138 
Fishery Bulletin 119(2-3) 
Figure 2 
Photographs of thin sections of otoliths from (A) redbait (Emmelichthys niti- 
dus) and (B) common jack mackerel (Trachurus declivis) collected off Kangaroo 
Island and New South Wales in Australia between 2014 and 2016. Images are 
shown at 3.2x magnifications under transmitted light. The white line indicates 
the location of measurement, and the horizontal white markings along the line 
indicate annual growth rings. 
with a reference age for increment measurement. Oto- 
liths were viewed under both reflected and transmitted 
light without focal adjustment or movement of the slide. 
Photographs were taken (with an Olympus DP73 digital 
microscope camera, Olympus Corp.) with the light sources 
overlaid on each other to enhance contrast and visibility 
of growth increments and allow increments to be accu- 
rately identified. 
Annual deposition of growth bands has been validated 
in common jack mackerel by using bomb radiocarbon 
analysis (Lyle et al.°) and in redbait by using marginal 
increment analysis (Ewing and Lyle, 2009). These valida- 
tions confirm that each growth increment, composed of a 
translucent zone and an opaque zone, represents 1 year 
of growth. Starting from the core, increment boundaries 
were delineated at the outermost edge of each opaque 
zone (Fig. 2), by using the image analysis software Olym- 
pus Stream (vers. 1.9.1, Olympus Corp.). The distance 
from the core to the first increment (which equates to 
growth from hatching to the first birthday; age 0) and 
from the final increment to the proximal edge (marginal 
increment) were excluded from all analysis because they 
did not represent a full year of growth. On the basis of 
the timing of spawning, a birthdate of 1 January was 
assigned for common jack mackerel and a birthdate of 1 
October was assigned for redbait (Ewing and Lyle, 2009; 
6 Lyle, J. M., K. Krisic-Golub, and A. K. Morison. 2000. Age and 
growth of jack mackerel and the age structure of the jack mack- 
erel purse seine catch. Fish. Res. Dev. Corp. Final Rep., Proj. 
1995/034, 49 p. Tasman. Aquac. Fish. Inst., Mar. Res. Lab., Univ. 
Tasman., Taroona, Australia. [Available from website.] 
Ward et al., 2016). Each increment was 
then assigned a calendar year of forma- 
tion, back calculated from the capture 
date. 
Growth analysis 
The AquaticLifeHistory package (vers. 
0.0.9000; Smart et al., 2016; Smart, 2019) 
was used in the statistical program R 
(vers. 3.4.0; R Core Team, 2017) to esti- 
mate growth parameters for all otoliths 
from each species and sampling location 
by fitting length-at-age data with the von 
Bertalanffy growth function (von Berta- 
lanffy, 1938; Beverton and Holt, 1957): 
silly > Ch, Ik =e, 
where t¢ = age in years; 
L, = length at age ¢; 
Ly = the length at age 0 (fixed at 0); 
L,, = asymptotic length; and 
k =the growth completion para- 
meter. 
The von Bertalanffy growth function 
was fit by using the nls function in R. 
Differences in growth between sampling 
locations were then compared for each species by using a 
likelihood-ratio chi-square test (Ogle, 2016). 
Chronologies and environmental variables 
The growth chronology data set was truncated for years 
for which less than 5 increment measurements were avail- 
able (Morrongiello and Thresher, 2015). A set of general- 
ized linear mixed-effects models were fitted with a gamma 
error structure and a log-link function. These were used 
to investigate the sources of growth variation, both intrin- 
sic and extrinsic, within species and regions (Table 1) 
(Morrongiello and Thresher, 2015). Sea-surface tempera- 
ture and chlorophyll-a (Chl-a) concentration (used as proxy 
for productivity) were both included as extrinsic variables 
because they have been proven to influence the physiology 
and somatic growth of fishes (Hughes et al., 2017). Local 
SSTs and Chl-a concentrations were obtained by defining 
boundary coordinates around all trawl tow positions for 
each region in the Integrated Marine Observing System 
of the Australian Ocean Data Network (IMOS”®). Sea- 
surface temperatures were obtained as 1-day composite 
data from polar-orbiting satellites (IMOS’) and then aver- 
aged to produce annual means. Chlorophyll-a concentra- 
tions were obtained from daily satellite images computed 
7IMOS (Integrated Marine Observing System). 2018. IMOS- 
SRS-SST-L3S-single sensor-1 day-day and night time-Australia. 
[Available from website, accessed December 2018.] 
8 IMOS (Integrated Marine Observing System). 2018. IMOS- 
SRS-MODIS-01 day-Chlorophyll-a concentration (OC3 model). 
[Available from website, accessed December 2018.] 
