Ray [3,4,5,6] described in detail the evolution of ecological interactions among the replicating creatures of the original Tierran world. Those interactions are not discussed in this paper, because this paper is focused on more macroscopic statistical properties of the system. However, readers familiar with the earlier work may notice in Figure 1 that the parasites (the lower clouds of points) generally die out by half way through the run. Yet, in the upper part of Figure 2, parasites persist throughout the run.
The run illustrated in Figure 2 was conducted with relatively lower mutation rates, than the runs shown in Figure 1. At higher mutation rates, optimization proceeds more rapidly. It is when optimization is well advanced that parasites tend to die out. In Figure 1 it can be seen that parasites tend to persist longer in instruction sets two and three which achieve only moderate optimization, than in sets one and four that achieve higher degrees of optimization. Once the algorithms have been significantly compacted through optimization, the evolution of ecological interactions becomes more difficult, and these interactions tend to disappear from the evolving communities.