Fox et al.: Juvenile Acipenser oxyrinchus desoto/ in the Apalachicola River in Florida 251 
marked fish were captured more than once within a single 
summer. Although we cannot prove that no fish rejected 
PIT tags, the repeated capture of tagged fish within and 
among summers indicates that sturgeon retained their 
PIT tags and that tags were not overlooked. 
Survival 
Overwinter survival of juvenile Gulf sturgeon from age 1 
to age 2 appears quite variable, with high survival in some 
years (e.g., 90% in 2014) and substantially lower rates in 
others (e.g., 40% in 2015 and 33% in 2016). However, these 
estimates are conservative (i.e., biased lower than the true 
values) because we were unable to differentiate actual 
mortality of tagged fish from transmitter failure or lack of 
detection. We were able to supplement acoustic detections 
with physical recapture events to confirm that—after as 
long as several years at large—some fish had actually 
survived. Although we are unable to ascertain the fates 
of individuals that were not detected or recaptured at age 2 
or beyond, it is unlikely that these fish utilized another 
aggregation site outside of our receiver array because 
they were not detected moving upriver in the spring or 
in the main stem Apalachicola River. No tagged juveniles 
were detected outside of the Brothers River in surveys 
with the portable receiver, and many tagged fish were 
eventually recaptured in gill nets within the aggregation 
site in the Brothers River. 
The probability of detecting tagged fish in our array 
was undoubtedly <1. Although results of preliminary 
testing of the detection range of receivers indicate that 
receivers could detect a Vemco V7-4X transmitter at 
ranges of up to 800 m, we recognize that in most cases 
detection range is probably substantially smaller. Detec- 
tion range also likely varied with condition and orienta- 
tion of receivers, with river conditions at each site in the 
array, and even with ambient weather. Therefore, fish 
could certainly have swum past acoustic receivers in 
the array without detection. Additionally, distribution of 
receivers within the array changed over the course of this 
study. Beginning in 2016, many receivers in the Brothers 
River were moved to locations lower in the estuary. Our 
transmitter specifications also differed among study 
years, both in signal delay and activation delay, creating 
additional differences in probability of detection across 
study years. 
Despite imperfect detection, our survival analysis 
required only that we detect each tagged fish at least 
once in the year after tagging—to indicate that it had 
survived and moved back upriver. Most surviving tagged 
fish likely would have been detected at least once during 
the several months in which this detection could occur. 
Indeed, 77.5% of all fish confirmed to have survived to 
ages 22 were detected by the array (and most of those 
fish were detected hundreds of times). Additionally, if a 
fish did survive, each year it remained at large provided 
an additional sampling season in which we might catch 
it and confirm survival. Continued long-term monitor- 
ing of Gulf sturgeon in the Apalachicola River may yet 
reveal the survival of some fish that are currently pre- 
sumed dead by this model, increasing the rate of survival 
beyond what we report in this paper. 
This study was not intended to identify sources of over- 
winter mortality for Gulf sturgeon in the Apalachicola 
River. However, some mortalities may have resulted from 
transmitter implantation, especially in substandard envi- 
ronmental conditions. In 2016, water temperatures reached 
as high as 31.3°C (Suppl. Table 1) (online only). Although 
we did not conduct surgeries in temperatures above 28°C, 
exposure to very warm water shortly after being tagged 
may have facilitated the infections we observed at surgery 
sites of recaptured fish in 2016. These infections likely 
explain the decreased survival of tagged fish in 2016. 
Commercial fisheries in the area may also be responsible 
for some mortality of Gulf sturgeon. Wooley and Crateau 
(1985) documented incidental capture of juvenile and 
adult Gulf sturgeon (>800 mm TL) in shrimp trawls and 
other commercial fishing gears; juveniles are also poten- 
tially susceptible to these threats. 
Despite the conservative bias and potential inaccuracies 
of our estimates of overwinter survival, examining mean 
annual survival still has utility. During this study, the mean 
annual survival rate was 60.3% (SD 27.8)—which trans- 
lates to annual mortality of 39.7% (mortality=1—survival). 
This rate of mortality is roughly comparable to previous 
estimates of mortality for juvenile Gulf sturgeon: Morrow 
et al. (1998) found an annual mortality of 34% for fish at 
ages 3—9 in the Pearl River, and Pine et al. (2001) esti- 
mated that annual mortality was 25% for Gulf sturgeon 
at ages 1-3 in the Suwannee River. If we omit data for 
2016 from our survival analysis (presuming that low 
survival in that year was due to surgery site infections), 
mean mortality across this study was 30.7%., a value that 
is even more congruent with the reports in the literature. 
Tate and Allen (2002) simulated responses of populations 
of Gulf sturgeon to several rates of juvenile mortality 
and found that, at 30% annual mortality, populations 
remained stable over 200 years but that, at 35% annual 
mortality, the population slowly collapsed. Refining the 
accuracy of survival estimates remains an important 
goal for researchers and managers of Gulf sturgeon, and 
the ability in future studies to adjust survival estimates 
for imperfect detection should lead to estimates of true 
overwinter survival that are more accurate and higher 
than those we have reported. 
Conclusions 
The results of this study indicate that direct estimates of 
annual recruitment of Gulf sturgeon are feasible. In the 
Apalachicola River, we observed age-1 juveniles recruit- 
ing to the population in every year of our study, but abun- 
dance of age-1 fish was low: only about 50 individuals in 
5 of the 6 years of this study. 
The methods employed in this study can be used to 
obtain recruitment data for other populations of Gulf stur- 
geon, once juvenile aggregation sites have been located. 
