Long-term genetic monitoring of a riverine dragonfly, Orthetrum coerulescens (Odonata: Libellulidae]: Direct anthropogenic impact versus climate change effects

Modern conservationists call for long term genetic monitoring datasets to evaluate and understand the impact of human activities on natural ecosystems and species on a global but also local scale. However, long-term monitoring datasets are still rare but in high demand to correctly identify, evaluate and respond to environmental changes. In the presented study, a population of the riverine dragonfly, Orthetrum coerulescens (Odonata: Libellulidae), was monitored over a time period from 1989 to 2013. Study site was an artificial irrigation ditch in one of the last European stone steppes and "nature heritage", the Crau in Southern France. This artificial riverine habitat has an unusual high diversity of odonate species, prominent indicators for evaluating freshwater habitats. A clearing of the canal and destruction of the bank vegetation in 1996 was assumed to have great negative impact on the odonate larval and adult populations. Two mitochondrial markers (CO1 & ND1) and a panel of nuclear microsatellite loci were used to assess the genetic diversity. Over time they revealed a dramatic decline in diversity parameters between the years 2004 and 2007, however not between 1996 and 1997. From 2007 onwards the population shows a stabilizing trend but has not reached the amount of genetic variation found at the beginning of this survey. This decline cannot be referred to the clearing of the canal or any other direct anthropogenic impact. Instead, it is most likely that the populations' decay was due to by extreme weather conditions during the specific years. A severe drought was recorded for the summer months of these years, leading to reduced water levels in the canal causing also other water parameters to change, and therefore impacting temperature sensitive riverine habitat specialists like the O. coerulescens in a significant way. The data provide important insights into population genetic dynamics and metrics not always congruent with traditional monitoring data (e.g. abundance); a fact that should be regarded with caution when management plans for developed landscapes are designed.