390 
Fishery Bulletin 106(4) 
middens indicate that there were slight or moderate 
declines in Pacific cod during this early period. On 
the basis of the remains in the middens, there was a 
sharp decline in the relative abundance of Pacific cod 
(compared to the other three taxa) beginning around 
520 AD. This declining trend intensified drastically in 
the 1030 AD assemblage, and was followed by a sharp 
increase in Pacific cod abundance at 1540 AD. 
The declining AIs strongly indicate that a form of 
resource depression occurred from 520 through 1030 
AD (for a similar interpretation of declining fish AIs, 
see Nagaoka, 2001, 2002). Resource depressions can 
be related to a number of factors, such as exploitation 
pressure, behavioral changes, and microhabitat (range) 
shifts (e.g., Charnov et al., 1976). The apparent decrease 
in Pacific cod abundance that we noted is intriguing 
given the lack of any technological, procurement, or 
other cultural changes (e.g., increased territoriality) 
that could have influenced the encounter or success rate 
of the Aleut Pacific cod fishery. Instead we suspect the 
decline in Pacific cod reflects an environmentally-driven 
natural change in abundance and note that the increase 
in mean length during this period corroborates this 
interpretation (see Shin et al., 2005). 
Interpreting the effects of climate 
on Pacific cod size and abundance 
Punctuated shifts in ocean climate (regime shifts) are 
believed to explain many of the changes observed in 
abundances of some species in the North Pacific Ocean 
during the past century (Hare and Mantua, 2000; 
Benson and Trites, 2002; Trites et al., 2007) and these 
shifts may explain the variations noted in Figure 4. 
Pacific cod populations may be susceptible to fluctuations 
in oceanic regimes to the extent that they periodically 
disappear in significant numbers from the ecosystem, 
only to reappear in greater numbers at a later date. This 
condition has deep historical roots; the ancient Aleut 
name for Pacific cod translates literally into “the fish 
that stops” because this species periodically disappears 
(Black, 1981), a situation that occurred, according to 
traditional Aleut knowledge, at least once in the mid-19 th 
century, and again in approximately 1942. 
Several small variations in mean Pacific cod length 
were evident over the temporal sequence. Correlating 
the shifts in body size with climatic shifts does not 
appear to explain the fluctuating sizes of Pacific cod 
between 2550 BC and 80 BC, which occurred during 
the generally cool and wet conditions of the Neoglacial 
period (the first major postglacial cooling period from 
approximately 2500 BC to AD 1). The average lengths 
of Pacific cod (Fig. 2) increased as the warming and 
drying of the Medieval climatic anomaly began (a period 
of hemispheric climatic fluctuations ca. AD 1000-1300 
across the Northern Hemisphere) and decreased slightly 
during the cool and wet conditions of the Little Ice Age 
(ca. AD 1400-1850). Marine productivity during the 
Medieval climatic anomaly and the Little Ice Age, as 
recorded by Finney et al. (2002) and Misarti (2007), 
appears to be inversely related to Pacific cod lengths 
but positively related with Pacific cod numbers (Fig. 4). 
Such inverse relationships between mean fish 
length, abundance, and productivity have been 
noted by Shin et al. (2005) (i.e., increased 
productivity and abundance is usually associ- 
ated with decreases in mean length, presum- 
ably because of the increase of juvenile fish in 
the population). The impact of these types of 
processes on fish lengths over the long time 
scales has never been assessed, and so this 
must remain a working hypothesis. Regard- 
less, it is also notable that the mean length 
of Pacific cod during the Medieval climatic 
anomaly was not statistically different from 
that of the modern era codfish (Table 1) — two 
of the warmest periods in the last 4500 years. 
The effects of oceanic regime shifts that 
have dominated the dynamics of marine eco- 
systems in the North Pacific for the past cen- 
tury (Benson and Trites, 2002; Polovina, 2005) 
are difficult to monitor in the prehistoric pe- 
riod because of an important complicating fac- 
tor — an overwhelming lack of data. Although 
climatic and oceanic warming AD 1000 to 
1300 have been documented around the North 
Pacific (Calkin et al., 2001; Hu et al., 2001), 
and its effects are beginning to be understood 
(Jones et al., 1999), the full effects of climate 
change on Pacific cod stocks over the entire 
period are difficult to measure because not a 
Figure 3 
Mean estimated fecundity of Pacific cod ( Gadus macrocephalus) over 
time based on the relationships between body length and fecundity 
established by Karp (1982). This graph indicates that estimated 
mean fecundity of Pacific cod has varied widely over the past 4500 
years, despite concomitantly small changes in mean Pacific cod 
length (compare with Fig. 2). Although the graph indicates an 
apparent trend toward decreasing fecundity, the relationship is 
not significant; r=-0.42, P=0.35. 
