Bumblebee Roadside Surveys: A Pilot Survey and Recommendations

43 %
57 %
Information about Bumblebee Roadside Surveys: A Pilot Survey and Recommendations

Published on August 18, 2009

Author: sdroege

Source: slideshare.net

Description

A pilot protocol for a roadside survey of bumblebees was tested along roads in Beltsville, Maryland, U.S.A.

Sam Droege Erika Tucker Prepared August 2009

Populations of some species of bumblebees in North America have declined to such an extent that in the last few years (current year = 2009) they remain undetected or appear to be absent from large reaches of their former range

Populations of some species of bumblebees in North America have declined to such an extent that in the last few years (current year = 2009) they remain undetected or appear to be absent from large reaches of their former range

There are ongoing efforts to inventory and re-inventory all bumblebee species from known historic localities There is some cataloging of museum bumblebee collections as reference points Bumblebee research workers are actively searching for and noting occurrences of now rare species No national or state level monitoring programs are currently in place

There are ongoing efforts to inventory and re-inventory all bumblebee species from known historic localities

There is some cataloging of museum bumblebee collections as reference points

Bumblebee research workers are actively searching for and noting occurrences of now rare species

No national or state level monitoring programs are currently in place

A meeting of bumblebee researchers from Canada, U.S. and Great Britain was held in June 2009 to discuss the status of bumblebees, current population research, and possible mechanisms and survey techniques for developing large regional/continental survey programs for the genus Bombus

A meeting of bumblebee researchers from Canada, U.S. and Great Britain was held in June 2009 to discuss the status of bumblebees, current population research, and possible mechanisms and survey techniques for developing large regional/continental survey programs for the genus Bombus

One possible survey methodology mentioned during the aforementioned meetings was a roadside survey of bumblebees Erika Tucker and Sam Droege thought this was a reasonable, but untested proposition, and embarked on some field tests…or, in reality, Erika did the field tests and Sam just made suggestions The results of those field tests are presented here and are meant as preliminary proof-of-concept and a starting place for discussions and further study

One possible survey methodology mentioned during the aforementioned meetings was a roadside survey of bumblebees

Erika Tucker and Sam Droege thought this was a reasonable, but untested proposition, and embarked on some field tests…or, in reality, Erika did the field tests and Sam just made suggestions

The results of those field tests are presented here and are meant as preliminary proof-of-concept and a starting place for discussions and further study

The model, analytical tools, and sampling frame for a roadside survey of the continent is already in place in the form of the North American Breeding Bird Survey (www.pwrc.usgs.gov/bbs) Roadside surveys only require access to public roads and should not require landowner permission Travel between sampling points is quick and a large number of survey points across a large landscape can be covered in a single day

The model, analytical tools, and sampling frame for a roadside survey of the continent is already in place in the form of the North American Breeding Bird Survey (www.pwrc.usgs.gov/bbs)

Roadside surveys only require access to public roads and should not require landowner permission

Travel between sampling points is quick and a large number of survey points across a large landscape can be covered in a single day

The technique for roadside surveys is attractive to volunteers The statistical, analytical, and sampling issues have been investigated many times for bird and amphibian groups Application to bumblebee surveys seems relatively straightforward After some preliminary tests we established a protocol and ran a set of survey points in Beltsville, Maryland, U.S.A.

The technique for roadside surveys is attractive to volunteers

The statistical, analytical, and sampling issues have been investigated many times for bird and amphibian groups

Application to bumblebee surveys seems relatively straightforward

After some preliminary tests we established a protocol and ran a set of survey points in Beltsville, Maryland, U.S.A.

Study Area: Roadsides within the USDA Beltsville Agriculture Research Center Habitats along roadsides were a mix of mature forest, farmland, and mown grass Edges of roads were often wide enough to support grass and some flowers even in predominantly wooded stops Surveys were conducted between noon and 3:30 p.m. on sunny days with low wind and warm weather

Study Area: Roadsides within the USDA Beltsville Agriculture Research Center

Habitats along roadsides were a mix of mature forest, farmland, and mown grass

Edges of roads were often wide enough to support grass and some flowers even in predominantly wooded stops

Surveys were conducted between noon and 3:30 p.m. on sunny days with low wind and warm weather

Stops were placed every half mile, with collections made at the nearest safe parking spot to the half mile point The count began when the observer left the car and continued for 10 minutes The observer walked to the nearest roadside location having or that looked like it might have bumblebees As many bumblebees as possible were collected and transferred to a centrifuge tube of soapy water for later processing* Lat, long, road name, and start of time were recorded and are available from Sam Droege (sdroege@usgs.gov) *Bumblebees were collected rather than visually identified due to the great difficulty both trained and untrained observers have in identifying bumblebees accurately on the wing

Stops were placed every half mile, with collections made at the nearest safe parking spot to the half mile point

The count began when the observer left the car and continued for 10 minutes

The observer walked to the nearest roadside location having or that looked like it might have bumblebees

As many bumblebees as possible were collected and transferred to a centrifuge tube of soapy water for later processing*

Lat, long, road name, and start of time were recorded and are available from Sam Droege (sdroege@usgs.gov)

*Bumblebees were collected rather than visually identified due to the great difficulty both trained and untrained observers have in identifying bumblebees accurately on the wing

Surveys occurred on 6 days between June 24 and July 7, 2009 38 sites were surveyed 14 (37%) of those sites recorded no bumblebees Lack of bees at a site appeared, anecdotally, to be largely associated with lack of floral resources

Surveys occurred on 6 days between June 24 and July 7, 2009

38 sites were surveyed

14 (37%) of those sites recorded no bumblebees

Lack of bees at a site appeared, anecdotally, to be largely associated with lack of floral resources

151 bumblebees were collected 3.97 bumblebees per site, dividing by all sites 6.29 bumblebees per site, on sites that had recorded bumblebees Minimum of 2 bumblebees, maximum of 17

151 bumblebees were collected

3.97 bumblebees per site, dividing by all sites

6.29 bumblebees per site, on sites that had recorded bumblebees

Minimum of 2 bumblebees, maximum of 17

Species Female Male Unknown Total B impatiens? 1     1 B. bimaculatus 9 40 49 B. fervidus 4 4 B. griseocollis 8 23 31 B. impatiens 64 64 B. perplexus   1 1 B. unknown   1 1 Grand Total 86 64 1 151

The next table presents species composition by site

The next table presents species composition by site

Site impatiens? bimaculatus fervidus griseocollis impatiens perplexus unknown nil Grand Total 1               1 1 2               1 1 3               1 1 4   3   2 7       12 5               1 1 6   13     2       15 7   7     2       9 8         3       3 9   1     1       2 10         4       4 11   1     3       4 12   3   1 4       8 13   2             2 14               1 1 15   3     3       6 16               1 1 17               1 1 18       7 1       8 19               1 1 20               1 1 21   1     4       5 22   4   1 1 1     7 23               1 1 24         9   1   10 25     1 1 1       3 26               1 1 27               1 1 28         5       5 29   1     4       5 30 1 1   1 1       4 31               1 1 32     2   3       5 33     1   5       6 34               1 1 35   7             7 36   2             2 37       1 1       2 38       17         17 Grand Total 1 49 4 31 64 1 1 14 165

Reasonable numbers and species of bumblebees were caught, and reflected bumblebees found in the region in what seemed to be appropriate proportions Stop methodology seemed straightforward We feel that the technique has merit and is worth further evaluations See next slides for potential negatives and some possible questions to test

Reasonable numbers and species of bumblebees were caught, and reflected bumblebees found in the region in what seemed to be appropriate proportions

Stop methodology seemed straightforward

We feel that the technique has merit and is worth further evaluations

See next slides for potential negatives and some possible questions to test

As the next few slides demonstrate*, different species have different peaks in numbers within a year Any survey will have to accommodate in both survey timing and analysis the fact that the date for peak population numbers for each species varies among species and likely varies across years More than one survey time period per year could help *Data came from a our database of all captures of bumblebees in the Mid-Atlantic region over the past 7 years

As the next few slides demonstrate*, different species have different peaks in numbers within a year

Any survey will have to accommodate in both survey timing and analysis the fact that the date for peak population numbers for each species varies among species and likely varies across years

More than one survey time period per year could help

*Data came from a our database of all captures of bumblebees in the Mid-Atlantic region over the past 7 years

 

 

 

 

 

 

 

Clearly there is a component of observer variability built into this type of survey Observers can vary by: Choice of where to survey at a site Ability to detect and catch bees Speed at which they can process specimens These issues are similar to those found in aural surveys of birds and amphibians and can be accounted for analytically to some extent, but decreasing such effects where possible would be important

Clearly there is a component of observer variability built into this type of survey

Observers can vary by:

Choice of where to survey at a site

Ability to detect and catch bees

Speed at which they can process specimens

These issues are similar to those found in aural surveys of birds and amphibians and can be accounted for analytically to some extent, but decreasing such effects where possible would be important

Studies of capture rates among observers would give us some idea about the magnitude of observer impacts on such a survey Testable Attributes of Observers: Speed in removing bumblebees from nets Proportion of successful captures per swing of net Variation in species composition and numbers by observers One possibility would be to keep all bumblebees in the net for the entire 10 minutes using a gloved hand to trap bees in the net between swings

Studies of capture rates among observers would give us some idea about the magnitude of observer impacts on such a survey

Testable Attributes of Observers:

Speed in removing bumblebees from nets

Proportion of successful captures per swing of net

Variation in species composition and numbers by observers

One possibility would be to keep all bumblebees in the net for the entire 10 minutes using a gloved hand to trap bees in the net between swings

Are inexpensive Are attractive to volunteers Require relatively little training (assuming all specimens go to a central sorting station and experts are available for identification) Likely would capture any major increases or decreases in those bumblebee populations that occur in road developed regions

Are inexpensive

Are attractive to volunteers

Require relatively little training (assuming all specimens go to a central sorting station and experts are available for identification)

Likely would capture any major increases or decreases in those bumblebee populations that occur in road developed regions

Roadside habitats are not a random reflection of the surrounding landscape Roads, particularly in mountainous and remote regions, are not random transects of the landscape Floral resources, on average, are likely to be higher near roadsides Trends in bumblebees found along roadsides would be both a reflection of roadside related processes and regional processes

Roadside habitats are not a random reflection of the surrounding landscape

Roads, particularly in mountainous and remote regions, are not random transects of the landscape

Floral resources, on average, are likely to be higher near roadsides

Trends in bumblebees found along roadsides would be both a reflection of roadside related processes and regional processes

Identification of bumblebees to species is not simple due to regional and individual variation in hair color Good preparation of alcohol preserved specimens is extremely important as hair color is hidden in matted specimens and bumblebees taken straight from alcohol do not make useful specimens Pinning thousands of bumblebees is time consuming and storage can be expensive

Identification of bumblebees to species is not simple due to regional and individual variation in hair color

Good preparation of alcohol preserved specimens is extremely important as hair color is hidden in matted specimens and bumblebees taken straight from alcohol do not make useful specimens

Pinning thousands of bumblebees is time consuming and storage can be expensive

Washing and Drying specimens can be largely automated (see slide shows on www.slideshare.net search on sdroege) Specimens can be dried, identified without pinning, databased, and then archived in Petri dishes Unusual or problematic specimens can be easily pinned after drying and sent to specialists

Washing and Drying specimens can be largely automated (see slide shows on www.slideshare.net search on sdroege)

Specimens can be dried, identified without pinning, databased, and then archived in Petri dishes

Unusual or problematic specimens can be easily pinned after drying and sent to specialists

Are the results presented here replicable elsewhere on the continent? (e.g., can you catch enough bumblebees along roads to make random surveys of roadsides worthwhile?) How variable are these counts? Knowing how stable these counts are gives us some perspective on how many samples will be needed to achieve good trend detection However, replication is to some degree affected by the first sample having removed some bees from the population

Are the results presented here replicable elsewhere on the continent? (e.g., can you catch enough bumblebees along roads to make random surveys of roadsides worthwhile?)

How variable are these counts?

Knowing how stable these counts are gives us some perspective on how many samples will be needed to achieve good trend detection

However, replication is to some degree affected by the first sample having removed some bees from the population

A comparison of on-road and off-road surveys would give some perspective on the relationship between the two landscapes A matrix of surveys with the roadside points going through the center of the grid would permit the modeling of both distance and habitat effects This design repeated over time would also permit the modeling of some of the variance components as well as stability of the relationships

A comparison of on-road and off-road surveys would give some perspective on the relationship between the two landscapes

A matrix of surveys with the roadside points going through the center of the grid would permit the modeling of both distance and habitat effects

This design repeated over time would also permit the modeling of some of the variance components as well as stability of the relationships

Because bumblebees are colonial there is at least some expectation (not really borne out within this small trial) that the number of bees caught at a site could be quite variable both as colony size varies and foraging from the nest varies with distance Consequently, traditional indices and measurements of population size used in vertebrate surveys are likely to yield difficult to interpret results or at least have wide confidence intervals…more in keeping with the natural history of bumblebees would be analytical techniques that investigate the number of sites occupied by a species rather than the population size of a species

Because bumblebees are colonial there is at least some expectation (not really borne out within this small trial) that the number of bees caught at a site could be quite variable both as colony size varies and foraging from the nest varies with distance

Consequently, traditional indices and measurements of population size used in vertebrate surveys are likely to yield difficult to interpret results or at least have wide confidence intervals…more in keeping with the natural history of bumblebees would be analytical techniques that investigate the number of sites occupied by a species rather than the population size of a species

Sam and Erika would like this dataset to be used by as many people as possible, please contact Sam Droege ( [email_address] ) for a copy We would also be interested in helping other groups establish tests or trials of this sort of survey…email or give us a call Work Phone: 301-497-5840

Sam and Erika would like this dataset to be used by as many people as possible, please contact Sam Droege ( [email_address] ) for a copy

We would also be interested in helping other groups establish tests or trials of this sort of survey…email or give us a call

Work Phone: 301-497-5840

Add a comment

Related pages

Bumble Bee Megatransect - Hands on the Land

During the Bumble Bee Megatransect, ... Center your swing on the Bumble Bee ... Bumblebee Roadside Surveys: A Pilot Survey and Recommendations by Sam Droege.
Read more

Bumblebee | LinkedIn

Bumblebee Roadside Surveys: A Pilot Survey and Recommendations. 2,944 Views. xiby. The Buzzard, The Bat, And The Bumblebee. 1,493 Views. salmanm92.
Read more

Pilot Tests and Focus Groups | Pew Research Center

Pilot Tests and Focus Groups. ... In addition, pilot tests for many surveys test the full implementation procedures (e.g., contact letters, ...
Read more

Best Practices - Survey Analysis: Reporting Your Findings ...

Survey Analysis: Reporting Your Findings. The final step of a survey project is reporting the findings. ... Conclusion and Recommendations ;
Read more

Presenting Survey Results - Drawing Conclusions

Presenting the conclusion and recommendations includes reviewing the survey goals and objectives and ... Presenting Survey Results ... 6.1 Pilot Survey;
Read more

Surveys 101: Survey Best Practices & Guidelines | SurveyMonkey

Surveys 101. Design surveys, ... Pilot Test Tips: How to Verify the Design and Settings; How to analyze a survey. Success! You’ve got survey results.
Read more

SurveyMonkey: Free online survey software & questionnaire tool

SurveyMonkey provides free online questionnaire and survey software. Sign In; Help; SurveyMonkey ...
Read more

Roadside | LinkedIn

Director at Midwest Roadside Safety Facility, Research Associate Professor at Midwest Roadside Safety Facility Past
Read more