Interactions of predators, habitat and conspecifics on the coastal distribution of age 0 juvenile cod (Gadus spp.) in Newfoundland

Newly settled age 0 juvenile cod, Atlantic cod Gadus morhua and Greenland cod Gadus ogac, coexist in nearshore areas around Newfoundland, but little is understood on the mechanisms affecting their distribution. In young marine fish, predation risk is often high yet there is generally an increasing capacity in larger individuals to evade predators. It is therefore difficult to distinguish the role of behaviour and differential survival on distribution of fish in their early life stages. In a series of field and laboratory experiments, I investigate how habitat complexity, predators and conspecific density interact to affect survival and behaviour in juvenile cod, and ultimately, their distribution in coastal Newfoundland. -- In a bay-scale study (~10 km²), I monitored patterns of settlement and distribution of G. morhua and G. ogac, following a large-scale alteration of nearshore eelgrass (Zostera marina) habitat. Comparisons of bi-weekly sampling between control and experimental sites from 1995-2001 indicated a significant increase in cod abundance at sites enhanced with simulated eelgrass and a corresponding decrease in cod numbers at sites where eelgrass had been removed. These data supported predictions, demonstrating that: 1) there was a sufficient supply of juvenile cod within the range of areas that have historically been unoccupied (i.e. sand) and 2) both species preferred to settle in complex habitats. However, G. ogac responded significantly to the removal of eelgrass in more comparisons than G. morhua (70% and 37% respectively), suggesting that G. ogac have a higher affinity for complex vegetative habitats than G. morhua at the scale of manipulation (ca. 800m²). Furthermore, despite an overall preference for eelgrass habitat, high within-site catch variation indicated that individuals of both species were not restricted to any single seine site. Rather, the highly variable data indicate that juvenile cod were mobile and occasionally aggregated throughout the study period. My results support previously described juvenile cod-eelgrass associations but contradict other published accounts of high site-attachment and restricted movement in G. morhua following settlement. -- In a second study, fish density, movement and schooling behaviour were examined in G. morhua and G. ogac using a combination of field-seines, mark-recapture-and laboratory experiments. Density estimates from seines (n=427) over five years (1996, 1998-2001) indicated that these species associated with eelgrass but periodically were detected in high abundance over sand. Within-site catch variation indicated both species aggregated more in sand than eelgrass habitat, and in G. morhua, aggregations over sand increased as density in eelgrass increased. Although such patterns in catch data could be interpreted as the effects of differential mortality between habitats, a mark-recapture experiment indicated that both species of cod were not site-attached and moved between seine locations. Furthermore, video-analysis from laboratory experiments demonstrated that cod formed tighter aggregations over sand compared to eelgrass habitats. The results demonstrate that juvenile cod modify their behaviour with changing density, possibly as a means of exploiting poor quality habitats when high quality habitats are saturated with conspecifics. Consequently, habitat suitability for G. morhua should be considered a dynamic rather than fixed variable in density-dependent habitat models. -- In a third study, I measured the effect of predator density and habitat patch size in age 0 juvenile cod in a nearshore field experiment. Eelgrass patch size was altered using artificial eelgrass mats of five sizes (0.32, 1.1, 5 .5, 11 and 22 m²) and subsequently deployed in duplicate at each of two sites in Newman Sound in Terra Nova National Park during summer-autumn, 1999 and 2000. Predator distribution was determined using a combination of weekly underwater transect surveys and biweekly seining. Relative predation rates were measured by tethering age 0 cod at the center of each patch and recording the incidence of predation (n=1116 tether sets). Predation rates were negatively correlated with patch size during both years, suggesting that larger patches reduce predator foraging ability. However, high predator densities in the largest eelgrass patch resulted in higher than expected rates of predation. Therefore habitat dimension affected predation risk in juvenile cod via two opposing mechanisms. The results emphasize the importance of considering both habitat areal extent and predator distribution when estimating the effects of habitat fragmentation on predation rates. -- In a final, small-scale study (1000m2, plays an important role in shaping distributions of age 0 juvenile cod in the nearshore. Behavioural decisions were largely made in response to the predation risk associated with predator distribution and refuge characteristics of habitat. Therefore, nearshore distributions of age 0 juvenile cod are not simply the result of differential mortality between areas of higher and lower risk. Consequently, these results provide information on home range, habitat preference and aggregation behaviour that would be relevant to future efforts in the conservation of Atlantic cod