Table 1. Transformation of movement data of individual Hudsonian godwits into scores that reflect the accumulation or dissipation of lateness during their annual cycles.

	Beluga Departure	Saskatchewan Arrival	Saskatchewan Departure	Amazon Arrival
Population Mean	6 July	8 July	20 August	25 August
Individual HX (raw data)	7 July	11 July	17 August	21 August
HX (Step 1 — Relative Timing)	+1	+3	−3	−4
HX (Step 2 — Rate of Change)	+1	+2	−6	−1

The first line displays the population mean (2009, n = 15) timing of arrival and departure for three consecutive sites in the godwit annual cycle. The second line displays the dates of the movements between those sites for one individual godwit, “HX.” The third line displays the relative timing of HX's movements in relation to the population mean. In this case, HX departed Beluga on 7 Jul and the population mean departure was 6 Jul; thus HX departed Beluga 1 day later than the mean (+1). HX arrived in Saskatchewan 3 days later than the mean and thus has a score of +3. In the fourth line is the rate of change of HX's movements. This score reflects the timing of HX's movements both in relation to the population mean, but also in relation to the timing of its previous movements. HX left Beluga 1 day later than the mean and, also, arrived in Saskatchewan an additional two days later than the mean (three days later in total), giving scores of +1 and +2. However, it departed Saskatchewan 3 days earlier than the mean, thus giving it a score of −6 (+3 to −3). Values calculated in Step 1 allowed us to account for inter-annual differences in the movements of the entire population. Values calculated in Step 2 allowed us to determine if an individual's rate of change from mean timing was part of an individually consistent schedule (i.e., an individual always departing three days later than the mean) or whether they reflected an individual becoming increasingly later (or earlier) — a potential manifestation of the existence of carry-over effects.