Both species, like the great majority of bdelloid taxa, are capable of surviving desiccation through anhydrobiosis, although this capacity may differ among the different species . A. r. and M. q. survive desiccation in high percentages and recover a few hours after the addition of culture medium. The two species, however, came from habitats with different frequencies of drought. A. r. was sampled in dry sediments of a billabong, that is fed by floods and may dry out during Australian summer and that represents a temporal water body. M. q. was originally collected in a spring-fed-pond that hardly dries out completely. When entering anhydrobiosis, both arrest activity, undergo morphological changes and stop reproduction . At re-hydration, both species resume reproduction ignoring the time spent dry, but while desiccation does not affect lifetime fecundity of M. q., it is known to promote post-anhydrobiosis fecundity of A. r. [13, 15]. In each species a strain was isolated and was desiccated at regular intervals. Due to the relatively short lifespan of the two species, it is unlikely that a rotifer experienced desiccation twice in its life, but this has been found to have no effect on its recovery probability . Recovery rates were mostly invariant during the different months for A. r. and randomly variable in M. q.; no trend of recovery rates was evident in either species. The fact that a strain after being desiccated reiteratively does not either increase or decrease its recovery capability suggests that neither selection nor adaptation were operating on it. This is not surprising since no genetic variability is expected to occur in a short time under strict ameiotic parthenogenesis.
Nevertheless, in both species 'constant conditions' produced progressive decrease of fitness components with time of cultivation, while the stress induced at regular intervals maintained life-history traits stable in time. How constant a laboratory environment can be throughout years is hard to state, and the condition experienced by a same line in different months may have produced changes of fitness traits. However, paired life table experiments of H and D cohorts were run at the same time under similar conditions: they shared thermostatic chamber, temperature, time of the year, medium, food and experimenters. Yet, the two lines differed, and the responses of the two species to the desiccation were similar, as a general trend. The constantly hydrated line of each species registered a progressive decline in all fitness-related traits, in particular mean life-time fecundity and early reproduction (both age at first reproduction and number of eggs till 10-d-old). In contrast, the strain that was regularly dried had higher fitness-related traits than its parallel H line, but the same traits remained constant if compared to those of the ancestor population, M0. Surprisingly, it is not the treatment that affected fitness, but the absence of treatment that impaired fitness.
It is commonly known in most organisms that processes like reproduction necessitate resources that are diverted from maintenance, with the consequence that high reproduction is associated with short lifespan . Although this does not seem universally true [e.g. ], reproduction and fecundity are known to trade-off in monogonont rotifers, as well [e.g. ]. However, no inverse association between fecundity and longevity is evident in both bdelloid species; lifespan appears unrelated to fecundity in A.r., and is directly, and not inversely, associated in M.q. Both species under lab conditions go through a post-reproductive time that may be long; thus, if an early death impedes the rotifer to produce more eggs, the opposite is not true. Surprisingly, recovered bdelloids produce offspring that lay more eggs AND live longer. In other words, the subpopulations of bdelloids that were regularly dried reproduced more and lived longer. But, also maintenance and eventually repair mechanisms are costly processes, so there might be a cost in recovering from desiccation. If this is true, it is not paid in reduction of fecundity or of lifespan.
The differences of fitness correlates between the parallel subpopulations, H and D, must be seen as phenotypic responses induced by the different treatments. We might advance the hypothesis that there could occur some mistake capable of accumulating unless a stressful event reveals its presence by promoting a check-up. Perhaps some 'mistake' reduces fitness but can be removed at recovery after some severe stress, like desiccation. Desiccation implies that the animal loses water, re-adjusting tissues, cells and organules, maybe DNA and ribosomes as well; at re-hydration all structures must resume their original shape and function. But it is not unlikely that structures need to be checked and repaired before activity is resumed. If this is true, then the action can also promote the control over 'damages' accumulated during active life. Thus we can hypothesize that continuous parthenogenetic reproduction, obligatory to the bdelloids, produces the accumulation of some 'mistake' in their cells, and that emergence from anhydrobiosis promotes the removal of the 'mistake' and the re-establishment of the original condition. Alternatively, it could be that bdelloids host viruses or parasites, whose load increases over generations reducing the fitness of the population if conditions are constantly suitable. If conditions become harsher and the virus or the parasite are less tolerant to desiccation than the bdelloids themselves, then the desiccation could represent a way to decrease the parasite/virus load and keep animal fitness constant.
During the duration of the experiment, we observed a regular and progressive decrease of fitness. If this result means that the bdelloid populations maintained permanently hydrated are committed to extinction in the long run, is premature to state. We can imagine that fitness may drop and then level off, possibly at a level sufficient to sustain long-term survival, or, alternatively, the decay of the population can be overcome by a large population size. As already stated, not all bdelloid species live in habitats that dry out, but a minority (less than 10% ) occur in permanent water bodies, and few of them were found not to recover after desiccation . Apparently these species do not become extinct in absence of anhydrobiosis. We might advance the hypothesis that these species are all recent ones and are doomed in the long run, but this aspect requires further investigation. On the other hand, based on a previous experience on another bdelloid species , even a single event of desiccation improves its fitness components and can thus rescue the hydrated population. On the basis of these observations, it seems unlikely that the decline is due to detrimental mutations, as predicted by current theory [see  and literature therein], but it seems more likely to be epigenetic in nature.