SOUTAR and ISAACS: ABUNDANCE OF PELAGIC FISH 



Figure 3. — Comparison of Santa Barbara regional rainfall, 

 smoothed Santa Barbara regional rainfall, and total varve 

 thickness. Spearman rank-correlation coefficient between 

 Santa Barbara regional rainfall and total varve thickness is 0.26 

 (P = 0.02, n = 99). The highly variable pattern of seasonal 

 rainfall is smoothed by the filter: 



Y,= 



i"^ 



a-4)i-,- 



where Yf = smoothed seasonal rainfall at year t; R/ = actual 

 seasonal rainfall at year i; and r = a time constant (years). Thus 

 Yf is the sum of^ of year t rainfall and exponentially decreasing 

 portions of previous year's rainfall. The function tends to damp- 

 en rapid oscillations and lag slower oscillations at a slightly 

 lower amplitude. The value of ris derived by multiple regression 

 analysis. Varve thickness at year t is the dependent variable and 

 rainfall at year t, year t-l, year t-2, etc. are the independent 

 variables. T is found by fitting a log linear line to the regression 

 coefficients. The Spearman rank-correlation coefficient between 

 the filtered rainfall and the total varve thickness is 0.50 {n = 99). 

 Due to the autocorrelation induced by the filter, no probability is 

 assigned. Mean Santa Barbara regional rainfall is 42.2 cm, and 

 mean varve thickness is 1.74 mm (at 60% water by weight). 



M moles at depth. Since the composition of fish 

 scales is essentially an intermixture of micro- 

 crystalline apatite and a collagen-ichthylepidin 

 matrix ( Wallin, 1957), a potential for the degrada- 

 tion of the scale record is present. However, 

 ichthylepidin, an albuminoid approaching keratin 

 in composition, is unlikely to be degraded in an 

 anaerobic environment (Kaplan and Rittenberg, 

 1963). Thus the organic matrix may contribute 

 significantly toward the preservation of scales. 

 The distinctly higher frequencies of scales at core 

 depths in excess of 2 m further suggest a non- 

 degraded record (Soutar and Isaacs, 1969). 



The record of scale deposition in the Santa Bar- 

 bara Basin (16 subcore-sample mean, 1810-1970) 

 and the Soledad Basin (4 subcore-sample mean, 

 1780-1970) is in part presented (Figure 5a, b) and 

 statistically summarized (Table 1) for: Sardinops 

 caerulea (Pacific sardine), Engraulis mordax 

 (northern anchovy), Merlucciusproductus (Pacific 

 hake), Colalabis saira (Pacific saury), and 

 Scomber japonicus (Pacific mackerel). 



Of particular interest are those portions of the 

 scale record covering the past few decades for 

 which population estimates exist. Considerable 

 attention has been directed towards the elucida- 

 tion of the historical population levels of the 

 Pacific sardine. Estimates of the population (fish 2 

 yr and older) derived from the solution of a fishery 

 catch equation extend from 1930 up to 1959 (Mur- 



phy, 1966). These biomass estimates presented in 

 single- and 5-yr averages (Figure 6) document the 

 historical decline of the fishery and the popula- 

 tion. Comparison of the 5-yr averages of the 

 biomass and the scale-deposition rate in the Santa 

 Barbara Basin sediment indicates a parallel but 

 offset decline. The derived age frequency of the 

 sedimented scales (Table 2) indicates that most 

 (92%) of the contributing fish were less than 2 yr 

 old, suggesting a relatively fast response on the 

 part of the sediment to particular year-class sizes. 

 Comparison of the 5-yr averages of the year-class 

 size (numbers of 2-yr-old fish entering the fishery) 

 and the scale-deposition rate indicates a direct 

 proportional relationship (Figure 7 and Table 3). 

 Population estimates for the central, southern, 

 and total populations of the northern anchovy for 

 the years 1951 to 1966 have been made (Smith, 

 1972). The estimates of the total spawning pop- 

 ulation presented in single- and 5-yr averages, 

 and the 5-yr averages for the subpopulations, 

 record the recent ascendancy of the anchovy. 

 Comparison is made of these population estimates 

 with the northern anchovy scale-deposition rate 

 in the Santa Barbara and Soledad Basin sedi- 

 ments (Figure 8 and Table 4). In the three 5-yr 

 intervals having sufficient information, the scale 

 deposition is proportional to the spawning 

 biomass. The direct relation to spawning biomass 

 may be associated with the relatively rapid mat- 



Table 1. — Statistical summary of scale deposition 

 (numbers/10='cm*/yr) for Santa Barbara Basin sediment, 

 1810-1970, and Soledad Basin sediment, 1780-1970.^ 



'In the case of the Santa Barbara sediment, the statistics are drawn from a 

 16subcoreset representing a combinedareaof 980 cm^. The statistics for 

 the Soledad Basin are drawn from a 4 subcore set having a combined area 

 of 260 cm2. 



261 



