568 



Fishery Bulletin 105(4) 



eggs because nesting behavior ended 89 days later, 

 which is a few days longer than the period needed for 

 eggs to incubate and hatch at the water temperatures 

 recorded by the archival tag (mean=4.9°C). If one con- 

 siders that this male may have already spent 2 months 

 establishing a nest before being captured and tagged, 

 the total nesting period would be 6.5 months. Given 

 the ambient water temperature and late July start of 

 spawning, there was insufficient time for the other 

 three male Atka mackerel to brood a new or existing 

 batch of eggs. Being displaced from an original nesting 

 territory or failing to attract a female to spawn may be 

 reasons for changing a nesting site or for abandoning 

 it altogether. Nesting colonies can have a broad depth 

 range, cover areas over many square kilometers, and 

 be physically and biologically diverse (Lauth et al., in 

 press); therefore the quality of some territories within 

 a nesting colony may be more suitable for attracting 

 females to spawn. 



In addition to males changing the depth and location 

 of their nesting territory, there is evidence to suggest 

 that some male Atka mackerel adopt already existing 

 nesting territories containing eggs. In the exhibit tank 

 at the ASLC, a non-nesting male displaced a territo- 

 rial nest-guarding male and adopted the egg masses 

 already present on the nest. Adoptive nest behavior is 

 also thought to be a characteristic of another hexa- 

 grammid, lingcod (Ophiodon elongates) (Withler et 

 al., 2004). Abandoning nests or allowing another fish 

 to assume nest-guarding responsibilities is a common 

 behavior in males that provide parental care of eggs 

 (Gozlan et al., 2003). Male fish that guard territories 

 containing eggs are more reproductively successful 

 because females prefer to mate with males that al- 

 ready have eggs (e.g., Marconato and Bisazza, 1986; 

 DeMartini, 1987; Pruett-Jones, 1992). 



There is a high energetic cost to male parental care 

 (Marconato et al., 1993; Gillooly and Baylis, 1999; 

 Steinhart et al., 2005), and protracted nest-guard- 

 ing severely restricts feeding opportunities for males. 

 Egg cannibalism by males is common in Atka mack- 

 erel (Zolotov, 1993; Yang, 1999) and may serve as a 

 means for nest-guarding males to sustain themselves 

 during the protracted nesting season (Zolotov, 1993). 

 Cannibalism of eggs within a male's own nesting ter- 

 ritory (filial cannibalism) is considered an adaptive 

 behavior in marine fishes that provide paternal care 

 of eggs and it is especially common in species with 

 protracted breeding cycles (Rowher, 1978; Kondoh and 

 Okuda, 2002). 



Commercial trawling on nesting habitat during the 

 spawning and breeding periods would probably re- 

 sult in high mortality of both nest-guarding males 

 and developing embryos. Atka mackerel aggregate for 

 spawning and nesting (Lauth et al., in press), and ag- 

 gregating behavior is generally predictable and makes 

 a population easier to target and more vulnerable to 

 fishing (Colin et al., 2003). The reproductive success 

 of a fish species is also negatively impacted by al- 

 terations to the structural habitat and the benthic 



community — alternations that are caused by fishing 

 gears (Auster and Langton, 1999). There is some spa- 

 tial and temporal overlap of commercial trawling and 

 nesting sites in the Aleutian archipelago (Fritz and 

 Lowe, 1998; Lauth et al., in press), but it is difficult 

 to assess the impacts of the fishery on the reproduc- 

 tive success of Atka mackerel without knowing the 

 total distribution of nesting habitat. Knowledge of 

 the temporal use of breeding habitat can be used as a 

 management tool for minimizing negative impacts of 

 the commercial fishery during crucial periods of Atka 

 mackerel life history. 



Conclusions 



This study provides detailed information about the 

 timing and duration of three specific phases of the 

 Atka mackerel's reproductive cycle: 1) establishment 

 of a nest, 2) spawning, and 3) egg brooding. An accu- 

 rate incubation rate from controlled laboratory experi- 

 ments, egg mass collections from nesting sites, and 

 water temperatures characteristic of those observed 

 at nesting sites were used in an incubation model that 

 showed that nest-guarding males use nesting habitat 

 for more protracted periods than previously known. 

 Model results for the start and end of the reproductive 

 cycle were validated with observations from a time- 

 lapse video camera and archival tags. The duration 

 of the reproductive cycle appears to be affected most 

 by the length of the male brooding phase, which can 

 double in duration within the range of ambient water 

 temperatures observed in our laboratory incubation 

 experiments. Knowledge about the timing and dura- 

 tion of the reproductive cycle is essential for conserv- 

 ing Atka mackerel populations and for providing a 

 solid framework for investigations into the physical 

 and biological processes influencing recruitment and 

 population dynamics. 



Acknowledgments 



This project was funded by grants from the North Pacific 

 Research Board (# F0417) and the National Marine 

 Fisheries Service (NMFS) Essential Fish Habitat (EFH) 

 fund. Rearing experiments at the Alaska Sealife Center 

 (ASLC) by J. Guthridge were part of the requirements 

 for his Master of Science thesis at the University of 

 Alaska, Fairbanks. R. Lauth thanks the skippers and 

 crew from the FV Morning Star. FV Sea Storm, FV Sea- 

 fisher, and RV Tigldx, especially K. Bell, S. Branstiter, 

 P. Deng, T. Meintz, and S. Clark for field support of this 

 research. Many thanks to S. Atkinson from ASLC for 

 assisting with the proposal and those who helped with 

 field support, scuba diving, or reviewing this manu- 

 script including D. Cooper, G. Duker, G. Hoff, J. Lee, L. 

 Logerwell, S. Lowe, S. McDermott, R. Nelson, K. Rand, 

 C. Rilling, D. Somerton, G. Stauffer, S. VanSant, and 

 two anonymous reviewers. 



