From the presented data it seems clear that thanatotic reactions of aphids are more diverse than previously studied (Wohlers 1981) and may be linked to various biological features and ecology of aphids. Such features, which seemed most related to variance in thanatotic response, were presented in Table 2. It seems that the aphids response to external stimuli in our experiments was correlated with the combination of their ecological adaptations rather than purely with the phylogenetic relationships of the aphids tested.
In case of A. fabae, we may consider its behaviour as typical “last chance” defensive, thanatotic response, with immobile and spread legs, probably making it difficult for predators to swallow the prey (Honma et al. 2006).
Within single subfamily Lachninae we have the whole spectre of reactions, which seem to depend on the life mode of aphids. A species which seems to have the most primitive, typical kind of response is E. rileyi which exhibits thanatosis for a significant period of time. This species feeds on needles of pines, without protection of ants and with a thin wax layer at most and high above the ground level. Additionally, individuals of this species live singly, not in the colonies (Kanturski et al. 2015, 2016). When disturbed, it escapes quickly towards branches or twigs, often falling off the tree. Also the weather conditions may influence their often drop off. In this case, it is not surprise, that aphids casted down by wind very often are adapted to feigning death after the fall, to avoid being treated as a prey. Opposite situation exists in the case of T. troglodytes, which lives underground and basically inside the nest of ants (intranidal). When the ant nest is disturbed by a potential predator, ants’ aggressive and protective behaviour seems to be the most significant defence mechanism applied by mutualistic aphids. There is even small need for ants to catch aphids and escape with them, as aphids may instantaneously escape to lower chambers of ant nest, which they actually do. Such opposite life modes may explain opposite reactions of these two species.
C. (Sch.) pineti shows the life mode similar to E. rileyi, both species sometimes even live in mixed colonies, but the former is protected by quite significant wax covering which serves as a protection layer against predators and weather conditions. Sometimes colonies encompass the whole needle. Probably defence mechanism in the form of dense wax cover and/or better adhesion to a needle is better than dropping off the plant. In this case the thanatotic response was not an adaptive feature undergoing selective pressure and was lost during evolution. Similar case concerns T. annulatus which is not covered with wax and lives separately, but on the underside of oak leaves. It seems that dropping off the broadleaved tree provides greater chances of falling on another leaf and hiding under. Also in case of L. roboris which lives on twigs and branches of oaks the escape and dropping off can be a good solution – as proven by unresponsiveness of 6 of 14 individuals (Table 1.). Escape is typical behaviour of these species when disturbed. However, in case of L. roboris and contrary to C. (Sch.) pineti and T. annulatus, it lives in obligatorily mutualistic relationship with ants. Close encounter of dropped off L. roboris with ant worker, often of the same species tending it on original feeding location on a tree, brings no harm to aphid. Moreover, it may be moved back to colony of proper feeding location if submissive – thanatotic. In case of C. (Sch.) pineti and T. annulatus such encounter may be more risky – quick escape can be better adaptation.
The observed double behaviour of L. roboris meets prerequisites for the evolution of an adaptive trait by natural selection expressed by Endler (1986) (Miyatake et al. 2004): variation among individuals and fitness differences. Both traits may be adaptive: thanatosis – either to encounter with ant worker or with predators and active dropping off/escape, in case of danger. On a tree, contrary to intranidal T. troglodytes, ants may not be sufficient protection against enemy or other disturbance.
The final case concerns S. graffii and M. submacula which both present very similar behaviour of curling legs towards the body, resembling pupa-like position of submissive ant workers. Both species are also obligatorily myrmecophilous, often living either in close proximity of ant nest e.g. on a tree or shrub situated directly on the ant nest, in the chambers of soil built around the stems, or inside the nest – at stem or trunk base (Depa 2012; Depa et al. 2012). In these cases aphids face very high possibility of encountering ant worker, yet the escape path to ant nest may not be obvious for a single aphid to walk alone. A further difficulty occurs in S. graffii, whose very long mouthparts disables them to be extracted quickly (it usually takes a few minutes) and quick escape. Furthermore, this genus is proven to produce pheromones imitating those of ants tending them (Endo and Itino 2013). In both cases, thanatotic reaction evolved into submissiveness and the trait exhibited partially by L. roboris had adaptive significance. It may be also suspected that aphid’s curled legs make the transportation in soil tunnels and chambers easier for ant worker. But we also know, that some aphids, particularly L. roboris communicate with ant workers by kick-like movements of their long, hind legs. They do such movements when they are ready to extract honeydew (Hölldobler and Wilson 1990). Perhaps curling legs is also a first communicate for ant workers that it encountered a submissive individual, inhibiting potential aggressive behaviour. After proper recognition by ant, it then allows transportation.
In our opinion, these two kinds of responses resemble two kinds of responses met in anuran amphibians (Toledo et al. 2010) where thanatosis and shrinking were observed. While thanatosis concerned non-toxic species, the shrinking seemed to be either adaptation to being swallowed and spit out after excreting toxins or to avoid further injuries in case of struggling. In aphids, we may consider the behaviour of S. graffii and M. submacula as shrinking, but being adaptable to mutualism – quick and efficient transportation to a safe place by ant worker.
It seems that the degree of the thanatotic reaction strongly results from strictly ecological adaptations and has undergone the independent development in particular evolutionary lineages in aphid subfamily Lachninae. It seems that two factors influenced its development the most: feeding location and relation with ants. Definitely the initial state was a thanatotic reaction to fall/drop off – similar to A. fabae or other non myrmecophilous Aphidinae (Wohlers 1981). It was later either enhanced, when aphids were strongly exposed to fall (E. rileyi) or lost, if not (living underground – T. troglodytes – although mutualism with ants is definitely involved) or other protective means were applied (C. (Sch.) pineti – wax cover, colonies). When very strict, mutualistic relationship with ants developed, including close proximity of ant nest, the defensive thanatotic response transformed into mutualistic adaptation to transfer by ant worker – shrinking.
Still, however, the subject requires further confirmation with experimental work concerning various morphs as well as various systematic groups of aphids. With a whole set of defensive mechanisms in apterous, sedentary morphs, an extent of variability in thanatotic response may be even larger. It is also important to test, whether in highly polymorphic aphids thanatosis is limited to aptrous morphs only or it occurs also in alate morphs, where fly off could be alternative way of avoiding danger (Ohno and Miyatake 2007).