SEARCH

SEARCH BY CITATION

Table S1. Effects of applying different colour separation layers (overall intensity, red, green and blue) on resulting grey-level characteristics of particles (bright spots resulting from firefly flashes) detected at two intensity thresholds. Data are mean ± SE, based on three bright spots from each of five selected sites (= 15).

Table S2. Effects of different colour separation layers on regression of particle analysis counts (y) against manual counts of firefly flashes (x) in image sequences taken at five sites. Data shown are the means of all sites, with ranges given in parentheses.

Figure S1. Part of the riverside habitat monitored for fireflies and equipment used for the technique. (a) Panoramic image (stitched) of a stretch of the Selangor River monitored for fireflies, showing separate image frames (cropped for overlap). (b) Flashes of light captured in an image of typical group display behaviour of Pteroptyx tener on Sonneratia caseolaris trees lining the riverbank (30 s exposure, ISO 3200). (c) Dual camera arrangement used for camera cross calibration, mounted on a tripod located at a fixed monitoring point on the riverbank opposite a stretch of riverside trees that is monitored for fireflies.

Figure S2. Effects of increasing firefly density in receptacles of fixed size on counts of flashes in images captured in a laboratory experiment.

Figure S3. Examples of single bright spots in images, showing variation in size and intensity when taken using different focal lengths at distances of approximately 60–120 m (exposure time 0.5 s). All images cropped to same ratio and enhanced by same magnitude: gamma = 2.00, contrast = 45 and brightness = −45.

Figure S4. Effects of lower detection threshold on the number of particles detected in particle analysis of vegetation on which the fireflies congregate (i.e. background sky and river water excluded). The threshold that gives a particle count that approximates manual counts (transverse line) is indicated by the vertical line.

Figure S5. Regression of total manual counts for each site and sampling period (nine camera locations × 4 months) against total counts predicted by three-point, zero intercept regression lines. Images with a high proportion of low intensity spots that fell below the detection threshold for particle analysis were excluded from particle analysis and their manual count values used in the predicted totals (6% of images and less than one third per site and sampling period).

Figure S6. Prediction of manual counts from particle analysis counts by fixed-intercept regression in the EOS 5DII. Different symbols represent different months. (a) Linear regression. (b) Multiple regression, where Ĉm is the best subsets regression, Ca = particle analysis count from the EOS 5DII, A = mean particle area (pixels), and Is = mean integral intensity.

Appendix S1. Reproducing camera location, orientation and settings.

Appendix S2. Cropping overlap in dark images.

Appendix S3. Dual camera set-up.

Please note: Neither the Editors nor Wiley-Blackwell are responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.

FilenameFormatSizeDescription
ICAD_157_sm_SupportingInformation-Rev.pdf162KSupporting info item

Please note: Neither the Editors nor Wiley Blackwell are responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.