Balancing precision and risk: should multiple detection methods be analyzed separately in N-mixture models? (Q245575)