Figure 3. Baffle design efficiency and optimization behaviour.

(A) FEA predicted SRE representing the effect of different leaf and hole size combinations (cartoons not to scale), overlaid with data from cricket-made baffles (black dots). Leaf and hole lengths are used as a proxy for size. In nature, as in our models, leaves retain their aspect ratio at different sizes, and length is an adequate measure. Vertical stippled lines mark the ranges of small and large leaf sizes offered to the crickets during choice experiments. Side graphs show SRE at average leaf size (115 mm) and average hole size, which is also the optimal size (10.1 mm). The red cross marks the optimal baffle. (B) Blue bars depict the proportion of males that made baffles on different leaf sizes in a no-choice experiment. The width of the bars depict the size range of the leaves in that size class. Grey bars depict the distribution of natural leaf sizes by depicting the proportion of leaves that fall into different size classes (N = 570 leaves). The two red lines depict the SRE associated with baffled calling (solid line), and with unbaffled calling from the leaf edge (wings parallel to the leaf surface, stippled line) at different leaf sizes. (C) Baffles created on real leaves were normalized to the average large leaf (length = 115 mm) and overlaid onto a map showing the percent SRE at different hole positions in relation to the maximum. Percent maximum SRE along horizontal and vertical transects are shown on side graphs. One male made two holes, first an eccentric, suboptimal one, followed by another at a more efficient location (white arrow). (D) Distributions of maximum SRE achievable on small and large leaves offered to the males, with SRE distributions of real baffles made by males (n = 15). Box and whisker plot depicts median (red line), the 25^th and 75^th percentile (blue box) and the range (whiskers).

Figure 3—figure supplement 1. Baffle making behaviour and its natural context (A) Box-plot showing the distributions of leaf size used by wild non-bafflers and by baffling males in the field.

The leaves used by both sets of males for calling were larger than the most commonly found leaves in the wild. The leaves used by the baffling males were, however, bigger than those used by non-baffling males, suggesting that males actively select larger leaves for making baffles (Mann-Whitney U test; Length: W = 668, p<0.005, numbers in boxes are the sample sizes). (B) All the baffle holes made on large H. suaveolens leaves by the male tree crickets (scale bars: 10 mm). Note that the holes are always placed between large leaf veins.