9. Sample nature without fooling yourself
Use quadrats, transects, plots, traps, point counts, mark-recapture, and repeated measurements while controlling bias. You will choose replication, randomization, controls, and timing that match the ecological question.