Fishery Bulletin 120(1) 
species were widespread in the study area, northern 
rock sole were in deeper, colder water than yellowfin 
sole (Hurst, 2016). In this study, the overlap coefficient 
was notably higher (SI=0.55) when calculated by using 
only fish collected at sampling locations where the spe- 
cies co-occurred. Interestingly, the opposite pattern was 
observed in coastal embayments in the GOA: dietary 
overlap was lower at stations where these species co- 
occurred than across all stations in the study area (Hol- 
laday and Norcross, 1995). 
Dietary overlap indicates the potential for compet- 
itive interactions between species, but index values 
should be interpreted with caution. Dietary overlap may 
occur between species because of the shared utilization 
of locally abundant resources as opposed to competi- 
tion for a limited resource (Carter et al., 1991). Alter- 
natively, when species have niche separation (i.e., low 
dietary overlap), it can be difficult to determine if the low 
overlap reflects differences in habitat preferences with 
diets reflecting availability of prey or vice versa (i.e., 
consumers occupy habitats because of the availability 
of preferred prey). The novel approach to application of 
PCA to multi-species diet composition data in our study 
provided additional insight into foraging dynamics of 
the northern rock sole and yellowfin sole. This analysis 
included both the biotic traits of the consumer (size and 
Fulton’s condition factor) and the abiotic aspects of their 
environment (depth, temperature, and sediment charac- 
teristics). The general separation between these predator 
species along the first axis of the PCA illustrates the dif- 
ference in overall foraging habitats, with northern rock 
sole consuming prey in deeper waters, with more sandy 
sediments, and at later times of day in comparison to the 
diets of yellowfin sole. The separation also indicates that 
certain prey were predominantly consumed in specific 
habitats. In particular, euphausiids were consumed in 
deeper water than other prey by both species; whereas, 
hyperbenthic copepods were consumed by both species in 
shallow waters. 
Foraging patterns and predation risk 
Flatfishes possess a unique combination of morphological 
and behavioral adaptations associated with their demer- 
sal lifestyle. These traits indicate that avoiding detection 
by predators is critical to the survival of juvenile flatfish. 
For flatfish species, foraging activity likely directly con- 
flicts with anti-predator behaviors that reduce conspicu- 
ousness. This trade-off is expected to be even greater when 
flatfish leave the sediment surface to forage in the water 
column because of the exposure of their unpigmented (and 
highly conspicuous) blind side (Hurst and Duffy, 2005). 
However, even within this guild, there are differences 
among species in behavioral characteristics. 
In earlier studies that compared life history strategies 
of 3 co-occurring flatfishes, northern rock sole were con- 
sidered the most “risk averse” in that they maintained 
the lowest body profile on the sediment surface and 
had the lowest activity rates (Lemke and Ryer, 2006). 
Consistent with the finding of this risk-averse lifestyle, 
laboratory and field observations indicate that foraging 
(Corcobado Onate, 1991; Hurst et al., 2007) and overall 
activity of northern rock sole were greatest during dusk 
periods (Hurst and Duffy, 2005). Although sampling 
during this study was conducted only during daylight 
hours, northern rock sole had a temporally restricted 
feeding pattern consistent with that observed in those 
previous studies. 
Comparatively little work has been conducted on the 
behavioral ecology of juvenile yellowfin sole. However, sev- 
eral pieces of evidence indicate that their strategy may 
be even more risk averse than that of northern rock sole. 
Yellowfin sole consumed less hyperbenthic and pelagic 
prey whose capture presumably would entail risky off- 
bottom swimming. In addition, Hurst (2016) found juvenile 
yellowfin sole at shallower depths than northern rock sole, 
indicating reduced encounters with large-bodied preda- 
tors, as articulated in the shallow-water refuge hypothesis 
(Paterson and Whitfield, 2000; Linehan et al., 2001). 
Conclusions 
The coexistence of yellowfin sole and northern rock sole 
on shared nursery grounds is likely facilitated by aspects 
of niche separation along habitat and dietary preferences. 
Yellowfin sole appear to be obligate benthic foragers with 
narrower diets than those of northern rock sole, indicating 
a greater sensitivity to environmentally driven changes 
in availability of demersal infauna and epifauna (Poiesz 
et al., 2020). These differences also indicate that these 
species may respond differently to environmental change. 
Because our sampling was conducted in a comparatively 
cold year (in 2012) in a rapidly warming ocean region 
(Stabeno et al., 2017), the data from our study serve as a 
useful baseline for evaluating effects of climate on food- 
web interactions in the SEBS. In addition to dietary differ- 
ences between the 2 species, growth rates of yellowfin sole 
appear more sensitive to temperature variation than those 
of northern rock sole (Matta et al., 2010; T: Hurst, unpubl. 
data). Interestingly, the distribution of juvenile yellowfin 
sole appears to be more stable in the face of fluctuating 
environmental conditions than that of northern rock sole 
(Yeung and Cooper, 2020). This combination of narrower 
diet breadth, greater thermal sensitivity, and stable distri- 
bution may result in stronger environmental sensitivity in 
the productivity of yellowfin sole in the future. 
Acknowledgments 
Thanks to A. Stoner, the crew of the FV Bountiful, and 
staff of Peter Pan Seafoods for sampling and logistical 
assistance. M. Ottmar, D. Cooper, and C. Hines assisted 
with laboratory dissections. This project was supported 
by grants (to T. Hurst) for essential fish habitat studies 
from the Alaska Regional Office of the National Marine 
Fisheries Service (NMFS). C. Yeung, S. Henkel, B. Laurel, 
