation in cassava (Manihot esculenta Crantz)

Cassava (Manihot esculenta Crantz) is an important crop in many parts of the tropics, being mainly cultivated for its storage roots. Farmers' yields are low and one of the constraints to higher yields is the lack of adequate clones. At the beginning of the 1970s an extensive cassava research programme was started at CIAT (Colombia). One of its aims was to develop high-yielding clones by genetic modification of the plant habitus.This thesis begins with a literature review in which the available information on the physiological determinants of the yield of cassava storage roots is described.Next, a series of experiments carried out to deepen and to broaden this knowledge on physiological causes of yield variation in cassava is described. MCol 1684 (the best cultivar of the CIAT cassava germ plasm bank) and MPtr 26 were used as the reference cultivars in the experiments, both in the field and in the greenhouse.Rate of leaf photosynthesis was measured by infrared gas analysis. Measurements of the photosynthetic rate were carried out using the youngest fully expanded leaf from plants growing outdoors that were 35 - 45 days old. Maximum photosynthetic rates varied from 0.74 x 10 -6to 0.81 x 10 -6kg CO 2 .m -2leaf.s -1. MCol 22 had the highest leaf photosynthetic rate. A relatively low photonflux density level was required for light saturation of the photosynthetic rate. This is characteristic for a plant species with a C 3 cycle. Photosynthesis increased only slightly from 1000 to 1500 μE .m -2.s -1PAR (photosynthetically active radiation). Light efficiency at low light intensities (α) varied from 9.0 x 10 -9to 12.4 x 10 -9kg CO 2 .J -1. The CO 2 concentration remained at an approximately constant level in the intercellular spaces, independent of the light level, being 212 vppm (0.387 x 10 -3kg.m -2). At a photonflux density of 1500 μE.m -2.s -1mesophyl resistance was higher than leaf resistance to CO 2 (335 s.m -1compared with 185 s.m -1). Transpiration rates did not differ between clones, but increased with light intensity. water use efficiency (WUE)varied from 15.1 to 17.1 mg CO 2 uptake per g H 2 O, and was most efficient for MCol 22.Linear relationships were found between total dry matter yield and the amount of intercepted PAR. Photosynthetic efficiency varied from 1.9% to 2.5%, based on PAR during the first six months of the growth period, and decreased markedly in older plants. The fraction of incoming intercepted PAR varied from 43% to 69% during the first six months. Cultivars had an extinction coefficient (K) of 0.72 to 0.88 and their leaves were dominantly planophile.A leaf area index (LAI) of 1.0 (about 50% light interception) was attained at 60 to 90 days from planting. An LAI = 3, which coincides with a light interception by the leaf canopy of approximately 90% was reached 120 to 150 days after planting, so about 40% of a growth period of one year had elapsed before complete ground cover was achieved. Genotypes with very different canopy characteristics reached an LAI of 3 in approximately the same time.Cassava has an indeterminate habit with sympodial branching. The length of the period until first branching depended on genotype and planting date. Large genetic differences were found in leaf life, leaf size, plant age at which maximum leaf size was reached and leaf formation rate per apex. Small variations in environmental conditions caused significant differences in canopy characteristics.During the growth period of cassava two periods with constant dry matter partitioning could be distinguished, with the apparent initiation of the filling of storage roots (AISS) being the crucial point. A constant proportion of the dry matter formed is distributed to the storage roots: this is the efficiency of storage root production (ESRP). Genetic differences were found for ESRP and AISS.The influence of daylength on growth and yield was studied. Daylength was increased by light bulbs, while the other growth conditions remained similar. Daylength had only a very slight influence on AISS value. ESRP was negatively influenced by long days. Differences in ESRP were the main cause of differences in yield of storage roots for plants grown at different daylengths. For MCol 22, the ESRP value was also considerably lower under long-day conditions, but the yield of storage roots was nevertheless only slightly reduced, because of the higher total dry matter yield. MCol 22 is the first detected cassava clone whose yield of storage roots is nearly dayneutral and thus it is suitable for cultivation at higher latitudes. Long-day conditions caused a large increase in LAI. The higher LAI values were caused by a higher leaf formation rate per apex and a larger number of apices per m 2. Photosynthetic efficiency was not influenced by daylength.The implications of the findings are discussed. Plant properties that could be useful for improving storage root yield are: a high ESRP value, a low AISS value, a high growth vigour (short period until 50% light interception), a light interception of about 90% as long as possible, and a good sink potential.Present knowledge of cassava cultivation techniques is such that high yields are possible. Potential yield of storage roots is about 30 t.ha -1.year -1on dry weight basis and about 90 t.ha -1.year -1on fresh weight basis. The highest recorded yield of fresh storage roots at the CIAT experimental farm is of that order (namely 82 t.ha -1), and was achieved after a growth period of one year.