際際滷

際際滷Share a Scribd company logo

Swaay butterfly monitoring analysis

Jacqueline Loos
Jacqueline Loos

Volunteers across Europe monitor butterfly populations by regularly counting butterflies along fixed transect routes. Over 200,000 records are collected annually in the Netherlands alone. Trend data from thousands of transects are analyzed to produce indicators of butterfly population changes over time. The data meet scientific standards and can measure progress towards conservation targets. Monitoring began in the UK in 1976 and now involves over 2,500 transects in 14 countries. The results are used to study relationships between butterflies and their habitat needs and environmental factors.

1 of 46
Downloaded 11 times
Butterfly Monitoring: analysis and scientific use of data Chris van Swaay, De Vlinderstichting / Dutch Butterfly Conservation Butterfly Conservation Europe (BCE) Statistics Netherlands (CBS) Thousands of volunteers Adapted by Martin Wiemers (UFZ, BCE)
What is butterfly monitoring? Collect information on the changes in butterfly abundance We have to follow a protocol to detect real trends Fieldwork Basis: samples Regular counts Fixed method In the Netherlands 400 transects generate 200000 records per year
Products 10 100 1000 1992 1994 1996 1998 2000 2002 2004 Woodland generalists Woodland specialists Species trends Indicators 10 100 1000 1992 1994 1996 1998 2000 2002 2004 2006 2008 GeelsprietdikkopjeThymelicus sylvestris
Criteria for indicators Scientific sound method Sensitive Affordable monitoring, available and routinely collected data Spatial and temporal coverage of data Measure progress towards target Policy relevance Broad acceptance
The start Ernie Pollard started the first BMS in the UK in 1976
1976: start of the first Butterfly Monitoring Scheme in the UK Well founded by many scientific papers Now at least 2500 transects in 14 countries Every year our European volunteers count once around the world (40.000 km)! Butterfly Monitoring available and routinely collected
Butterfly Monitoring Spatial coverage New countries join in every year Most of them done every year
Butterfly Monitoring Temporal coverage 0 2 4 6 8 10 12 14 16 18 20 1975 1980 1985 1990 1995 2000 2005 2010 NumberofBMS
From transects to European indicator Location of the transects Quality of the observer Quality of the observations Validation of the observations Calculating trends Building indicators
From transects to European indicator Location of the transects Quality of the observer Quality of the observations Validation of the observations Calculating trends Building indicators
Choice of locations Free choice of transects (e.g. in the UK, Netherlands, Germany) Pro: appealing to volunteers, easy to keep them motivated, rare species included Con: data is biased (but can be corrected by weighting) (Partly) random (e.g. France) Pro: less bias Con: sometimes transects on unattractive sites, no trends of rare species (often the ones with high conservation value) Regular grid (e.g. Switzerland) Pro: almost no bias Con: hard to achieve (only on professional basis); no trends of rare species (often the ones with high conservation value)
Swaay butterfly monitoring analysis
Swaay butterfly monitoring analysis
From transects to European indicator Location of the transects Quality of the observer Quality of the observations Validation of the observations Calculating trends Building indicators
Basic idea We realise we cant count all butterflies But by taking samples we can estimate trends As a consequence we dont know the population size But we can calculate changes in the population size efficiently With random or grid sampling transects are properly distributed over the country But in many countries recorders have a free choice Solution: weighting
Why weighting? Not all species are equally distributed over the country Not all transects are equally distributed over the country. In the Netherlands especially the dunes are oversampled, agricultural areas in the clay and peat regions are undersampled.
Weighting by Dutch physical geographic region and main habitat type Habitat types: Woodland Heathland Agriculture Open dunes Urban Moorland
Distribution of the population over the strata The distribution of each species per stratum is calculated. For example: Hipparchia semele Dunes - mainland Dunes - Waddensea Heathland - north Heathland - centre Heathland - south
Distribution of the transects over the strata Dunes- mainland Dunes- Waddensea Heathland- north Heathland- centre Heathland- south distribution
Big difference between weighted and unweighted indexes 1 10 100 1992 1994 1996 1998 2000 2002 2004 2006 Weighted Unweighted
The trend in the dunes is different from the trend on the heathlands distribution transects 1 10 100 1000 1990 1993 1996 1999 2002 2005 Heathland Coastal dunes
From transects to European indicator Location of the transects Quality of the observer Quality of the observations Validation of the observations Calculating trends Building indicators
Grassland Butterfly Indicator: main habitat for European butterflies For 57% of the species, grasslands are their main habitat. Grassland; 280 Woodland and scrub; 153 Heath, bog and fen; 25 others; 31
17 species make the indicator 7 widespread species: Ochlodes sylvanus, Anthocharis cardamines, Lycaena phlaeas, Polyommatus icarus, Lasiommata megera, Coenonympha pamphilus Maniola jurtina 10 specialist species: Erynnis tages, Thymelicus acteon, Spialia sertorius, Cupido minimus, Maculinea arion, Maculinea nausithous, Polyommatus bellargus, Cyaniris semiargus, Polyommatus coridon Euphydryas aurinia
From national trends to a European trend 0 25 50 75 100 125 150 175 1990 1994 1998 2002 2006 2010 Index(firstyear=100) France The Netherlands Spain - Catalonia United Kingdom 0 20 40 60 80 100 120 1990 1994 1998 2002 2006 2010 Index(1990=100) + 9 other countries
European species trends 0 20 40 60 80 100 120 1990 1994 1998 2002 2006 2010 Index(1990=100) 0 20 40 60 80 100 120 140 160 1990 1994 1998 2002 2006 2010 Index(1990=100) 0 50 100 150 200 250 1990 1994 1998 2002 2006 2010 Index(1990=100)
European trends Species Trend in Europe Trend in EU Phengaris nausithous decline decline Erynnis tages decline decline Lasiommata megera decline decline Lycaena phlaeas decline decline Thymelicus acteon decline decline Ochlodes sylvanus decline decline Coenonympha pamphilus decline decline Cupido minimus decline decline Anthocharis cardamines decline stable Polyommatus icarus decline stable Maniola jurtina stable stable Polyommatus coridon stable stable Cyaniris semiargus uncertain stable Polyommatus bellargus uncertain uncertain Spialia sertorius uncertain uncertain Euphydryas aurinia uncertain uncertain Phengaris arion uncertain uncertain
European Grassland Butterfly Indicator 0 20 40 60 80 100 120 140 1990 1994 1998 2002 2006 2010 Butterfly Conservation Europe / Statistics Netherlands
Main drivers 1. Intensification
Main drivers 2. Abandonment
Swaay butterfly monitoring analysis
Swaay butterfly monitoring analysis
Relationships between butterflies and environmental indicators Plants: Ellenberg values Like plants some species have a preference for rich or wet situations, others for poor or dry places Butterfly monitoring gave us info on the presence of butterflies We made vegetation surveys at transects and calculated the average Ellenberg value for Nutrient, acidity and moisture.
Field data of occurrence of a species vs soil pH
Logistic regression: sigmoid relationship
Logistic regression: gaussian relationship
0 10 20 30 40 50 60 0 1 2 3 4 5 6 7 8 9 Probabilityofoccurrence(%) Acidity-value (Ellenberg scale) Pmax=37% Tolerance=2.3 Optimum=5.0 Response curve of Araschnia levana for Ellenbergs acidity- value, showing the Optimum (U), the maximum probability of occurrence (Pmax) and the Tolerance (T).
0 20 40 60 80 0 2 4 6 8 10 Prob/Freqofoccurrence(%) Nutrient-value (Ellenberg scale) (a) observed expected 0 20 40 60 80 0 2 4 6 8 10Prob/Freqofoccurrence(%) Nutrient-value (Ellenberg scale) (b) observed expected Two examples of response curves of butterflies on Ellenbergs nutrient value, showing the calculated logistic regression model (expected) and the observed frequency of the species in the relev辿s falling in nutrient value classes with a width of 0.25: (a) the unimodal (Gaussian) response of Thymelicus lineola and (b) the sigmoidal response of Ochlodes sylvanus.
0 20 40 60 80 100 0 5 10 Probabilityofoccurrence(%) Nutrient-value (Ellenberg-scale) (a) M. alcon A. levana C. selene P. icarus P. rapae
0 20 40 60 0 5 10 Probabilityofoccurrence(%) Acidity-value (Ellenberg-scale) (b) C. tullia I. io E. tages A. agestis C. pamphilus
0 20 40 60 0 5 10 Probabilityofoccurrence(%) Moisture-value (Ellenberg-scale) (c) V. optilete M. alcon E. tages I. lathonia L. megera
Use butterfly monitoring results for site information 0 1 2 3 4 5 6 7 8 1990 1995 2000 2005 2010 Moisture/Nitrogenvalue Moisture index Nitrogen index Luttenbergerven
Multiple relationships Give the relationship between the three indicators When more than one is significant, we get a multi- dimensional plane or surface For a site it can give an insight in the effects of a changing environment on butterflies
Calculate national annual nitrogen index for butterflies 5.7 5.8 5.9 6 6.1 6.2 6.3 6.4 6.5 1990 1995 2000 2005 2010 CNI
Calculate national annual nitrogen index for butterflies y = 0.0131x - 20.205 5.7 5.8 5.9 6 6.1 6.2 6.3 6.4 6.5 1990 1995 2000 2005 2010 CNI
De Vlinderstichting Dutch Butterfly Conservation www.vlinderstichting.nl Statistics Netherlands www.cbs.nl

More Related Content

Swaay butterfly monitoring analysis

  • 1. Butterfly Monitoring: analysis and scientific use of data Chris van Swaay, De Vlinderstichting / Dutch Butterfly Conservation Butterfly Conservation Europe (BCE) Statistics Netherlands (CBS) Thousands of volunteers Adapted by Martin Wiemers (UFZ, BCE)
  • 2. What is butterfly monitoring? Collect information on the changes in butterfly abundance We have to follow a protocol to detect real trends Fieldwork Basis: samples Regular counts Fixed method In the Netherlands 400 transects generate 200000 records per year
  • 3. Products 10 100 1000 1992 1994 1996 1998 2000 2002 2004 Woodland generalists Woodland specialists Species trends Indicators 10 100 1000 1992 1994 1996 1998 2000 2002 2004 2006 2008 GeelsprietdikkopjeThymelicus sylvestris
  • 4. Criteria for indicators Scientific sound method Sensitive Affordable monitoring, available and routinely collected data Spatial and temporal coverage of data Measure progress towards target Policy relevance Broad acceptance
  • 5. The start Ernie Pollard started the first BMS in the UK in 1976
  • 6. 1976: start of the first Butterfly Monitoring Scheme in the UK Well founded by many scientific papers Now at least 2500 transects in 14 countries Every year our European volunteers count once around the world (40.000 km)! Butterfly Monitoring available and routinely collected
  • 7. Butterfly Monitoring Spatial coverage New countries join in every year Most of them done every year
  • 8. Butterfly Monitoring Temporal coverage 0 2 4 6 8 10 12 14 16 18 20 1975 1980 1985 1990 1995 2000 2005 2010 NumberofBMS
  • 9. From transects to European indicator Location of the transects Quality of the observer Quality of the observations Validation of the observations Calculating trends Building indicators
  • 10. From transects to European indicator Location of the transects Quality of the observer Quality of the observations Validation of the observations Calculating trends Building indicators
  • 11. Choice of locations Free choice of transects (e.g. in the UK, Netherlands, Germany) Pro: appealing to volunteers, easy to keep them motivated, rare species included Con: data is biased (but can be corrected by weighting) (Partly) random (e.g. France) Pro: less bias Con: sometimes transects on unattractive sites, no trends of rare species (often the ones with high conservation value) Regular grid (e.g. Switzerland) Pro: almost no bias Con: hard to achieve (only on professional basis); no trends of rare species (often the ones with high conservation value)
  • 14. From transects to European indicator Location of the transects Quality of the observer Quality of the observations Validation of the observations Calculating trends Building indicators
  • 15. Basic idea We realise we cant count all butterflies But by taking samples we can estimate trends As a consequence we dont know the population size But we can calculate changes in the population size efficiently With random or grid sampling transects are properly distributed over the country But in many countries recorders have a free choice Solution: weighting
  • 16. Why weighting? Not all species are equally distributed over the country Not all transects are equally distributed over the country. In the Netherlands especially the dunes are oversampled, agricultural areas in the clay and peat regions are undersampled.
  • 17. Weighting by Dutch physical geographic region and main habitat type Habitat types: Woodland Heathland Agriculture Open dunes Urban Moorland
  • 18. Distribution of the population over the strata The distribution of each species per stratum is calculated. For example: Hipparchia semele Dunes - mainland Dunes - Waddensea Heathland - north Heathland - centre Heathland - south
  • 19. Distribution of the transects over the strata Dunes- mainland Dunes- Waddensea Heathland- north Heathland- centre Heathland- south distribution
  • 20. Big difference between weighted and unweighted indexes 1 10 100 1992 1994 1996 1998 2000 2002 2004 2006 Weighted Unweighted
  • 21. The trend in the dunes is different from the trend on the heathlands distribution transects 1 10 100 1000 1990 1993 1996 1999 2002 2005 Heathland Coastal dunes
  • 22. From transects to European indicator Location of the transects Quality of the observer Quality of the observations Validation of the observations Calculating trends Building indicators
  • 23. Grassland Butterfly Indicator: main habitat for European butterflies For 57% of the species, grasslands are their main habitat. Grassland; 280 Woodland and scrub; 153 Heath, bog and fen; 25 others; 31
  • 24. 17 species make the indicator 7 widespread species: Ochlodes sylvanus, Anthocharis cardamines, Lycaena phlaeas, Polyommatus icarus, Lasiommata megera, Coenonympha pamphilus Maniola jurtina 10 specialist species: Erynnis tages, Thymelicus acteon, Spialia sertorius, Cupido minimus, Maculinea arion, Maculinea nausithous, Polyommatus bellargus, Cyaniris semiargus, Polyommatus coridon Euphydryas aurinia
  • 25. From national trends to a European trend 0 25 50 75 100 125 150 175 1990 1994 1998 2002 2006 2010 Index(firstyear=100) France The Netherlands Spain - Catalonia United Kingdom 0 20 40 60 80 100 120 1990 1994 1998 2002 2006 2010 Index(1990=100) + 9 other countries
  • 26. European species trends 0 20 40 60 80 100 120 1990 1994 1998 2002 2006 2010 Index(1990=100) 0 20 40 60 80 100 120 140 160 1990 1994 1998 2002 2006 2010 Index(1990=100) 0 50 100 150 200 250 1990 1994 1998 2002 2006 2010 Index(1990=100)
  • 27. European trends Species Trend in Europe Trend in EU Phengaris nausithous decline decline Erynnis tages decline decline Lasiommata megera decline decline Lycaena phlaeas decline decline Thymelicus acteon decline decline Ochlodes sylvanus decline decline Coenonympha pamphilus decline decline Cupido minimus decline decline Anthocharis cardamines decline stable Polyommatus icarus decline stable Maniola jurtina stable stable Polyommatus coridon stable stable Cyaniris semiargus uncertain stable Polyommatus bellargus uncertain uncertain Spialia sertorius uncertain uncertain Euphydryas aurinia uncertain uncertain Phengaris arion uncertain uncertain
  • 28. European Grassland Butterfly Indicator 0 20 40 60 80 100 120 140 1990 1994 1998 2002 2006 2010 Butterfly Conservation Europe / Statistics Netherlands
  • 33. Relationships between butterflies and environmental indicators Plants: Ellenberg values Like plants some species have a preference for rich or wet situations, others for poor or dry places Butterfly monitoring gave us info on the presence of butterflies We made vegetation surveys at transects and calculated the average Ellenberg value for Nutrient, acidity and moisture.
  • 34. Field data of occurrence of a species vs soil pH
  • 37. 0 10 20 30 40 50 60 0 1 2 3 4 5 6 7 8 9 Probabilityofoccurrence(%) Acidity-value (Ellenberg scale) Pmax=37% Tolerance=2.3 Optimum=5.0 Response curve of Araschnia levana for Ellenbergs acidity- value, showing the Optimum (U), the maximum probability of occurrence (Pmax) and the Tolerance (T).
  • 38. 0 20 40 60 80 0 2 4 6 8 10 Prob/Freqofoccurrence(%) Nutrient-value (Ellenberg scale) (a) observed expected 0 20 40 60 80 0 2 4 6 8 10Prob/Freqofoccurrence(%) Nutrient-value (Ellenberg scale) (b) observed expected Two examples of response curves of butterflies on Ellenbergs nutrient value, showing the calculated logistic regression model (expected) and the observed frequency of the species in the relev辿s falling in nutrient value classes with a width of 0.25: (a) the unimodal (Gaussian) response of Thymelicus lineola and (b) the sigmoidal response of Ochlodes sylvanus.
  • 39. 0 20 40 60 80 100 0 5 10 Probabilityofoccurrence(%) Nutrient-value (Ellenberg-scale) (a) M. alcon A. levana C. selene P. icarus P. rapae
  • 40. 0 20 40 60 0 5 10 Probabilityofoccurrence(%) Acidity-value (Ellenberg-scale) (b) C. tullia I. io E. tages A. agestis C. pamphilus
  • 41. 0 20 40 60 0 5 10 Probabilityofoccurrence(%) Moisture-value (Ellenberg-scale) (c) V. optilete M. alcon E. tages I. lathonia L. megera
  • 42. Use butterfly monitoring results for site information 0 1 2 3 4 5 6 7 8 1990 1995 2000 2005 2010 Moisture/Nitrogenvalue Moisture index Nitrogen index Luttenbergerven
  • 43. Multiple relationships Give the relationship between the three indicators When more than one is significant, we get a multi- dimensional plane or surface For a site it can give an insight in the effects of a changing environment on butterflies
  • 44. Calculate national annual nitrogen index for butterflies 5.7 5.8 5.9 6 6.1 6.2 6.3 6.4 6.5 1990 1995 2000 2005 2010 CNI
  • 45. Calculate national annual nitrogen index for butterflies y = 0.0131x - 20.205 5.7 5.8 5.9 6 6.1 6.2 6.3 6.4 6.5 1990 1995 2000 2005 2010 CNI
  • 46. De Vlinderstichting Dutch Butterfly Conservation www.vlinderstichting.nl Statistics Netherlands www.cbs.nl