Simultaneous electronic and close‐up video recordings were made of behaviour during the initial 15 min of plant contact by adult, apterous Aphis fabae Scopoli on tick bean seedlings (Vicia faba Moench). Electronic techniques accurately determined stylet penetration of plant tissue and there was a close correlation between penetration and periods during which the insect antennae and body were immobile (r = 0.994, n = 60). Video techniques were then used alone to infer stylet penetration and the behaviour of aphids after various treatments was monitored. In particular, the time to first penetration, the number of penetrations, the mean duration of penetrations and the total time of penetration were observed. Behavioural differences were recorded between tethered (as required for electronic recording) and freely‐moving insects as well as between fed and starved insects. The behaviour of starved aphids placed on beans treated with the plant‐derived antifeedant, polygodial could not be distinguished from aphids on solvent‐treated control beans. However, there were significant differences in behaviour of aphids which had previously been exposed to polygodial on plant or green/yellow paper surfaces for 24 h when compared with insects exposed to solvent alone. The possible modes of action of polygodial are discussed.

Final fruit size is the consequence of complex metabolic events of final fruit size by cell division is 3-hydroxy-3-methylglutaryl coenzyme A reductase (HMGR) and activity of the sucrose that occur between fruit set and maturation. Disruption of these non-fermenting 1-related protein kinase (SnRK1) complex. It biochemical and molecular processes at any stage during fruit growth will impact on final fruit size. Because fruit size is a is argued that sugar content and composition of sink cells impact on SnRK1 (and hexokinase) to modulate expression of function of cell number rather than cell size, factors affecting sugar-metabolizing enzymes, HMGR and molybdenum cofac-cell division cycle activity assume importance. In this paper, we focus attention on the metabolic control of fruit growth using tor (MoCo)-containing enzymes. These changes, in turn, impact on hormone metabolism by affecting allocation of the avocado as a model system. Three areas of current interest are purine-derived MoCo to aldehyde oxidase and thus the endoge-highlighted, viz. the contribution by isoprenoid metabolism in the control of cell proliferation, the role played by carbohydrate nous concentration of indole-3-acetic acid, abscisic acid and cytokinin (CK) to alter plant hormone homeostasis. These content and composition in signalling changes in metabolite status and gene expression and maintenance of plant hormone aspects are integrated into a model to explain the metabolic homeostasis. Central to the process of fruit growth and control control of avocado fruit growth and final fruit size.either directly or indirectly to alter gene expression. Appreciation of the pleiotropic effects of plant hormones, however, suggests that no single growth regulator can account for a complex process such as fruit morphogenesis. This is because plant hormones exert multiple control on organ development by alterations in concentration and as a result of changes in sensitivity of the affected tissues (Bradford and Trewavas 1994). Therefore, development must be considered the result of intricate spatial and temporal interactions between the resources required for growth and hormonal mediation through the regulation of gene expression. Even so, the fruit developmental programme remains obscure. As stated by Gillaspy et al. (1993): 'Despite centuries of intensive genetic selection of agriculturally valuable fruit, we still lack most information about how fruits develop, how this development is coordinated with embryonic development and seed formation, and the molecular, cellular, and physiological events that control fruit growth and differentiation.'

a b s t r a c tLosses in the production of avocado (Persea americana (Mill.)) are incurred due to Phytophthora root rot (PRR), a disease of the feeder roots that results in tree-dieback and eventual tree death. Avocado is also a flood-sensitive species and flooding exacerbates the effects of PRR. The avocado industry relies on the use of rootstocks tolerant to PRR to minimise losses. The present study compared the gas exchange and chlorophyll fluorescence responses of avocado rootstock plants of 'Dusa™', the current South African industry standard, with 'Duke 7', and the selections R0.12 and R0.06 which show reduced and superior tolerance to PRR, respectively. A decline in stomatal conductance (g s ) and net CO 2 assimilation (P N ) over the 30 day evaluation period were early responses to flooding. 'Dusa™', the more tolerant rootstock plants, demonstrated a better recovery in P N and g s in response to inoculation; however, both rootstocks performed poorly under flooded conditions. A decline in P N in infected 'Duke 7' plants appeared to be associated with stomatal limitations due to reduced stomatal conductance. The decline in P N and g s was not apparent in infected 'Dusa™' plants. Non-stomatal limitations to P N in rootstock plants exposed to flooding were also evident as indicated by increases in the ratio of internal to atmospheric CO 2 concentrations (C i /C a ). Impaired photosynthetic capacity in flooded rootstock plants was reflected by reduced photosystem II efficiency and photochemical quenching. In comparison to 'Dusa™', R0.12 rootstock plants showed reduced P N and g s following inoculation with Phytophthora cinnamomi whereas the more tolerant R0.06 rootstock plants revealed sustained photosynthetic activity. Interestingly R0.06 was the only rootstock able to maintain P N and g s in non-inoculated, flooded plants.

Exceptionally high yielding (>100 t ha-1) apple orchards (Malus domestica Borkh.) are becoming common in South Africa and elsewhere in the world. However, no accurate quantitative information currently exists on the water requirements of these orchards. Information is also sparse on the water use of young apple orchards. This paucity of data may cause inaccurate irrigation scheduling and water allocation decisions, leading to inefficient use of often limited water resources. The aim of this study was therefore to investigate the dynamics of water use in eight apple orchards in South Africa planted to Golden Delicious and the red cultivars i.e. Cripps' Pink, Cripps' Red and Rosy Glow in order to understand how canopy cover and crop load influence orchard water use. Four of the orchards were young (3-4 years after planting) and non-bearing, while the other four were mature high yielding orchards. Transpiration was monitored using sap flow sensors while orchard evapotranspiration (ET) was measured during selected periods using eddy covariance systems. Scaling up of ET to seasonal water use was done using a modified Shuttleworth and Wallace model that incorporated variable canopy and soil surface resistances. This model provided reasonable estimates in both mature and young orchards. The average yield in the two mature 'Cripps' Pink' was ~110 t ha-1 compared to ~ 88 t ha-1 in the 'Golden Delicious' orchards. However, average transpiration (Oct-Jun) was ~ 638 mm for the 'Cripps' Pink' and ~778 mm in the 'Golden Delicious' orchards. The peak leaf area index was ~2.6 and ~ 3.3 for the mature 'Cripps' Pink and 'Golden Delicious' orchards. So, canopy cover rather than crop load was the main driver of orchard water use. Transpiration by the young orchards ranged from 130 to 270 mm. The predicted seasonal total ET varied from ~ 900 to 1100 mm in the mature orchards and it was ~500 mm in the young orchards. Orchard floor evaporation accounted for ~18 to 36% of ET in mature orchards depending on canopy cover and this increased to more than 60% in young orchards.

The role of cytokinins in the promotion of flowering in the endangered species Kniphofia leucocephala Baijnath. was investigated using shoots maintained in culture for 3 years. The highest percentage flowering (65%) was obtained on media containing 20 lM benzyladenine (BA). The inclusion of isopentenyladenine and zeatin in the media also resulted in flowering, but these treatments were less effective than BA in inducing flowering. The effect of cytokinins on flowering was dose-dependent, with high concentrations of BA inhibiting flower formation. Treatments that resulted in rooting of explants produced no flowers. The resulting inflorescences in all treatments did not mature and senesced prematurely, even when gibberellic acid (GA 3 ) was applied post-flower-emergence.