Scientific Literature: Searchable Database

We weighted krill density values as data are overdispersed (i.e., its variance ishigher than its mean), so that using the mean in the analysis would probably lead to a misinterpretation of thecorrelation between variables being tested. Cross-correlation methods allow the identification of the time lag thatmaximizes the correlation between the explanatory and the response variables (Legendre & Legendre, 1998) andwas run in PAST software version 3.0 (Hammer et al., 2001). The result was later used in a regression model run inR software version 4.0.2 (R Core Team, 2020) to test the relationship between krill biomass and RBR weighted byeffort (number of trips of each year). We considered a 0.05 level of significance in the analysis.Relative birth rates of humpback whales were computed from data collected during 278 whale-watching trips(M = 39.7 ± 7.9 per year) (Table 1). A total of 127 calves (M = 18.1 ± 13.2 per year) and 1,637 noncalves (M = 233.8± 130.4 per year) were recorded, excluding resights. The RBR (M = 0.07 ± 0.03) was significantly correlated with krilldensity in the previous year (i.e., lagged by 1 year; r2= 0.9, p = .02; Figure 2). The relationship between RBR and krilldensity in the previous year balanced by the number of trips was also positive and significant (r2= 0.8, p = .02).This is the first time that the relationship between environmental conditions around the Antarctic Peninsula andcalf production by humpback whales from Breeding Stock G has been investigated.