Corresponding author : John Jacobs, NOAA/NCCOS Cooperative Oxford Lab, 904 South Morris Street, Oxford, MD 21654, USA, Tel: 410-226-5193-135; Fax: 410-226-5925; E-mail: email@example.com
Received: January 31, 2013 Accepted: March 29, 2013 Published: April 05, 2013
Citation:Rhodes M, Bhaskaran H, Jacobs J (2013) Short-Term Variation in the Abundance of Vibrio vulnificus and Vibrio parahaemolyticus in a Tidal Estuary. J Mar Biol Oceanogr 2:2. doi:10.4172/2324-8661.1000109
We examined the influence of tide stage and depth on the
abundance of Vibrio vulnificus and Vibrio parahaemolyticus in the
Chesapeake Bay. Samples were collected every 3 hours following
predicted tides at a fixed location over 3 separate days and Vibrio
concentrations analyzed by qPCR. Multi-way Analysis of Variance
suggest that sampling day explains the vast majority of the variance
in abundance for both species (p<0.0001) with limited influence
of tide and depth. The physio-chemical parameters that define a
sampling day were further explored with environmental gradient
analysis. Gradients in daily photosynthetic activity and turbidity
(PC1) and temperature and salinity (PC2) explained 75% of the
environmental variability, and 50% of Vibrio vulnificus abundance.
However, these same gradients did not explain a significant
proportion of variation in the abundance of Vibrio parahaemolyticus
(P>0.05). These results suggest that within day variability is not
as important as that associated with environmental changes
over time, and further highlight the need for species specific and
mechanistic approaches to the study of vibrio ecology.
Vibrio spp.; Ecology; Monitoring; qPCR
Vibrio spp. are gram negative, flagellated, heterotrophic bacteria
indigenous to the estuarine environment. Several species, including
V. cholerae, V. vulnificus, and V. parahaemolyticus are capable of
causing severe and occasionally life threatening infections in humans
both through water contact and consumption [1,2]. For example, V.
vulnificus is responsible for 95% of all seafood related mortalities,
and carries a 50% mortality rate with primary septicemia . Over
the past three years of data availability (2007-2009), an average of
over 650 cases have been reported annually through the Center for
Disease Control’s COVIS system . Further, there is evidence that
the incidence has increased . Accordingly, significant effort has
been devoted to monitoring Vibrio abundance, enhancing detection
and differentiation methods, and developing predictive models [6-
15]. Many studies have examined the relationship between Vibrio
abundance and environmental factors [7-9,12,13,16-23]. It is clear
from these efforts that temperature is the major temporal driver, while salinity, chlorophyll, and zooplankton govern spatial variability
and may differ in importance among species. For example, Vibrio
vulnificus is strongly governed by preferred salinity in Chesapeake
Bay [8,23], while Vibrio parahaemolyticus occurs in a wider salinity
range with attempts to examine correlation equivocal [23-25].
However, what is less clear is the influence of short term variability
perhaps influenced by factors such as tidal stage and depth of sample
collection. These may be of particular significance to large scale, coastal
water monitoring efforts which routinely collect surface samples and
are unable to control for tide. Previous research found tide, depth,
and day to all significantly influence Vibrio’s in a Florida estuary,
however no attempt was made to identify individual species .
Thus the aim of this study was to determine what factors influence
the concentrations of Vibrio vulnificus and Vibrio parahaemolyticus over short time scales.
Materials and Methods
Field sampling was conducted on the Tred Avon River at the
Cooperative Oxford Lab (Oxford, MD). Water samples were collected
at the surface, 0.1 meters above the bottom, and the mid depth of the
water column (3 meter total depth) over two complete tidal cycles
(Low AM, Flood, High, Ebb, Low PM) and was replicated over the
course of three days in July and August of 2010 (n=45). Surface water
samples were collected by submerging autoclaved Nalgene bottes and
rinsing 2x times prior to sample collection. Sub-surface samples were
collected with a Niskin bottle following the same rinsing procedure at
depth. The samples were thoroughly mixed and 200 ml of water was
filtered through 0.22um sterivex filters, all water removed, and stored
at -80°C, and DNA extracted as previously described . Physical
water quality parameters were measured concurrently at each depth
in-situ with a YSI datasonde (YSI Incorporated, Yellow Springs,
Ohio, USA). Parameters measured are listed in Table 1.
Table 1: Physio-chemical water parameters measured in this study by sampling
day. Mean and standard deviation are presented.
Primers tlh F (5’-ACTCAACACAAGA AGAGATCGACAA-3’)
and tlh_R (5’-GATGAGCGGTTGATGTCCAA-3’) were used in
conjunction with the probe tlh_TXRD (5’- /TxRED/CGCTCGCGTTCACGAAA
CCGT/3BHQ_2/-3’) for the detection of V. parahaemolyticus.
A unique internal control (IC) was incorporated simultaneously
into the assay to test for the presence and influence of
inhibitors . Primers vvh_F (5’-TTCCAACTTCA AACCGAACTATGA-
3’) and vvh_R (5’-TTCCAGTCGATGCGAATACGTTG-3’)
were used in conjunction with the probe vvh874 (5’-/56- FAM/ AACTATCGTGCAC
GCTTTGGTACCGT /3BHQ_1/-3’) for the detection
of V. vulnificus . A unique internal control was also incorporated
into this assay .
Assay performance testing was carried out in a manner similar
to that as previously described . Primers and probe were
tested against strains of Vibrio vulnificus, V. parahaemolyticus,
Enterococcus faecium, Hematodinium spp. and 17 species of the genus
Mycobacterium. In both cases the primers and probes were specific
only to the organism of interest and negative results were obtained for
all other species. Recovery and repeatability estimates as well as the
effects of freezing are previously published . Standard curves of Ct
values versus concentration yielded an assay efficiency of 104.66% ±
0.96 (standard deviation; n=3) and 88.60% ± 5.90 (standard deviation; n=4) for Vp and Vv (respectively). Detection limits of the assays are
48 and 190 CFU/200ml as determined from spiked water samples. No
inhibitors were observed in any environmental samples, based on the
amplification of the internal control.
Three-way Analysis of Variance was used to examine the influence
of sampling day, tidal stage, sampling depth, and interactions on log
transformed count data for each of the two pathogens (Proc GLM, Sas
Inc., Carey NC). For the purpose of this study the term ‘day” is used
to describe the physio-chemical variables and their measurements
within a given 24 hour period. Where necessary, least square
means comparisons were used to determine significance within a
factor (LSmeans procedure, pdiff option, Tukey’s adjustment). All
models were examined for normality and homogeneity of variance.
Spearmans rank correlation analysis was employed to evaluate
relationships between environmental variables and Vibrio abundance
as an initial approach to describe variability among sampling
days. Because of a high degree of collinearity among water quality
variables, Principal Component Analysis (PCA) was employed to
describe the environmental gradients present during the study, and
scores subsequently used as composite variables in multiple linear
Results and Discussion
The main intent of this effort was to examine the relative
importance of within-day variability on Vibrio spp. concentration,
with the particular hypothesis that they may be influence by tide
and depth. Tide did not significantly influence the abundance of
V. vulnificus, but accounted for 10% of the total variance for V.
parahaemolyticus (Table 1). However, this finding is somewhat
misleading as the interaction of tide and day is most significant
(Table 1B, 20% of total variance), particularly the August 16th flood
tide event (Figure 1). Similar results have been previously reported
for sucrose negative Vibrios with respect to tide . Depth similarly
accounted for a minority of the variance but was significant for both
species (Table 1). Concentrations obtained from samples at the
bottom of the water column were 30% greater on average for V.
vulnificus than those from the middle and surface (Figure 2) with a
similar trend noted for V. parahaemolyticus (Figure 1). These results
are supported by others  and may represent re-suspension from
sediments, which are known to act as a reservoir for several species
of Vibrios .
Figure 1:Vibrio vulnificus abundance by tide, day and depth.
Figure 2: Vibrio parahaemolyticus abundance by tide, day, and depth.
The vast majority of the total variance in both V. vulnificus
(p, 0.0001, f=54.02, 2df) and V. parahaemolyticus concentrations
(p<0.0001, f =6.076, 2df) was explained by the physio-chemical
environment that defines a sampling day (Tables 1 and 2, Figures 1
and 2). In this study, a variety of water quality measurements were obtained simultaneously to offer further explanatory variables for
use in describing environmental drivers of vibrio abundance and
discriminating among sampling days. As is typical with measures of
water quality, however, many of these variables are highly correlated
with each other. Indeed, multiple regression analysis initially
attempted with our data proved futile due to variance inflation
associated with multicollinearity. Principal component analysis
is a particularly valuable data reduction technique in multi-variate
analysis where many of the explanatory variables are co-correlated.
The orthogonal transformation results in a series of principal
components that are linearly uncorrelated and defined in a manner
where the majority of the variance is explained by the first component,
and less by each successive component. The resulting variables serve
as composite explanatory variables for further data exploration.
Table 2: Analysis of variance for fixed effects of day, depth, tide and interactions
for Vibrio vulnificus (A) and Vibrio parahaemolyticus (B).
Ordination clearly demonstrates the difference between sampling
days in the physio-chemical environment (Figure 3). The majority of
variance in the environment was explained by the first two principal
components (75%). PC1 (horizontal gradient, Figure 3) represents
45% of the total variance and describes a gradient in photosynthesis
and turbidity as defined by increasing concentration of chlorophyll,
dissolved oxygen, and pH. These factors, as well as PC1 in general
are increasing during the course of the day in concert with algal
production and respiration (PC1 correlation, r=0.49, P=0.001). PC2
(vertical gradient) represents 30% of the environmental variation
and represents a gradient in temperature and salinity. As a whole,
individual sampling days are clearly differentiated on PC2 by
variation in temperature and salinity, and to a lesser extent PC1.
Figure 3: Environmental gradient analysis of the physio-chemical patterns
associated with each sampling day. Green = July 19, Blue = August 3, and
Red = August 16. Asterisks (*) represent either Vv (panel A) or Vp (panel B)
samples which exceed the median concentration obtained in this study (8 and
2 CFU/ml respectively).
By overlaying vibrio abundance on the environmental
gradient analysis, we demonstrate that the two species respond
in a dissimilar fashion. The warmer, less saline, and more turbid
environments (Figure 3A) nearly always supported Vibrio vulnificus
at concentrations greater than the median concentration (8 CFU/
ml) with a decreasing proportion associated with cooler, more saline
days. Regression analysis using PC1 and PC2 as explanatory variables
explains 49% of the variation (P <0.0001) in Vv abundance as follows:
LnVv=1.90 – (0.45 x PC1) + (0.52 x PC2). We have previously
demonstrated that the presence of V. vulnificus declines rapidly with
distance from optimal salinity (11.5 ppt) in Chesapeake Bay , and
similar results were obtained here even within a relatively narrow
salinity range (11.4 – 12.8ppt). This is supported by other work noting
a general preference of 5-15 ppt in Chesapeake Bay and other coastal
systems [9,23]. Turbidity is also an important component of current models used to forecast the distribution and abundance of Vv in the
Chesapeake system (Authors unpublished data).
Vibrio parahaemolyticus did not demonstrate the same degree
of association with environmental gradients described in this study
(Figure 3B). Regression analysis using PC1 and PC2 demonstrates a
stronger, negative association with PC1 than PC2, however the model
is insignificant and explains a paucity of variance (LnVp=1.04 –(0.12
x PC1) – (0.02 x PC2), P=0.34, R2=0.05). While temperature is clearly
a driver of Vp abundance seasonally, it is clear that other factors not
directly measured in this effort are influential .
While of interest in describing the general ecology of V.
parahaemolyticus and V. vulnificus, the significance of this effort is
in understanding site specific variability on short time frames and its
application to monitoring and forecast efforts. This work confirms that
changes in physio-chemical variables over time exert a much stronger
influence on V. vulnificus and V. parahaemolyticus abundance than
within-day variability associated with tides or sampling depth. This
is important information for monitoring programs and modeling
efforts in that a daily sample or prediction, regardless of depth or tide taken, is generally representative of overall conditions at that
location for the given day. However, the study also points to the
need for moving towards a more mechanistic understanding of the
ecology of Vibrio spp and relationships with the ecosystem as a whole.
While general habitat preferences have been well defined for many
species and described through empirical relationships, detailed study
of the interactions with the microbial community, other biota, and
ecological processes over short time scales may greatly improve our
understanding of the ecology and dynamics of Vibrio spp.
This work was supported by a summer internship and funding provided by
NOAA’s Teacher in the Lab Program. We are grateful for the efforts of Jennifer
Hammond and LeeAnn Hutchinson in arranging this opportunity for teacher
education and collaborative research.
Oliver JD, Kaper JB (2001) Food microbiology: fundamentals and frontiers. ASM Press, Washington, DC.