• Acoustic interference;
  • community structure;
  • frogs;
  • null models;
  • signal evolution

Abstract Competitive partitioning of ‘community’ signal space has long been suggested to underlie diversification of mating signals. Selection or competitive exclusion is expected to reduce overlap of signals, minimizing destructive interference or reducing mismating. We used null models backed by simulation of type I and II error rates to test for evidence of structuring within 11 frog advertisement call assemblages. Within three assemblages, we found significant over-dispersion and regularity-of-spacing in dominant frequency and in pulse rate, consistent with a signal interference hypothesis and signal confusion hypothesis, respectively. Observed partitioning could represent signal evolution or could result from selection on assemblage composition. Most assemblages showed no acoustic partitioning possibly because: (i) partitioning is more readily apparent in female preference, calling times or sites, rather than call attributes; (ii) assemblages have not yet accommodated recently arrived species, or are compositionally unstable so that acoustic accommodation cannot occur; and (iii) evidence of partitioning is only likely where the acoustic space is densely packed.