Butterfly Surveying In France

The national butterfly survey scheme in France began in 2005 and is called the Suivi Temporel des Rhopalocères de France (STERF). The aim of the project is to find out what species occur where and whether populations are increasing, decreasing or stable. Volunteers and professionals all over the country gather data from (April) May to August (September) and send it in to the French National Natural History Museum biodiversity recording unit. They do the number crunching and produce graphs showing the population trends and distribution maps. The data also feeds in to the annual European report on the state of grassland butterflies and the International Red Lists of Threatened Species.  Butterflies make good indicator species because they are particularly sensitive to habitat change. They are also easy to survey because they are generally easy to identify without capturing.

One of the squares I have been given to survey.
Each surveyor is allocated one or more random 2km² patches within a 10km radius of their home. They then decide on 5 - 15 transects within that square. A transect is a strip 5m wide and of varying length, depending on how rich the habitat is. For the French butterfly survey, you must be able to walk it in around 10 minutes, counting and identifying all the butterflies within the 5m strip as you go. In practice this means that the average transect length is 200m. Each transect must be of an homogenous habitat, or, if you are surveying woodland edge or crop margins, you must specify that the transect is one habitat on the left and another on the right. If a random square is more than 50% broadacre farming, then 3 of the transects must be along field edges. STERF provide a list of codes to use for a range of habitat types. You must only survey in good weather, as it is pointless going out in conditions when your target will be keeping out of sight. Surveyors may also choose sites to survey. These are generally nature reserves.

The average number of species observed per transect visit is 3.8.  The richest habitat in terms of numbers of species is suburban (although this varies considerably, with some extremely poor transects and some extremely good); chalk grassland; and pine woods (often associated with chalk grassland).  Good habitats, but not quite so rich are broadleaf forests; brown field, wasteland and fallow sites; other grasslands; and forest edges.  The poorest habitats are urban; agricultural land; and the fringes of non-forested habitat.

Polyommatus bellargus Adonis Blue (le Bel-Argus)
one of the target grassland species.
The most abundant butterflies in France are (in descending order of total individuals counted): Meadow Brown Maniola jurtina, Marbled White Melanargia galathea, Small / Large White Pieris rapae / napi, Gatekeeper Pyronia tithonus, Small Heath Coenonympha pamphilus, Chalkhill Blue Polyommatus coridon, Adonis Blue P. bellargus (and Painted Lady Vanessa cardui in 2009, catapulted to second place by the mass migration witnessed that year).  The most frequently observed (that is, recorded by the most surveyors in the most transects) on the other hand are:  Meadow Brown, Small Heath, Small White, other white Pierids, Speckled Wood Pararge aegeria, Common Blue Polyommatus icarus, Painted Lady and Marbled White. Chalkhill and Adonis Blues drop out of this second list because they are specialists of calcareous grassland, rather than generalists. 

At present the national results are skewed because of a predominance of recorders being located in suburban and / or northern France.  Also, some species (e.g. Whites Pieris spp, 'golden' Skippers Thymelicus spp) are difficult to identify in the field whilst surveying. The Marbled Whites are particularly gregarious, as are some Blues, Meadow Brown and Gatekeeper.  So are Lulworth Skipper Thymelicus acteon and Spanish Gatekeeper Pyronia bathseba, but difficulties with identification weaken the data. No clear trend is evident regarding emergence dates.

Of particular concern at the moment are butterflies that inhabit grasslands. It is clear that their numbers are plummeting in almost every case (a total population decrease in Europe of 70% in the last 10 years, and it is believed that the population was already greatly reduced when surveying first started). The causes are the increased intensification of farming (eg. broadacre farming, improved pasture) and conversely, the abandonment of small, steep sites which were once grazed, but are now considered uneconomic. On these latter sites, butterflies thrive for a couple of years, but with no grazing stock the grassland gets rank and is slowly replaced by scrub, and the butterflies disappear. Intensive farming on the other hand, uses almost every inch of soil for monocultured crops, with no space for the mosaic of wild flowers and wild grasses that the butterflies need. Improved pastures replace native flower rich grasslands with a much less complex mix of much lower habitat value for butterflies. Even haymeadows don't provide as high a quality of habitat as they used to, as in the modern farming calendar hay is cut earlier (and silage even earlier).

This is the middle of my other survey square.
As a result, many STERF surveyors are told that 3 of their transects in a square must be through agricultural land. The aim is not to concentrate on high quality nature reserves, but to get a broad and realistic picture of the situation throughout France. The clearer the problem becomes, the more chance the project has of influencing policy. Environmentalists believe that the Common Agricultural Policy must be reformed, to encourage and allow farmers to become ecological stewards rather than ruthless exploiters of the landscape.

What are the ideal dates to conduct your surveys?
This isn't a simple question and the answer is complex.

The dates of your surveys depend first and foremost on your availability and the weather. Doing a survey in bad weather results in incomplete data and even biases (for example, overcast weather will favour the recording of Satyrinae and those species that stay put on flowers such as some of the Blues. The same goes for windy weather, which disadvantages other species.

The survey protocol envisages that you will visit your survey sites once a month in May, June, July and August, and -- if your availability and the weather co-operate -- in April and September. These visits should be roughly a month apart (at least a fortnight apart, and at most, 6 weeks apart).

From a statistical point of view the most useful records are those which are taken year after year on more or less the same dates, allowing conclusions to be drawn about long term developments such as climatic change. This is therefore recommended as best practice. Likewise you should try to visit each transect at the same time of day each time, doing them in the same order each time, so the record is consistent and comparable in a statistically meaningful way. This provides a reproducible methodology to the surveys. However, be aware that visiting a transect too early in the day may artificially skew your records so that the site seems poorer than one visited in the middle of the day.

From an entomological point of view it would be ideal to do the surveys to coincide with the peaks of emergence of each of the species. However, since the peak period for each species will be different but the surveys are done for all species together, this rarely works out in practice. On the other hand, you can take into account whether the season is advanced or late relative to the norm and take that into consideration when choosing dates. It is generally agreed that the way around this is to survey every fortnight, so annual variances in seasonal abundance get picked up with less chance of misleading data. Surveying like this avoids the erroneous conclusion that a species is in regression, when in fact it is more that your survey dates coincide with an intergenerational gap.

It's all a compromise whilst trying to make the surveys reproducible and trying not to vary the dates too much from year to year, but still taking into account annual seasonal and weather variations.

If you are interested in monitoring butterflies and can commit to about an hour a month between May and September, please contact STERF. In 2010 there were only about 120 surveyors, checking 660 transects. Some départements don't have a single surveyor yet. You don't have to live in France (or even speak fluent French) but if you have a holiday house here and visit once a month during the summer you could participate. If you have any questions don't hesitate to get in touch with me or with STERF. I have translated the STERF abridged guidelines for surveying into English to encourage Anglophones to get involved. And if you see me wandering about the Chaumussay area or the Parc de Boussay with a clipboard, you will know what I'm up to.