Fig 1.
Antlion immobility duration (s) against the sequential number of the immobility period since predator contact for each of the three substrates: Paper, Shallow sand, Deep sand.
Both axes are on a log scale; blue “violins”: mirror density plots with horizontal lines representing the median, upper and lower quartile, red line: overall relationship (predicted fixed effects from the best LMM), grey lines: relationships for individual antlions (predicted random effects from the best LMM).
Table 1.
Post-hoc pair-wise comparisons between slopes for the best LMM for immobility duration (s).
Fig 2.
Antlion movement duration (s) against the sequential number of the movement period since predator contact for each of the three substrates.
Both axes are on a log scale; blue “violins”: mirror density plots with horizontal lines representing the median, upper and lower quartile, red line: overall relationship (predicted fixed effects from the best LMM), grey lines: relationships for individual antlions (predicted random effects from the best LMM).
Table 2.
Post-hoc pair-wise comparisons between slopes for the best LMM for movement duration (s).
Fig 3.
Empirical complementary cumulative distribution for immobility duration (s) for each of the three substrates fitted with each of the models: (a) Exponential, (b) Power law (c) Log-normal; grey square: data for Paper, blue circle: data for Shallow sand, red triangle: data for Deep sand; black line: fit for Paper; blue line: fit for Shallow sand; red line: fit for Deep sand. Both axes are on a log scale. Here each model is fitted with its own best-fit lower-bound (xmin) value to show how much of the data were discarded in each case but for the pair-wise comparisons with Vuong’s test, xmin was the same for the members of each pair of compared models.
Table 3.
Complementary cumulative distribution for immobility duration (s): Goodness-of-fit tests comparing the exponential with alternatives.
Fig 4.
Empirical complementary cumulative distribution for movement duration (s) for each of the three substrates fitted with each of the models: (a) Exponential, (b) Power law (c) Log-normal; grey square: data for Paper, blue circle: data for Shallow sand, red triangle: data for Deep sand; black line: fit for Paper; blue line: fit for Shallow sand; red line: fit for Deep sand. Both axes are on a log scale. Here each model is fitted with its own best-fit lower-bound (xmin) value to show how much of the data were discarded in each case but for the pair-wise comparisons with Vuong’s test, xmin was the same for the members of each pair of compared models.
Table 4.
Complementary cumulative distribution for movement duration (s): Goodness-of-fit tests comparing the exponential with alternatives.
Fig 5.
Path (mm) and instantaneous speed (mms-1) over time for one of the antlions on each of the three substrates.
(a)-(b) Antlion 19 on Paper, (c)-(d) Antlion 21 on Shallow sand and (e)-(f) Antlion 6 on Deep sand; dotted black line: path from the spot in the centre of the arena where the antlion larva was dropped (Paper: light blue circle; Shallow sand: light green circle; Deep sand: light yellow circle) to the wall (Paper: red circle) or end of up to 16 moves (Shallow sand: pink circle; Deep sand: salmon pink circle) with superimposed stops (Paper: dark blue circle; Shallow sand: dark green circle; Deep sand: dark yellow circle) and displacement segments between stops (Paper: light blue line; Shallow sand: light green line; Deep sand: light yellow line); black line: instantaneous speed over time (s) calculated from the image-analysis track, cyan highlight: the superimposed observed movement periods up to reaching the wall or end of up to 16 moves.
Fig 6.
Start-to-finish displacement (mm) for antlions on each of the three substrates: Paper (N = 22), Shallow sand (N = 8), Deep sand (N = 8).
The horizontal interrupted red line at 159 mm is half the diagonal of the experimental arena and represents the maximum possible start-to-finish displacement given that the antlions were delivered in the arena centre. Box plots show 25% and 75% quartiles (boxes), medians (lines in the boxes), outermost values within the range of 1.5 times the respective quartiles (whiskers), outliers (grey circles), and individual measurements (pink circles). A small amount of jitter was applied to the circles to minimise any occlusion. The two outliers represent antlions 10 and 13, that hardly moved throughout the experiment on Paper.
Fig 7.
Dispersion rate of antlion larvae on the three substrates Paper (N = 22), Shallow sand (N = 8), Deep sand (N = 8).
(a)-(b) mean squared displacement (MSD, mm2) and mean instantaneous speed (MIS, mms-1) over time since the end of PCI on Paper; (c)-(d) MSD and MIS on Shallow sand; (e)-(f) MSD and MIS on Deep sand; filled black circles: mean value for each of the 14 log-binned time intervals (see Methods for calculation details), red triangle: upper 95% CI limit of the mean for the time interval, solid blue line: line of best-fit from a segmented linear regression model with one break point, with estimates for the two slopes and their 95% CIs in brackets, red point and red horizontal line: estimate and 95% CI for the break point, interrupted grey line: the line of unity (y = x), solid yellow line: line of best fit from a simple linear regression with estimate and 95% CI for the slope; the number of mean values for Shallow sand is 13 and the number of upper 95% CI limits of the mean for both Shallow and Deep sand is 12 rather than 14, because there were fewer than two entries for the last two intervals for both sand substrates (see Methods for details of time binning).