The Evolution of Migration Routes in Birds and Insects

 

Consider an archipelago. Island inhabitants within an archipelago gain in fitness by moving between islands or between islands and mainland. This is done to exploit regularly changing conditions. Such behavioural adaptations lead to the evolution of migration routes. Migration, often over spectacular distances, is well known for many bird and mammal species. Typically, migration routes follow genetically fixed compass bearings as shown for birds in Fig. 1d, from (Davies, Krebs, & West, 2012).

 

 

 

Though we usually picture islands as land surrounded by water, reflection leads to the important generalisation that essentially all habitats are isolated patches surrounded by inhospitable terrain. In other words they are islands. Different species have different ecological requirements, so will perceive different places as habitable islands.

 

      In this essay I want to consider rain pools. There are many kinds of rain pool but all are transient habitat islands drying up frequently. When full they are exploited by a variety of aquatic animals, in prodigious numbers endemic to their kind of pool. Examples include pools in the foot prints of animals such as elephants, those in the axils of leaves, in pitcher plants, in tree hollows and on the surfaces of rocks (McLachlan & Ladle, 2001). It is the last of these, specifically those on sheets of rock in tropical Africa (Fig. 2), that is the focus of my interest here.  

 

FIG. 2. An example of pools (P1 and P2), near a river (r), flowing over a sheet of rock in Malawi in tropical Africa. In general such pools may be true rain pools while others are filled by river water as the river recedes during the dry season. From (McLachlan, 1988).  

 

 

There is quite a lot known about the ecology of pools on rock surfaces in Africa (hereafter called rock pools), and the adaptations of the indigenous animals inhabiting them (McLachlan & Ladle, 2001). I here contend that adaptive migration behaviour as seen in birds can be expected to evolve among the inhabitants of rock pools but not among the inhabitants of other ephemeral habitats. This is because rock pools are always to be found in the same place, which is not true of other rain pools. I return to this point below.

 

Principal among the inhabitants of rock pools are aquatic insect larvae that cannot survive the drying of their home pool. Individuals that manage to reach adulthood before the home pool dries must migrate to find a pool with water in which to oviposit. I here consider two such species, both midges in the Genus Chironomus. i.e. C. pulcher and C. imicola. Note that for rain pool species there is strictly one species in any particular pool and the association persists, when the pool is full, from year to year and presumably for thousands, even millions of years. It is worth noting too, that the genus Chironomus is an African genus of invaders, quick to exploit new opportunities but poor at competition with other species (Armitage, Cranston, & Pinder, 1995). In this sense these two rock pools breeding insects are preadapted (M. J.  West-Eberhard, 1992) p15.

 

 

 

     Now for some details:  C pulcher is confined to pools near rivers. Such pools are effectively permanent because even in the dry season new pools are formed as the river recedes, exposing new pools full of river water. The pools in Fig 2 are of this kind. During the rains, pools remote from rivers also fill and provide opportunities for invaders. It is these remote pools that are exploited by C. imicola. C. pulcher does not move from the archipelago of near–river pools, which are refuges for both species in the dry season. Figs. 3 and 4 below show the sampling sites where the work discussed here were done between 1970 and 2004.

 

 

 

 

FIG. 3. The study area, seen from the NW, is indicated by a broken line. It includes a heterogeneous landscape of rivers (r) and mountains (m) rising to a height of 250m above thinly wooded plains (dark patches are cloud shadows). X indicates the highest mountain within the study area. The photo was taken from 800m up Zomba massive. Mulanji massive (M), rising to a height of 2300m above the plain is barely visible in the far background. C, Chancellor College, University of Malawi.

 

FIG. 4. The distribution of pools inhabited by Chironomus pulcher (black dots), and those inhabited by Chironomus imicola (red dots), (a) in a dry season and (b) in the next wet season. Each point represents a single rock pool within the sampling are shown in Fig. 3. Pools dry at the time of sampling are not shown. Mountains are represented by contours at 50m intervals. The red arrow shows a predicted wet season migration route for C. imicola from river refuges. The principal refuge, measured as number of full pools at the time of sampling is indicated by a red circle.

 

 

I have made the point that rock pools are different from other rain pools and even different from other ephemeral insect habitats such as fallen fruit, carrion and the dung of mammals and that the evolution of migration routes can therefore be predicted for insects indigenous to rock pools, here C. pulcher and C. imicola. How might the evolution of migration routes have originated for these two species? The proximate requirement is that migration direction is genetically determined and that migration direction varies with the population. Individuals with migration direction resulting in the successful location of appropriate remote pools pass on their migration direction genes to the next generation, thus leading to the evolution of migration routes. Note that I am not here proposing adaptive phenotypic flexibility (M. J. West-Eberhard, 2003), but well understood genetically fixed phenotypic behaviour (Davies et al., 2012).

 

The hypothesis of migration routes for C. imicola can, in principal, be tested by tagging with a radio isotope such as P³². The whole population of larvae in a pool is tagged and allowed to reach adulthood. Egg masses from all pools within the sampling area are later collected to determine the distribution of P32 labelled egg massed from migrating females.

 

 

 

Now we arrive at last at my central point. Because of their spatial predictability over geological time scales, rock pools, possibly along with deep sea fumaroles, may be habitats unique on this earth. Their spatial relationships are presumably not even interrupted by tectonic place movements. We should not forget either, that rock pools are candidate places for the origin of life on earth (Mclachlan 2017). For these reasons it would be interesting to determine the phylogeny of these two strange rock pools breeding species. I predict that they will prove to be sister species, but if an hypothesis of sister species is rejected another interesting conclusion is indicated. I mean the independent evolution of two species adapting to an ephemeral habitat with low levels of interspecific competition and high levels of spatial predictability (McLachlan & Ladle 2001).

 

 

references

Armitage, P., Cranston, P. S., & Pinder, L. C. V. (1995). The Chironomidae. The biology and ecology of non -biting midges. . London: Chapman & Hall.

Davies, N. B., Krebs, J. R., & West, S. A. (2012). An Introduction to Behavioural Ecology. (4 ed.). Oxford: Wiley-Blackwell.

McLachlan, A. J. (1988). Refugia and habitat partitioning among midges (Diptera: Chironomidae) in rain-pools. Ecological Entomology, 13, 185-193.

McLachlan, A. J., & Ladle, R. (2001). Life in the puddle: behavioural and life-cycle adaptations in the Diptera of tropical rain pools. Biological Reviews 76, 377-388.

McLachlan, A. J. (2017). The living puddle. http://atholmclachlan.blogspot.com.

West-Eberhard, M. J. (1992). Adaptation: Current Usages. In E. Fox Keller & E. A. Lloyd (Eds.), Kewords in Evolutionary Biology. London: Harvard University Press.

West-Eberhard, M. J. (2003). Developmental Plasticity and Evolution. Oxford: Oxford University  Press.