From the Conclusions ... 'The conclusions of this study are not readily generalised to other types of savanna. The fact that the savanna never burns in its entirety each year, and never the same day, maintains a richer and more varied population.A certain number of species are probably present only because of the existence each year of unburned zones, e.g. Delegorguella phalerata and Thoria rotundata disappear almost entirely from the burned zones each year.Even the abundance of certain species in the burned savanna seems equally linked to the spared zones, those which hide in unburned zones in large numbers during one part of the year, waiting for the burned areas to reach a certain maturity so they can reproduce (e.g. Gellia dilatata, Lobopeltista guineensis, Thoria gillonae).If the savanna burned entirely each year, logically the population should become impoverished in sciaphilous species, since they could recover only by reproduction of the low levels of survivors.We have seen that if the population of pentatomids can accomplish an entire annual cycle after the passage of fire, it has time to become complex and rich before the next fire. Normally in this region fire is most probable in the dry season: December, January, or February. Out-of-season fires which affect the population late in the year and prevent its complete development before the following dry-season fire are rare and localized and usually deliberately set. Their impoverishing effect on the species diversity is zero.Fire is not a very destructive factor. It simplifies the habitat which thus maintains a population of pentatomids of heliophilous type. Certain authors (e.g. Adjanohoun, 1964) believe that fire maintains the savanna, and in its absence the savanna would evolve towards a wooded savanna and then a forest. However, in natural conditions, if fire does not occur in several successive years, the habitat changes indeed, but its chances of burning increase since the quantity of combustible material increases each year. It is only at the beginning of a certain stage in its succession that ligneous species are able to develop and thereby reduce the likelihood of fire. When this happens, the population of pentatomids continues to change, and that of the unburned savanna of this study is one of the first steps. At the limit of development, the heliophilous species tend to disappear at the same time as fire diminishes in probability. The population then becomes an exclusively sciaphilous type linked to the sub-forest area.The absence of fire appears to liberate the population to develop towards a different type. Because of this, fire seems to be necessary for the equilibrium of the pentatomid population so that it will remain similar from year to year. Fire cannot be excluded from this habitat without gravely disturbing this equilibrium which has been established for several hundreds of years.Thus, one can see how much the natural equilibrium of a group of animals can depend on factors in the habitat, and how a modification in only one is sufficient to upset it.'