Aphid-plant interactions at phloem level, a behavioural study.

Aphid-plant interactions occurring during plant penetration, or probing, have determined the evolution of these insects as well as the abilities of plants to survive huge densities of these insects. For aphids, as phloem sap feeders, the interactions at sieve element level must be of special importance. During the phloem phase of probing, there are two aphid activities, categorised as waveforms El and E2 in electrical penetration graphs (EPGs), which have been studied here. The initial E1 activity appeared to be responsible for the inoculation of a persistent plant virus (BYDV), thus reflecting saliva injection into the sieve element (Chapter 2). No apparent sap ingestion was found during E1. The relation of the subsequent E2 activity with phloem sap ingestion was confirmed in this study, as the E2 duration appeared to be positively correlated with the acquisition success of the virus. Moreover, the results supported the idea that concurrent salivation during E2 does not reach the sieve element. Instead, it is presumably mixed with the ingested sap in the fused apical part of the two stylet canals.The "obligatory" occurrence of sieve element salivation (E1) before ingestion (E2) starts suggested a function in suppressing wound reactions of the sieve elements, such as phloem protein gelation or callose formation, or some other activity in preparation of phloem feeding. Therefore, the extensive saliva injection of an aphid colony might influence the probing behaviour of the individuals and explains the sometimes observed better performance. Hence, the occurrence and duration of phloem salivation (E1) was subsequently studied under this condition (Chapter 3). The aphid Aphis fabae showed reduced phloem salivation when probing on previously colonised bean leaves (Vicia faba), which seems in agreement with this hypothesis. However, no shortened E1 was found in the combination Rhopalosiphum padi-Triticum aestivum. Thus, no general conclusion can be drawn. Nevertheless, a beneficial effect of plant changes at phloem level could be inferred from the observed shorter phloem salivation and the more continuous phloem sap ingestion by A.fabae on previously infested plants. The ingestion rate, as reflected by recorded honeydew excretion, remained unaltered but the ingestion periods were prolonged.The effects of aphid manipulation during the experiments, including interruption of feeding, moving to a new plant depriving from food, etc., on the experimental results were investigated (Chapter 4). Also, the impact of attaching them to a wire for EPG recording was examined. Increasing the interval between removal from the plant and giving access to the test plant led to a reduction of time until the first probe on test plants. A shorter time until the first phloem phase was found in A.fabae when deprived of food for 1 min and replaced on the same plant, suggesting a short "aphid memory" and/or a better plant acceptability of the already probed plant. Also, a short "memory" effect was observed in the 1 minute intervals, slightly reducing the time until the first phloem phase. A general effect, however, was that an interruption of feeding resulted in a "reset" of probing behaviour, appearing as a similar sequence, duration and pattern of probing activities, irrespective of the length of the interval between 1 and 100 minutes. Confinement of an aphid by a wire slightly increased pathway activities, but once phloem phase was reached it did not affect phloem feeding, at least in tests with suitable host plants as were used here.The final chapter describes some more detailed aspects of the phloem phase activities. Phloem salivation was supposed to be prolonged when compared to a susceptible cultivar, during aphid probing on a line of Triticum monococcum with resistance to the aphid Sitobion avenae. This plant-aphid combination was used as a model system in order to investigate further the biological function of phloem salivation. However, it appeared that there was no prolonged phloem salivation when the parameters were selected critically. Additional honeydew excretion experiments with S.avenae and R.padi showed that E l activities were not related to honeydew excretion, i.e. no honeydew secretion occurred during this waveform. Also, no relation was found between variations in the phloem ingestion waveforms (E2), frequencies and amplitudes of certain details, and excretion rate. If these waveform variations reflect fluctuations in ingestion rate, the excretion rate seems not adequate to measure them