Dr. Murai Éva szerk.: Parasitologia Hungarica 22. (Budapest, 1989)
The basic rules of simulation were the following, (il Only flies staying on the cattle were considered. Only this active minority (almost exclusively females) collecting the females in a certain stage of their ovarial cycle is important from our point of view. The largest part (99-99.5%) of the population was handled as a black box called "pasture compartment" (see PAPP and GARZÔ 1985). (ii) The flies included in the simulation move randomly and they do not make any difference between the cows. If a fly had already foraged it is contaminated with bacteria, (iii) The duration of staying on the eye is also random, (iv) If a fly succeeded to stay on the eye a long time (more than one minute) for four times, then it exits from the active minority into the "pasture compartment". This quitting specimen is replaced with an other non-contaminated one from the pasture compartment. This stipulation decreases the transitional ability, but its reality can be justified with observations. Namely, this fly possibly had sucked up enough protein to make its eggs ripen so it leaves the cow. On the body of the fly Moraxella bovis is viable for 3 days. At a constant 30°C temperature the ovarial cycle of the fly takes three days. In Hungary, even on the hottest days the ovarial cycle takes at least 5 days. Thus, when the fly returns to the cattle, the bacteria taken up previously are not alive, (v) If a fly leaves a cow, its probability to return to the same cow is ten times higher than that of its landing on another cow. We have no concrete observations confirming this idea (we plan to measure this probability). We keeo much lower values to be true, but we consciously overestimated this probability in order to reduce the number of changing hosts in the simulation. The frequency of host change, flies occurring on the eye of cows were the target variables of our simulation model. The simulation was done on an IBM AT compatible computer, with Monte Carlo stochastic simulation. The program was written in TURBO PASCAL language. Since large amounts of random number were generated, and this is a time-consuming process, 20 cows were considered in our program. According to the field data, about 13 flies per cow were counted. The simulation b.eing stochastic, this number naturally fluctuated under the simulation. One running of the program simulated 12 hours (in Hungary, M. autum nalis imagoes are active for a longer time in summer). The program was run 100 times. COMPUTER SIMULATION RESULTS The results are summarized in diagrams 5-10. According to the simulation results, about three host changes per cow per hour can be counted (Diagram 5: mean 55.683 per 20 cows). The most frequent short stay (less than 10 seconds) was 68 flies per cow per hour. The most frequent middle stay (from 10 seconds to 59.9 seconds) was 8-9 flies per cow per hour. 35 flies per cow occurred for more than one minute. Since counting averages may omit individual differences, the number of long stay visits (Diagram 10) were examined more thoroughly. As one can see, these data have very large standard deviation. During the simulation there occurred a cow with 112 long stay visit (see more below). From the simulation results we can infer that in summer on Hungarian lowland pastures only the activity of M. autumnalis imagoes is enough to keep the infection on each animal In a herd if there were animals contaminated with the infectious agent (for example Moraxella bovis) . We have to say that the simulated fly densities were far under the Hungarian maximum (cf. PAPP and GARZÔ, 1985) and the model parameters were mainly underestimated. The simulation results suggest that, in the presence of other predisposing factors, the role of populations of M. autumnalis imagoes in causing IBK is a kind of the last drop into the cup to overflow.
