FIGURE 2.

Boxplots showing the hatching success of spotted salamander embryos (Ambystoma maculatum) under eight different treatments in a lab experiment with two independent variables fully crossed: four predation levels (no, low, medium, and high) and two jelly‐layer levels (jelly layers removed or intact). Predation treatments consisted of varied numbers of wood frog tadpoles (Lithobates sylvaticus). Spotted salamander embryos are surrounded by thick jelly layers, which were gently removed to create “removed jelly‐layer” treatments. There was a significant interactive effect between the two independent variables. Within the removed jelly‐layer group, hatching success was highest under no predation followed by that under low predation, while no embryos successfully emerged under medium and high predation treatments (statistical differences indicated by red alphabets). Within the intact jelly‐layer group, hatching success under high predation was greater than those of any other predation treatments (statistical differences indicated by blue alphabets).