The member countries in the European Union is obliged to report the conservation status for species listed in Habitat directive every 6th ^{year. To gather information about population sizes and trends for the species in Sweden, the Swedish Environmental Protection Agency funds a monitoring namned "Biogeographical monitoring". The County Adminstrative Board of Östergötland has been assigned to coordinate the biogeographical monitoring of dragonflies (Odonata) and diving beetles (Dytiscidae) listed in the Habitat directive in Sweden. As a first step for establishing a monitoring programme for Ophiogomphus cecilia, a survey and method test was performed during the summer 2015. This report presents the result from the survey and suggests a strategy for monitoring the species.}

Ophiogomphus cecilia is restricted to the rivers (with tributaries) Råneälven, Kalixälven and Torneälven in the Sweden. These rivers are situated in the northeastern part of Sweden and are large rivers (often >100 meters width), unregulated and mainly surrounded by woodland, but near the coast to some extent also by cultivated land. The species is redlisted as "Near threatened" (NT) in Sweden due to small distribution area.

The aim of this study was to test exuviae-search as a survey method and act as a pilot study to receive more information about timescale, costs and practical considerations. If the survey turned out well, it could also serve as a first sample in the monitoring programme. The survey was performed by Tommy Karlsson, the County Administrative Board of Östergötland and Ulf Bjelke, the Species Species Information Centre during the period 20-24 July 2015. A large number of previously known localities for the species were surveyed as well as new, potential localities. The latter category also includes rivers from which the species not was known.

In total, 57 localities were surveyed och O. cecilia was found at 13 of these (Tab. 1 and Fig. 2). Exuviae were found at all localities, in total 50 exuviae. They were mainly found up to 30 cm height at straws of Carex sp. and within 50 cm from the shoreline. However, 15 exuviae were found 100-400 cm from the shoreline. This was probably due to a higher water level when the dragonflies emerged. Adults were found at three localities, in total four individuals, all newly emerged and close to the exuvia. In addition, an adult was found on the roadside during car driving and wings from an individual, probably predated by a bird, was found at one locality. All visited localities are shortly described and showed in a map in annex 1 (Bilaga 1). Localities where O. cecilia were found is also presented with a picture of the habitat.

We assess searching for exuviae as the most appropriate method for monitoring O. cecilia in Sweden. It is a quick method since the exuviae are easy to find and the risk of mixing up with other dragonflies species is very low. No other species of the family Gomphidae coexist with O. cecilia in Sweden, and the other species that you could find exuviae of at localities for O. cecilia (Aeshna sp. Somatochlora metallica, Cordulegaster boltonii) are easy to distinguish (Fig. 3-5). Furthermore, big advantages with surveying exuviae compared to adults is that it is not dependent of the weather and that records of exuviae gives information about the reproduction habitat for the species.

Searching for exuviae is a common survey method for O. cecilia in several other European countries, and the number of observed exuviae during a certain length of a river (e.g. 50 or 100 meters) is often used to estimate status and trend for a population. However, the trend for the Swedish population is assumed to be stable and annual variation, when the surveyed is performed in relation to emergence and variation between surveyors would probably influence the result more than possible trends with the economic resources available for monitoring today. An appropriate strategy for monitoring O. cecilia in Sweden is instead to recurring detect occurrence of the species at a selected number of localities. In table 2 and figure 6, 20 localities for monitoring are suggested.