Mesoscale activity facilitates energy gain in a top predator

Abstract

How animal movement decisions interact with the distribution of resources to shape individual performance is a key question in ecology. However, links between spatial and behavioural ecology and fitness consequences are poorly understood because the outcomes of individual resource selection decisions, such as energy intake, are rarely measured. In the open ocean, mesoscale features (approx. 10-100 km) such as fronts and eddies can aggregate prey and thereby drive the distribution of foraging vertebrates through bottom-up biophysical coupling. These productive features are known to attract predators, yet their role in facilitating energy transfer to top-level consumers is opaque. We investigated the use of mesoscale features by migrating northern elephant seals and quantified the corresponding energetic gains from the seals' foraging patterns at a daily resolution. Migrating elephant seals modified their diving behaviour and selected for mesoscale features when foraging. Daily energy gain increased significantly with increasing mesoscale activity, indicating that the physical environment can influence predator fitness at fine temporal scales. Results show that areas of high mesoscale activity not only attract top predators as foraging hotspots, but also lead to increased energy transfer across trophic levels. Our study provides evidence that the physical environment is an important factor in controlling energy flow to top predators by setting the stage for variation in resource availability. Such understanding is critical for assessing how changes in the environment and resource distribution will affect individual fitness and food web dynamics.

Keywords: Lagrangian coherent structures; body condition; elephant seal; energy transfer; foraging; resource selection.

Figure 1.

Lagrangian coherent structures such as mesoscale fronts, eddies and filaments are identified as linear ridges of high absolute values in FSLE fields (a). Relationships between point FSLE values (a) and spatial mean (b), maximum (c) and standard deviation (d) values within a 25 km (5 pixel) radius denoted by the top left circle are demonstrated in the additional panels. The centre of the circle is highlighted in the mean, maximum and standard deviation panels with arrows.

Figure 2.

(a) Migration track from 2012 of an adult female northern elephant seal in relation to backward-in-time FSLEs in September 2012. The black box indicates the 95% bounding box of September telemetry locations for all individuals for measuring third-order resource selection (selection of habitat areas within a home or subpopulation range). (b) Overlap between telemetry locations (black dots) and time-matched FSLE map to explore fourth-order resource selection (selection of specific sites within a habitat area). White dots indicate foraging dive locations as captured by time-depth recorders.

Figure 3.

Response curves of GAMMs for n = 142 adult female northern elephant seals relating (a,b) foraging behaviour to absolute value of FSLEs and standard deviation of FSLE within a 25 km radius and (c,d) maximum dive depth and bottom time (calculated as time spent below 95% of the maximum dive depth) to the probability of a dive occurring in an FSLE-derived Lagrangian coherent structure. The vertical lines in (a) and (b) denote the average background absolute value of FSLE (0.047 d⁻¹) and FSLE spatial standard deviation (0.034 d⁻¹), respectively, within the study area. (Online version in colour.)

Figure 4.

(Top) Response curve of GAMM relating daily fat gains to the spatial standard deviation of FSLE for n = 29 seals. (Bottom) A histogram shows that daily lipid gains modelled from Schick et al. [35] are normally distributed. Response curves of GAMM to FSLE (left) and FSLE spatial standard deviation (right) for fourth-order selection for foraging behaviour of the same 29 seals. The vertical line denotes the average background values of FSLE and its spatial standard deviation. (Online version in colour.)