The study was conducted in the northern lagoon of Moorea, French Polynesia (17 30’S, 149 50’W) at the Gump Biological Research Station between February and August 2007, using a grid of 28 live-coral patch reefs in water 2 to 4 meters deep.

Experimental Design and Execution
To test for priority effects, we experimentally manipulated the sequence and timing of arrival of T. hardwicke and T. quinquevittatum onto reefs. We used recent settlers (13.8 mm SL; SD = 2.3; approximately two weeks post-settlement) captured from reefs roughly 4 km from the study site. All captured fish were held in tanks with running seawater for 6–12 h, then individually tagged with different colors of Visible Implant Elastomer forward of the caudal peduncle. After tagging, we returned fish to aerated aquaria for 6-12 hr before measuring them to the nearest 0.1 mm SL, and deploying them in the field. We used newly collected and tagged fish in two experiments. Experiment 1 examined interspecific priority effects between T. hardwicke and T. quinquevittatum. Experiment 2 examined intraspecific priority effects within T. hardwicke.

Experiment 1: Interspecific Priority Effects between T. hardwicke and T. quinquevittatum
We experimentally manipulated the sequence and timing of the arrival of T. hardwicke and T. quinquevittatum settlers; one portion of this experiment was designed to quantify effects of T. hardwicke (the prior resident) on T. quinquevittatum (the focal species); the other was designed to test the effects of T. quinquevittatum (as the prior resident) on T. hardwicke (as the focal species). For both cases, we simulated settlement pulses by introducing three fish of the focal species onto reefs where three tagged fish of the prior resident were either absent; had been introduced 12-days earlier than, 5-days earlier than, or simultaneously with (0-days) the focal species. All fish were collected and tagged in the 24 hours preceding their deployment in the field. We ran the experiment in two temporal blocks (11 to 30 April and 2 to 21 May), randomly assigning treatments to reefs, and fish to treatments, yielding eight replicates (four in each temporal block) for each of our seven treatments: 1) T. hardwicke without T. quinquevittatum; 2) T. hardwicke and T. quinquevittatum introduced simultaneously; 3) T. hardwicke with T. quinquevittatum introduced 5-days previously; 4) T. hardwicke with T. quinquevittatum introduced 12-days previously; 5) T. quinquevittatum without T. hardwicke; 6) T. quinquevittatum with T. hardwicke introduced 5-days previously, and 7) T. quinquevittatum with T. hardwicke introduced 12 days previously. Note that treatment 2 served as the simultaneous arrival treatment for both species. For each experimental run, we size matched focal individuals and prior residents, so that all fish were the same size at the time they were added to reefs. Size differences between prior residents and focal individuals were then a function of the length of prior residency (e.g., growth advantage conferred upon 12-day prior residents relative to 0-day prior residents was 12 days). Because recently settled fish often experience high rates of mortality, we initially introduced 6 prior residents to reefs in the 12-day and 5-day treatments. Prior to introducing focal fish to these treatments, we haphazardly removed excess residents when there were more than three residents (18 instances out of 32 reefs), and pressed residents at the same time as focal individuals were added when there were less than three residents (6 instances out of 32). We surveyed reefs twice daily (approximately 8 am and 4 pm) for seven days after we introduced focal individuals.

Experiment 2: Intraspecific Priority Effects between T. hardwicke Individuals
We used a similar experimental design to examine intraspecific priority effects for T. hardwicke (i.e., T. hardwicke recruits served as both the focal and resident individuals). Because the intraspecific design had four instead of seven treatments, all seven replicates of the four treatments were conducted together once (6 July to 25 July).

Behavioural Observations
To help elucidate the mechanisms driving priority effects we conducted five-minute behavioral observations at the time we added focal individuals to reefs. After allowing fish to acclimate to the observer’s presence for approximately three minutes, behavioral observations were conducted at a distance of approximately two meters from the reef. We recorded three response variables: 1) The number of chases between focal individuals and both prior residents and the background community; 2) The number of fin bites inflicted during chases; and 3) Time spent inside Pocillopora by focal individuals.