McDonald et al. : Spatial and seasonal abundance of Cynoscion arenarius and C. nothus off the coast of Texas 
27 
Table 2 
Summary of ANOVA results of abundance data for sand seatrout (Cynoscion arenarius ) and silver seatrout ( C . nothus) from gulf 
areas (three-factor ANOVAs) and abundance of sand seatrout compared between locations, pass or no pass, and bays with pass 
presence (two-factor ANOVAs) averaged across twenty years (1987-2006) of trawl capture monitoring by Texas Parks and Wild- 
life. Log-transformed abundance data were the dependent variable in all analyses. Degrees of freedom (df), mean squares (MS), 
F-values (F), and P-values (P) reported, ns = P>0.05; * = P<0.05 
Dependent variable 
Factors 
df 
MS 
F 
P 
Sand seatrout and silver seatrout 
Gulf areas 
4 
18.27 
60.41* 
<0.0001 
(Log 10 abundance) 
Season 
3 
4.15 
13.73* 
<0.0001 
Species 
1 
34.66 
114.58* 
<0.0001 
Gulf areas x season 
12 
1.79 
5.93* 
<0.0001 
Gulf areas x species 
4 
10.24 
33.86* 
<0.0001 
Season x species 
3 
5.48 
18.12* 
<0.0001 
Gulf areas x season x species 
12 
0.40 
1.34 ns 
0.1906 
Sand seatrout (Log 10 abundance) 
Location 
1 
6.60 
101.16* 
<0.0001 
Season 
2 
0.35 
5.31* 
0.0062 
Location x season 
2 
0.14 
2.17 ns 
0.1192 
Sand seatrout (Log 10 abundance) 
Pass 
1 
15.47 
192.74* 
<0.0001 
Season 
2 
0.70 
8.70* 
0.0003 
Pass x season 
2 
0.17 
2.11 ns 
0.1255 
Sand seatrout (Log 10 abundance) 
Bays 
5 
3.68 
22.72* 
<0.0001 
Season 
2 
1.79 
11.08* 
<0.0001 
Bays x season 
10 
0.18 
1.14 ns 
0.3337 
major bays) and seasons (n = 3; fall, spring and summer), 
and interaction (the winter season was excluded from 
this analysis because of the invariably low counts of sand 
seatrout within both locations, across all years). 
To determine whether abundance differed between 
bays with direct GOM passes and bays either without 
these passes or with limited offshore access, a two-fac- 
tor ANOVA was used involving the following factors: 
pass presence (n= 2; bays with a direct offshore pass 
and bays without a direct pass) and seasons (n- 3), with 
interaction. Analyses were then focused on bays with 
direct passes in order to determine whether distribution 
differed among individual bays with passes, seasonally. 
This analysis employed a two-factor ANOVA, involving 
the following factors: bays (n = 6; all major bays with 
passes) and seasons (n- 3), with interaction. Length- 
frequency histograms of inshore sand seatrout were 
then created in a similar fashion to that used for the 
previously created offshore length-frequency histograms 
in order to qualitatively evaluate differences in monthly 
cohort size and size classes of sand seatrout between 
locations (i.e., offshore and inshore). 
Statistical analyses 
All data were first averaged across seasons for each 
year and then analyzed across years for all dependent 
variables in all parametric tests. All dependent variables 
in these analyses were first tested for normality by 
using a Shapiro-Wilk test; however, in the case of non- 
normality, data were log 10 -transformed before analysis. 
All abundance data for ANOVAs involved all catch (zero 
catches included) so that catches were not overestimated. 
Statistical analyses and length-frequency histograms 
were carried out with SAS software (SAS vers. 8.02, 
SAS Inst., Inc., Cary, NC) and illustrated by using 
SigmaPlot (SigmaPlot vers. 10.0, Systat Software, Inc., 
Point Richmond, CA). 
Results 
Distribution of sand seatrout and silver seatrout 
Sand seatrout abundance was significantly lower (13.2 
±6.7) than that of silver seatrout (37.2 ±17.7) (t=- 8.55, 
PcO.OOOl). In addition, significant spatial and seasonal 
differences existed between sand and silver seatrout 
abundance, including significant interactive effects, and 
differences between species accounted for a majority 
of the variance in abundance in trawls in the 3-factor 
model (Table 2). The first interaction, seasonxspecies, 
revealed a lower abundance of silver seatrout through- 
out the summer season (Fig. 2A). Sand seatrout abun- 
dance was high during July and decreased by August, 
whereas silver seatrout abundance peaked in April, 
declined in June, and was minimal by July (Fig. 3). 
The second interaction, gulf area x species, revealed 
a high abundance of sand seatrout and low abundance 
of silver seatrout in the gulf area A (Fig. 4). The gulf 
area x season interaction although significant, explained 
only a minimal amount of variance in abundance and 
