Author:

Abstract:

Insects are the fastest declining group of species and among the most threatened by anthropogenic activities. However, they are largely overlooked in conservation policies and conservation research. Wetlands have greatly declined over the last centuries and continue to face important threats, impacting all sectors of biodiversity, including insects. Odonates (damselflies and dragonflies) are valuable indicators of wetlands' conservation status. Habitat restoration (including habitat recreation) has been raised as a priority strategy during the current decade. Such conservation actions are facilitated using flagship species, but to be efficient, a thorough knowledge of their ecological requirements is necessary. This doctoral thesis focused on the threatened damselfly Lestes macrostigma, as a flagship species to implement the conservation and restoration of temporary brackish ponds. There was limited knowledge concerning the influence of hydroperiods and water salinity levels on eggs and larvae, which are key stages in odonate population dynamics. With the goal of undertaking evidence-based conservation actions, we aimed to fill these knowledge gaps by conducting four studies in the Camargue, southern France. This area is characterized by salt deposits and a Mediterranean climate, creating brackish and temporary wetland habitats which harbour L. macrostigma breeding populations. First, to improve our knowledge of the egg biology, we studied the embryonic development and hatching phenology and success in two temporary brackish ponds. We found that the eggs -laid in plant shoots during late spring- remained at an early development stage during summer, autumn and winter. This likely increases their resistance to extreme temperatures and illustrates the adaptation of L. macrostigma to temporary habitats with uncertain timing of flooding. Yet, embedment in the ice of the plant shoots decreases egg survival, suggesting that higher water levels above the shoots would reduce egg mortality caused by lower winter temperatures. Hatching lasted from mid-March to mid-April and was synchronised. Differences in hatching phenology among ponds were most probably due to differences in water temperatures. Water salinity is a major driver of insect communities. Hence, we studied in the laboratory the effect of salinity levels experienced during the egg and larval stages. We found no effect of sea salt on egg survival (hatching success) in the range 2-9.5 g/L. By contrast, larval survival increased with salinity level from 0.5 to 8 g/L but was reduced to zero at 16 g/L. Increasing salinity levels augmented development time until the adult stage, decreased larval growth rates and reduced adult size at emergence. The salinity level experienced at the larval stage had no effect on adult thorax muscle mass, flight performance, investment in immune function and available energy. However, energy consumption increased with higher salinity levels, resulting in a lower net energy budget in the adult stage. Our results therefore indicated that Lestes macrostigma is a saline generalist (fundamental niche) of which the realised niche shifted toward brackish waters. Other factors, especially biotic interactions, experienced during the larval stage may have consequences on larvae and carry over to affect the adults. We therefore led a field study across 33 ponds, testing for the effect of 14 biotic and abiotic variables on L. macrostigma larval abundance and adult size at emergence. We found that larvae were more abundant in smaller ponds and with later flooding in autumn-winter. This seemed mediated by predatory aeshnid dragonfly larvae, which were more abundant and bigger in earlier flooded ponds. A high susceptibility of L. macrostigma larvae to predators could be related to their high activity levels associated with their fast growth and development rate imposed by the drying out of their temporary habitats. Further, L. macrostigma tended to emerge larger from ponds with higher water levels, most probably because the lower water temperatures extend the larval growth period. By contrast, we found no effects of water salinity levels and availability of oviposition plants on L. macrostigma larval abundance. In order to investigate whether the preference adult have for certain plants to lay their eggs benefit the progeny, we used semi-controlled conditions to study the effect on hatching success according to the flooding date and the type of plant shoot where the eggs were laid. We found that wetting was mandatory for hatching. Eggs were able to cope with desiccation until flooding as late as April (i.e. after 10 months in unflooded conditions) -yet with low survival rate- but later flooding resulted in no hatching. Laying eggs in the Sea club-rush Bolboschoenus maritimus provided higher survival rate and faster embryonic development. Given that B. maritimus grows in deeper parts of the pond, thus experiencing earlier flooding, suggests that the preference of adults for this plant is an adaptive response to the uncertain flooding conditions of L. macrostigma typical habitat. The results of these studies, which integrated key stages of the life cycle of L. macrostigma, provide a scientific basis to develop conservation actions targeting not only the species, but also the restoration of Mediterranean temporary brackish ponds when using L. macrostigma as flagship species. Setting a hydroperiod from November to June is crucial in reducing egg desiccation risks and in favouring its larvae over that of other odonate species in the study region. This may require ecological engineering at the watershed scale, based on modelling the pond hydrology using meteorological and topographic data. Given climate change and the increase in the unpredictability of rainfall in the Mediterranean area, flooding some ponds artificially by irrigation may guarantee the sustainability of a number of perennial reproduction sites. Irrigation may also help control salinity levels and maintain high water levels during the flooding period in order to increase the species' fitness. The colonisation of newly created ponds by B. maritimus should be accelerated by transplanting tubers and ramets collected in situ. The existing vegetation cover should be maintained by implementing an adaptive grazing scheme by livestock linked to the target hydroperiod. Considering the past degradation of temporary brackish ponds, a network of these habitats should be restored or recreated using L. macrostigma as a flagship species. Their long-term monitoring would help identifying the reasons of the project's success or failure, and improve the design of future projects. In this regard, new research about reintroducing L. macrostigma in some restored ponds is necessary. This PhD provides an important case study about how to take into account threatened insects within biodiversity conservation strategies.