Preliminary observations on the age and growth of dog snapper (Lutjanus jocu) and mahogany snapper (Lutjanus mahogoni) from the Southeastern U.S.

Age estimation and timing of opaque zone formation

Dog snapper and mahogany snapper were opportunistically sampled from fisheries operating offshore of North Carolina through Key West, FL. All specimens used in this study were killed as part of legal fishing operations and were already dead when sampled by the port agents; thus all research was conducted in accordance with the Animal Welfare Act (AWA) and with the US Government Principles for the Utilization and Care of Vertebrate Animals Used in Testing, Research, and Training (USGP) OSTP CFR, May 20, 1985, Vol. 50, No. 97. All fish were captured by either conventional vertical hook and line gear or by spears. Sagittal otoliths were collected from 62 dog snapper and 57 mahogany snapper by National Marine Fisheries Service port agents sampling the recreational headboat and commercial fisheries from 1979 to 2015. Total (TL) and fork lengths (FL) of specimens were recorded in millimeters (mm), and whole weight was recorded in grams (g) for fish sampled by the Southeast Region Headboat Survey (SRHS). Weights were generally unavailable for fish landed by the commercial fisheries, as these fish were eviscerated at sea. Sagittal otoliths were removed through the otic bulla inside the gill cavity and stored dry in coin envelopes. Otoliths were sectioned on a low-speed saw, following the methods of Potts & Manooch III (1995). Three serial 0.5 mm sections were taken, with at least one of them encompassing the otolith core. The sections were mounted on microscope slides with thermal cement and covered with histological mounting medium before analysis. The sections were viewed under a dissecting microscope at 12.5X using reflected light (Fig. 1). Each sample was assigned an opaque zone count by two readers. Counts were compared between readers. An index of average percent error (APE) was calculated, following the methodology of Beamish & Fournier (1981). Where two readings for a specimen disagreed, the sections were viewed again. If agreement was reached the sample was retained; otherwise, the sample was discarded from further analysis.

We assessed opaque zone periodicity using otolith edge, or margin, analysis by visual categorization. The edge type of the otolith was noted: 1 = opaque zone forming on the edge of the otolith section; 2 = narrow translucent zone on the edge, generally less than 30% of the previous translucent zone; 3 = moderate translucent zone on the edge, generally 30% to 60% of the previous translucent zone; 4 = wide translucent zone on the edge, generally greater than 60% of the previous translucent zone (Harris et al., 2007). All samples were assigned an age based on edge frequency analysis metrics, opaque zone count and time of capture. The zone count was increased by one, to reflect the calendar age of the fish, if the specimen was caught before increment formation and had an edge that was a moderate to wide translucent zone (type 3 or 4).

Theoretical growth

Von Bertalanffy (1938) growth parameters were estimated from the observed length-at-age data. Calendar age was used since little information on reproduction of these species in U.S. waters was available to assign a biological age. Growth parameters were derived initially by minimizing the negative sum of log-likelihoods with the AD Model Builder estimation software (Otter Research Ltd., Sidney, B.C., Canada). We examined parameters computed from a freely-estimated model, assuming normal distribution of lengths about each calendar age, and a model that adjusted for size selectivity bias within the fishery. Because samples for this study were obtained from fishery landings, the estimate of growth at the youngest ages may be skewed due to minimum size regulations imposed on the fishery or selection of the fish by the fishers. As an alternate strategy to model growth, the von Bertalanffy growth parameters were estimated using a left-truncated normal probability density function on length for fish subjected to the minimum size limit (305 mm TL) regulation, as developed by McGarvey & Fowler (2002). For samples in this study not subject to minimum size limit, the full, untruncated normal likelihood was used.

Body size relationships

Important metrics of fish species include the relationship between weight and length and between various length measurements. Whole fish weight (g) was regressed on TL using data for all dog snapper and mahogany snapper measured by the SRHS from 1979–2015 (n = 78, L. jocu; n = 79, L. mahogoni). We evaluated both direct fits using nonlinear least squares regression (SAS Institute, Cary, NC, USA) and a linearized fit of the log-transformed data, examining the residuals to determine which fit was more appropriate. We used the same SRHS data to examine the relationships between TL and FL using linear regression (n = 48, dog snapper; n = 65, mahogany snapper).

Age estimation

A total of 56 sagittal otoliths from L. jocu were sectioned (six otoliths were broken in storage). Fifty-four otoliths from L. mahogoni were sectioned, (three were broken in storage). The distribution by fishery and state of age samples is shown in Table 2. Dog snapper samples were primarily from southeastern Florida and Florida Keys waters (25 from commercial fishery landings and 25 from headboat angler landings) but did include five samples from the North Carolina commercial fishery and one sample from the South Carolina commercial fishery. All mahogany snapper came from southeastern Florida and Florida Keys waters, with 80% coming from recreational headboat fisheries and 20% coming from commercial landings.

Sectioned otoliths for both species were clear and easy to interpret (Fig. 1). Opaque zones were counted on all sectioned otoliths. Initial readings by the two authors resulted in 61% and 69% agreement for dog snapper and mahogany snapper, respectively. When we used ±1 year, agreement increased to 95% and 98% respectively. APE was 3.29% for dog snapper and 1.80% for mahogany snapper. These were both well within the criteria of 5% determined acceptable by Campana (2001). The age readers reviewed the samples where there was disagreement on the age and were able to reach consensus. Thus, all age data were used in subsequent analysis.

Timing of opaque zone formation

Opaque zones on the otolith marginal edge occurred May through July for both species (Fig. 2). Otoliths from dog snapper were without an opaque zone on the edge from August through April, while a single mahogany snapper caught in November exhibited an opaque edge. Dog snapper exhibited the least amount of translucent edge from August through October, with the width of the translucent edge increasing until reaching a maximum January through April, prior to opaque zone deposition in May. Mahogany snapper exhibited a similar pattern. We concluded that opaque zones on otoliths of both species were annuli.

We assigned calendar, or chronological, ages as follows: for fish caught January through July and having an edge type of 3 or 4, the annuli count was increased by one; for fish caught in that same time period with an edge type of 1 or 2 and for fish caught from August to December, the calendar age was equivalent to the annuli count.

Growth

Dog snapper in this study ranged from 200–836 mm TL and from ages 2–33, but only four fish were older than age-12 (Table 3). While 50 out of 57 fish (88%) were from Florida, the largest and oldest fish was from South Carolina. Predicted size-at-age of dog snapper agreed well with mean observed size-at-age (Fig. 3A). The Von Bertalanffy parameters (±1 Std. Err.) are presented in Table 4. The growth model parameters estimated when correcting for the left-truncated normal distribution imposed by the minimum size limit regulation (McGarvey & Fowler, 2002) are also in Table 4. The model appeared to have a more realistic estimate of initial growth (Fig. 3A; Table 4). Comparing negative log likelihood values from each model, the model incorporating the correction for the truncated normal distribution due to the size limit regulation was a better fit to the data than when assuming no bias (-log likelihood: 294.65 v. 343.95).

Mahogany snapper ranged from 270–416 mm and ages 2–18. All samples were from south Florida waters. Predicted size-at-age for the freely estimated model run agreed reasonably well with mean observed size-at-age (Fig. 3B), and the von Bertalanffy parameters (±1 Std. Err.) are presented in Table 4. As with dog snapper, the mahogany snapper specimens available for this study lacked smaller fish due to the fishery-dependent nature of the samples; the resulting growth parameters using the method of McGarvey & Fowler (2002) are in Table 4 and graphically presented in Fig. 3B. As with dog snapper, the model assuming the truncated normal distribution on lengths at the youngest ages was a better fit to the data than the freely estimated model (-log likelihood: 222.85 v. 245.57).

Weight–Length relations

When performing the statistical analyses of W–L relations, a multiplicative error term (variance increasing with size) in the residuals was revealed for both dog and mahogany snappers. A linearized ln-transform fit of the data was appropriate, and the resulting equations were: ${\displaystyle \text{Dog snapper—Ln}\left(W\right)=3.028\text{ln}\left(L\right)-11.249;\left(n=78,r^2=0.99\right)}$ ${\displaystyle \text{Mahogany snapper—Ln}\left(W\right)=3.154\text{ln}\left(L\right)-12.136;\left(n=79,r^2=0.79\right)}$

where W = whole weight in grams and L = total length in mm. The equations were transformed back to the power equation form, W = a L^b, after adjusting the intercepts for log-transformation bias with the addition of one-half of the mean-square error (1/2 MSE) (Beauchamp & Olson, 1973). The resulting regression equations were:

Length–Length relations

The relationships between TL and FL as determined by linear regression are described by the following equations: ${\displaystyle \text{Dog snapper}:\text{TL}=1.05\text{FL}+6.40\left(n=48,r^2=0.99\right)}$ ${\displaystyle \text{FL}=0.95\text{TL}-5.17\left(n=48,r^2=0.99\right)}$ ${\displaystyle \text{Mahogany snapper}:\text{TL}=1.03\text{FL}+9.27\left(n=65,r^2=0.93\right)}$ ${\displaystyle \text{FL}=0.89\text{TL}+12.01\left(n=65,r^2=0.93\right).}$