Biological control of fruit-tree red spider mite

During the last decade, integrated pest control systems have been developed for several crops. One of the main fields of research in integrated control has been the control of orchard pests. Experience with modified spraying programmes in apple orchards, the increasing resistance of spider mites to acaricides, and knowledge on the bionomics of many pest species have been major factors in promoting the development of integrated pest control systems. Attempts were made to introduce predatory mites or to improve their effectiveness in the control of the fruit-tree red spider mite, one of the major pests in commercial fruit orchards. These experiments clearly showed how these natural enemies can reduce and maintain spider mite populations below the economic threshold level. At present predacious mites are widely applied in the control of spider mites in apple orchards. However, the resulting changes in the system have still not been quantitatively assessed, and it is only speculation to explain the mode of operation of the system if there is no information about the underlying ecological processes. For a stable pest control system one must know how spider mite and predacious mite populations interact with each other and with the host plant, and how the system is influenced by abiotic factors (temperature, relative air humidity, wind and rain) and by cultivation methods (including the use of fertilizers as well as insecticides and fungicides). In several countries with a developed agriculture, research has therefore been started to monitor the effect of predators on pest populations.

This study presents basic models for the fruit-tree red spider mite (Panonychus ulmi) and the native predacious mite, Amblyseius potentillae. The models are constructed according to the state variable approach, as is described in Chapter 3. The models developed with this technique bridge the gap between biological control with predacious mites in the field and the analytical methods of natural sciences, thus assisting in the introduction and management of biological control agents of the fruit-tree red spider mite.

The simulation models are based on extensive knowledge of the effect of temperature, humidity, food and daylength on the prey as well as the predator. The relations between rates of development, mortality, oviposition and diapause with temperature and other physical factors were determined from literature studies, estimation and many laboratory experiments, Chapter 5. Many of the temperature responses of rates proved to be linear and reacted momentaneously to temperature fluctuations.

The predator-prey interaction (between predacious mite and fruit-tree red spider mite) in these models, which closely approximates the field situation, is based on a detailed analysis of the predation process. This predator-prey interaction is very complex. Five developmental stages of the prey (larva, protonymph, deutonymph, adult male and female), and four developmental stages of the predator (protonymph, deutonymph, adult male and female) are involved. The attractiveness of the different stages of the prey varies and depends partly on the satiation level of the predator. For example, the adult female predator (the most voracious stage) shows a strong preference for the younger stages of the prey, but 'hungry' predators are much less selective. The rate of ingestion and the utilization of a killed prey also depends on the satiation level of the predator. Fransz's detailed analysis of the predation process in the system two-spotted spider mite and predacious mite and the explanatory models he developed for this process showed that a simple system (one standardized predator and a constant number of preys) reaches an equilibrium within a few hours. Hence the degree of filling of the gut of the predator oscillates with a small amplitude, at a level depending on predator and prey density and on the temperature of the system. This enables the complex predation process to be incorporated in a model for a population of higher order by simply expressing relative predation rate and prey utilization as functions of temperature and state of the predator. The satiation level of the predator can be quantified visually, because well-fed predators are dark, while hungry predators are whitish and transparent. A colour scale has been developed which relates the behaviour of the predator expressed in success ratio (number of successful encounters to the total number of encounters) to the quantity of leaf and animal pigments in the predator, which together constitutes its colour.

Experiments were carried out to determine the rate of decrease in colour value, which is supposed to equal the digestion rate, the relation between predation rate and prey density, and the relation between predation rate and colour value at various temperatures. The required relations for relative predation rate and prey utilization are easily derived from these functions. Oviposition rate and the development rate from egg to adult of the predator (numerical response) also dependon its satiation level and on temperature. These relations are also experimentally quantified, Section 6.3.

The details of information required on the driving variables, temperature and food condition were determined by experiment and simulation. The effect of nutritive condition of the tree on the prey was determined in water culture experiments and related to the nitrogen content of the leaves in commercial apple orchards. It is shown, Section 7.2, that the nutritive condition of the trees in practice do not affect the rates of development and oviposition of the fruit-tree red spider mite. To determine the required details about micro-weather an adapted and verified micro-weather simulator was coupled to the population model. The small differences between simulation results when leaf temperatures are the driving variables and simulations in which air temperature is the driving variable justified further calculations in the field with air temperature, Section 7.3.

The assumptions in the model underlying the treatment of the predation process were verified by comparing results of an independent experiment on predation in replacement series of 'prey stages', with simulation results. It is also shown that the procedure, for determining yields of a plant species growing in competition, from sowing density experiments in monoculture, may be applied for calculating predation rates of a species in 'mixed cultures' from its functional response curve in monoculture (see Section 9.2).

The models for hatching winter eggs, for population growth throughout the season and for diapause are verified at different levels of integration by independent population experiments. The most simple verification is the measurement of population growth in small ecosystems under controlled conditions in situations with and without predators and then to compare results with those of simulation.

Verification in the field is done by comparing simulation results with population measurements in several orchards. The correspondence in general pattern of population fluctuations of prey and predator and the good correspondence between simulated and measured colour values of the predators enables the model to be used for sensitivity analysis and management.

Sensitivity analysis showed that particular key factors are absent and that a wide range of initial prey-predator ratio's may be tolerated. It is further shown that the predation activity of the younger stages and the adult males is relatively unimportant and that the female predator is the important regulator due to its high predation capacity, its long lifespan and an increase in rate of oviposition until the predator is well fed. The system is rather sensitive to length of prey's juvenile period, predation rate, and oviposition rate of the adult female predator and the delay in development of the predator due to insufficient food.