In the present study, the social immunity behaviors of C. formosanus were enhanced when they were acting as hosts to A. farris. Termite grooming and vibration frequency were increased in the higher phoresy group. Termite vibration and grooming behaviors play a key role in social immunity. When termites are exposed to pathogens, they show distinct alarm behaviors, typically consisting of 2–7 s bursts of a rapid longitudinal vibration [6, 26, 33]. Following this, other termites approach and groom exposed nestmates to remove the pathogens from their bodies [42]. Changes in termite social immunity behaviors suggest that termites can recognize their phoretic status and use behavioral responses to remove mites and alert nestmates through high grooming and vibration frequency. The enhanced defensive behaviors were not an immediate state but an ongoing behavior, so we recorded termite behavior after 1 day. And the frequencies of their defensive behaviors were always high. It means the mite attached on termite very tightly and are difficult to remove, so it might not cost much time for the duration of each grooming episode. So, the frequency was increased, while the duration of each groom was decreased.
Phoretic mites show a preference for certain host attachment sites [3, 28, 32]. In the present study, there were only 1–2 mites attached to each termite (Figs. 4, 5.). Further supplementary research (Additional file 1: methods and results) found almost all A. farris attached to the head of the termite (Additional file 1: Table S1; Fig S1), and this phenomenon has also been observed among some Acaridae species and termites [39]. The head of C. formosanus is therefore considered a suitable site for A. farris attachment, and consequently, there is a limited amount of space available for mites to attach to. Differences in head shapes and microstructures between termite castes may lead to mites developing a preference for attaching to specific castes. Soldier termites, with drop-shaped, shiny, harder heads, may be more conducive to mite attachment, as mites are less likely to become dislodged during travel through various environments. Conversely, the oval-ellipsoidal shape and relatively soft texture of a worker’s head may limit the tightness with which mites can attach to the termite, causing them to be more easily removed (Additional file 1: Figs. S2, S3). This may explain the higher phoresy proportion in soldiers than workers (Figs. 4, 5). Further, workers feed themselves, while soldiers are fed by worker trophallaxis, potentially providing mites that attach to soldiers more opportunities for subsequently moving onto other termites.
Phoretic mites in the Acaridae hypopode stage are not generally believed to cause direct damage to the carrier, as their mouthparts are degenerate [15, 22, 29]. The mouthpart of the mite used in the present study is degenerate, and therefore the mite cannot feed on live C. formosanus. When frequency increases, duration decreases. This may be indicative of a trade-off, and hence a cost to grooming, i.e., grooming is costly so can only do it in short spurts, and when termites groom more frequently, the consequence is short duration of grooming bouts. Consequently, frequent social immunity behavior may cost higher energy and reduce termite fitness and hasten death. Thompson et al. [37] found termite social contact to be enhanced during exposure to stress, and energy loss was compensated by increasing trophallaxis after removing the stressor. In the present research, the enhanced defensive behaviors were not an immediate state but an ongoing behavior, so we recorded termite behavior after 1 day. And the frequencies of their defensive behaviors were always high. So, termites were under sustained stress, significantly increasing fitness costs may be one of many possible hypotheses and the flat abdomen of termites also provide evidence. The relationship between A. farris and C. formosanus was, therefore, closer to parasitism than to commensalism, suggestive of the phoretic mite being an intermediate precursor to the evolution of parasitism.
These findings provide direct evidence that mites reduced termite fitness. The findings of previous studies have also reported termite behavior to have been affected by mites, typically with negative effects on termites. Korb and Fuchs [20] reported that individuals were less active in termite colonies where mites were present. In a study by Phillipsen and Coppel [30], the hypopus of Acotyledon formosani was found to impede C. formosanus feeding. Wang et al. [39] found no significant increase in Reticulitermes flavipes mortality associated with phoresy by the mite Australhypopus sp., and the mite was not deemed to be a good candidate for the biological control of termites. The methodology used differed from the present study in several ways. Wang et al. [39] used feeding-stage mites, as opposed to hypopus-stage mites, so the effect observed was not phoretic. Further, while they did not report a significant increase in mortality, they observed that high mite densities occurred in weak termite colonies and that the termites exhibited lower body weights and flatter abdomens. As such, it could be inferred that high mite densities weaken termite colonies. The phoretic mite differs from predatory or typically parasitic organisms in that it indirectly disturbs the termites by attaching itself to their surface. Once the termite dies, the A. farris hypopus soon transforms into a tritonymph, a trophic developmental stage, and feeds on the body of the dead termite. Consequently, some mite species may be effective as indirect biological agents to control termite populations.
The present study found that in the field, A. farris was commonly attached to C. formosanus. In the laboratory, the phenomenon was more clearly observable, with phoresy proportions of 34.81% observed in workers and 60.41% in soldiers. Releasing mites into termite colonies has significant negative effects on termites. Chouvenc et al. [7] suggested that attempting the biological control of termites based only on small-scale laboratory experiments was unrealistically optimistic. However, the biological control agents they referred to were fungi and nematodes, which can be groomed away. The nature of mites is different from that of fungi and nematodes, having a greater potential to control termites as it cannot so easily be groomed away, therefore ensuring that it can bypass the social immunity defenses of termites. Further, A. farris can easily be propagated by being fed on cheese, stored rice, and wheat [16, 17, 23, 36], and so could potentially be mass-produced. As such, A. farris could provide a useful biological agent for the control of termite populations in specific circumstances such as the protection of ancient buildings and celebrated trees. Further research is necessary to determine its effectiveness and explore how to apply this technique in the field.