Isolation and Diversity of Actinomycetes in the Chesapeake Bay

Chesapeake Bay was investigated as a source of actinomycetes to screen for production of novel bioactive compounds. The presence of relatively large populations of actinoplanetes (chemotype II/D actinomycetes) in Chesapeake Bay sediment samples indicates that it is an eminently suitable ecosystem from which to isolate actinomycetes for screening programs. Actinomycetes were isolated from sediment samples collected in Chesapeake Bay with an isolation medium containing nalidixic acid, which proved to be more effective than heat pretreatment of samples. Actinomycete counts ranged from a high of 1.4 × 10⁵ to a low of 1.8 × 10² CFU/ml of sediment. Actinomycetes constituted 0.15 to 8.63% of the culturable microbial community. The majority of isolates from the eight stations studied were actinoplanetes (i.e., chemotype II/D), and 249 of these isolates were obtained in a total of 298 actinomycete isolates. Antimicrobial activity profiles indicated that diverse populations of actinoplanetes were present at each station. DNA hybridization studies showed considerable diversity among isolates between stations, but indicated that actinoplanete strains making up populations at nearby stations were more similar to each other than to populations sampled at distant stations. The diversity of actinoplanetes and the ease with which these organisms were isolated from Chesapeake Bay sediments make this a useful source of these actinomycetes.     

Effects of replacing the unreliable cDNA microarray measurements on the disease classification based on gene expression profiles and functional modules

MOTIVATION: Microarrays datasets frequently contain a large number of missing values (MVs), which need to be estimated and replaced for subsequent data mining. The focus of the paper is to study the effects of different MV treatments for cDNA microarray data on disease classification analysis. RESULTS: By analyzing five datasets, we demonstrate that among three kinds of classifiers evaluated in this study, support vector machine (SVM) classifiers are robust to varied MV imputation methods [e.g. replacing MVs by zero, K nearest-neighbor (KNN) imputation algorithm, local least square imputation and Bayesian principal component analysis], while the classification and regression tree classifiers are sensitive in terms of classification accuracy. The KNNclassifiers built on differentially expressed genes (DEGs) are robust to the varied MV treatments, but the performances of the KNN classifiers based on all measured genes can be significantly deteriorated when imputing MVs for genes with larger missing rate (MR) (e.g. MR > 5%). Generally, while replacing MVs by zero performs relatively poor, the other imputation algorithms have little difference in affecting classification performances of the SVM or KNN classifiers. We further demonstrate the power and feasibility of our recently proposed functional expression profile (FEP) approach as means to handle microarray data with MVs. The FEPs, which are derived from the functional modules that are enriched with sets of DEGs and thus can be consistently identified under varied MV treatments, achieve precise disease classification with better biological interpretation. We conclude that the choice of MV treatments should be determined in context of the later approaches used for disease classification. The suggested exclusion criterion of ignoring the genes with larger MR (e.g. >5%), while justifiable for some classifiers such as KNN classifiers, might not be considered as a general rule for all classifiers.     

Sediment denitrifier community composition and nirS gene expression investigated with functional gene microarrays

A functional gene microarray was used to investigate denitrifier community composition and nitrite reductase (nirS) gene expression in sediments along the estuarine gradient in Chesapeake Bay, USA. The nirS oligonucleotide probe set was designed to represent a sequence database containing 539 Chesapeake Bay clones, as well as sequences from many other environments. Greatest nirS diversity was detected at the freshwater station at the head of the bay and least diversity at the higher salinity station near the mouth of the Bay. The most common OTUs from the sequence database were detected on the array with high signal strength in most samples. One of the most abundant OTUs, CB2-S-138, was identified as dominant at the mid-bay site by both microarray and quantitative PCR assays, but it comprised a much smaller fraction of the assemblage in the north and south bay samples. cDNA (transcribed from total RNA extracts) targets were hybridized to the same array to compare the profiles of community composition at the DNA (relative abundance) and mRNA (gene expression) levels. Only the three dominant denitrifying groups (in terms of relative strength of DNA hybridization signal) were detected at the mRNA level. These results suggest that the most actively denitrifying groups are responsible for most nirS expression as well.     

Functional prediction of unidentified lipids using supervised classifiers

Mass spectrometry (MS)-based metabolomics studies often require handling of both identified and unidentified metabolite data. In order to avoid bias in data interpretation, it would be of advantage for the data analysis to include all available data. A practical challenge in exploratory metabolomics analysis is therefore how to interpret the changes related to unidentified peaks. In this paper, we address the challenge by predicting the class membership of unknown peaks by applying and comparing multiple supervised classifiers to selected lipidomics datasets. The employed classifiers include k-nearest neighbours (k-NN), support vector machines (SVM), partial least squares and discriminant analysis (PLS-DA) and Naive Bayes methods which are known to be effective and efficient in predicting the labels for unseen data. Here, the class label predictions are sought for unidentified lipid profiles coming from high throughput global screening in Ultra Performance Liquid Chromatography Mass Spectrometry (UPLC^TM/MS) experimental setup. Our investigation reveals that k-NN and SVM classifiers outperform both PLS-DA and Naive Bayes classifiers. Naive Bayes classifier perform poorly among all models and this observation seems logical as lipids are highly co-regulated and do not respect Naive Bayes assumptions of features being conditionally independent given the class. Common label predictions from k-NN and SVM can serve as a good starting point to explore full data and thereby facilitating exploratory studies where label information is critical for the data interpretation.     

Comparisons of different hypervariable regions of rrs genes for use in fingerprinting of microbial communities by PCR-denaturing gradient gel electrophoresis

Denaturing gradient gel electrophoresis (DGGE) has become a widely used tool to examine microbial diversity and community structure, but no systematic comparison has been made of the DGGE profiles obtained when different hypervariable (V) regions are amplified from the same community DNA samples. We report here a study to make such comparisons and establish a preferred choice of V region(s) to examine by DGGE, when community DNA extracted from samples of digesta is used. When the members of the phylogenetically representative set of 218 rrs genes archived in the RDP II database were compared, the V1 region was found to be the most variable, followed by the V9 and V3 regions. The temperature of the lowest-melting-temperature (T(m(L))) domain for each V region was also calculated for these rrs genes, and the V1 to V4 region was found to be most heterogeneous with respect to T(m(L)). The average T(m(L)) values and their standard deviations for each V region were then used to devise the denaturing gradients suitable for separating 95% of all the sequences, and the PCR-DGGE profiles produced from the same community DNA samples with these conditions were compared. The resulting DGGE profiles were substantially different in terms of the number, resolution, and relative intensity of the amplification products. The DGGE profiles of the V3 region were best, and the V3 to V5 and V6 to V8 regions produced better DGGE profiles than did other multiple V-region amplicons. Introduction of degenerate bases in the primers used to amplify the V1 or V3 region alone did not improve DGGE banding profiles. Our results show that DGGE analysis of gastrointestinal microbiomes is best accomplished by the amplification of either the V3 or V1 region of rrs genes, but if a longer amplification product is desired, then the V3 to V5 or V6 to V8 region should be targeted.     

An assessment of the hypervariable domains of the 16S rRNA genes for their value in determining microbial community diversity: the paradox of traditional ecological indices

Amplicon length heterogeneity PCR (LH-PCR) was investigated for its ability to distinguish between microbial community patterns from the same soil type under different land management practices. Natural sagebrush and irrigated mouldboard-ploughed soils from Idaho were queried as to which hypervariable domains, or combinations of 16S rRNA gene domains, were the best molecular markers. Using standard ecological indices to measure richness, diversity and evenness, the combination of three domains, V1, V3 and V1+V2, or the combined V1 and V3 domains were the markers that could best distinguish the undisturbed natural sagebrush communities from the mouldboard-ploughed microbial communities. Bray-Curtis similarity and multidimensional scaling were found to be better metrics to ordinate and cluster the LH-PCR community profiling data. The use/misuse of traditional ecological indices such as diversity and evenness to study microbial community profiles will remain a major point to consider when performing metagenomic studies.     

Predictability of Vibrio cholerae in Chesapeake Bay

Vibrio cholerae is autochthonous to natural waters and can pose a health risk when it is consumed via untreated water or contaminated shellfish. The correlation between the occurrence of V. cholerae in Chesapeake Bay and environmental factors was investigated over a 3-year period. Water and plankton samples were collected monthly from five shore sampling sites in northern Chesapeake Bay (January 1998 to February 2000) and from research cruise stations on a north-south transect (summers of 1999 and 2000). Enrichment was used to detect culturable V. cholerae, and 21.1% (n = 427) of the samples were positive. As determined by serology tests, the isolates, did not belong to serogroup O1 or O139 associated with cholera epidemics. A direct fluorescent-antibody assay was used to detect V. cholerae O1, and 23.8% (n = 412) of the samples were positive. V. cholerae was more frequently detected during the warmer months and in northern Chesapeake Bay, where the salinity is lower. Statistical models successfully predicted the presence of V. cholerae as a function of water temperature and salinity. Temperatures above 19 degrees C and salinities between 2 and 14 ppt yielded at least a fourfold increase in the number of detectable V. cholerae. The results suggest that salinity variation in Chesapeake Bay or other parameters associated with Susquehanna River inflow contribute to the variability in the occurrence of V. cholerae and that salinity is a useful indicator. Under scenarios of global climate change, increased climate variability, accompanied by higher stream flow rates and warmer temperatures, could favor conditions that increase the occurrence of V. cholerae in Chesapeake Bay.     

Semi-Supervised Projective Non-Negative Matrix Factorization for Cancer Classification

Advances in DNA microarray technologies have made gene expression profiles a significant candidate in identifying different types of cancers. Traditional learning-based cancer identification methods utilize labeled samples to train a classifier, but they are inconvenient for practical application because labels are quite expensive in the clinical cancer research community. This paper proposes a semi-supervised projective non-negative matrix factorization method (Semi-PNMF) to learn an effective classifier from both labeled and unlabeled samples, thus boosting subsequent cancer classification performance. In particular, Semi-PNMF jointly learns a non-negative subspace from concatenated labeled and unlabeled samples and indicates classes by the positions of the maximum entries of their coefficients. Because Semi-PNMF incorporates statistical information from the large volume of unlabeled samples in the learned subspace, it can learn more representative subspaces and boost classification performance. We developed a multiplicative update rule (MUR) to optimize Semi-PNMF and proved its convergence. The experimental results of cancer classification for two multiclass cancer gene expression profile datasets show that Semi-PNMF outperforms the representative methods.     

Profile-based string kernels for remote homology detection and motif extraction

We introduce novel profile-based string kernels for use with support vector machines (SVMs) for the problems of protein classification and remote homology detection. These kernels use probabilistic profiles, such as those produced by the PSI-BLAST algorithm, to define position-dependent mutation neighborhoods along protein sequences for inexact matching of k-length subsequences ("k-mers") in the data. By use of an efficient data structure, the kernels are fast to compute once the profiles have been obtained. For example, the time needed to run PSI-BLAST in order to build the profiles is significantly longer than both the kernel computation time and the SVM training time. We present remote homology detection experiments based on the SCOP database where we show that profile-based string kernels used with SVM classifiers strongly outperform all recently presented supervised SVM methods. We further examine how to incorporate predicted secondary structure information into the profile kernel to obtain a small but significant performance improvement. We also show how we can use the learned SVM classifier to extract "discriminative sequence motifs"--short regions of the original profile that contribute almost all the weight of the SVM classification score--and show that these discriminative motifs correspond to meaningful structural features in the protein data. The use of PSI-BLAST profiles can be seen as a semi-supervised learning technique, since PSI-BLAST leverages unlabeled data from a large sequence database to build more informative profiles. Recently presented "cluster kernels" give general semi-supervised methods for improving SVM protein classification performance. We show that our profile kernel results also outperform cluster kernels while providing much better scalability to large datasets.     

Comparisons of Different Hypervariable Regions of rrs Genes for Use in Fingerprinting of Microbial Communities by PCR-Denaturing Gradient Gel Electrophoresis

Denaturing gradient gel electrophoresis (DGGE) has become a widely used tool to examine microbial diversity and community structure, but no systematic comparison has been made of the DGGE profiles obtained when different hypervariable (V) regions are amplified from the same community DNA samples. We report here a study to make such comparisons and establish a preferred choice of V region(s) to examine by DGGE, when community DNA extracted from samples of digesta is used. When the members of the phylogenetically representative set of 218 rrs genes archived in the RDP II database were compared, the V1 region was found to be the most variable, followed by the V9 and V3 regions. The temperature of the lowest-melting-temperature (T_m(L)) domain for each V region was also calculated for these rrs genes, and the V1 to V4 region was found to be most heterogeneous with respect to T_m(L). The average T_m(L) values and their standard deviations for each V region were then used to devise the denaturing gradients suitable for separating 95% of all the sequences, and the PCR-DGGE profiles produced from the same community DNA samples with these conditions were compared. The resulting DGGE profiles were substantially different in terms of the number, resolution, and relative intensity of the amplification products. The DGGE profiles of the V3 region were best, and the V3 to V5 and V6 to V8 regions produced better DGGE profiles than did other multiple V-region amplicons. Introduction of degenerate bases in the primers used to amplify the V1 or V3 region alone did not improve DGGE banding profiles. Our results show that DGGE analysis of gastrointestinal microbiomes is best accomplished by the amplification of either the V3 or V1 region of rrs genes, but if a longer amplification product is desired, then the V3 to V5 or V6 to V8 region should be targeted.     

Ecology, serology, and enterotoxin production of Vibrio cholerae in Chesapeake Bay.

A total of 65 isolates of Vibrio cholerae, serotypes other than O--1, have been recovered from water, sediment, and shellfish samples from the Chesapeake Bay. Isolations were not random, but followed a distinct pattern in which salinity appeared to be a controlling factor in V. cholerae distribution. Water salinity at stations yielding V. cholerae (13 out of 21 stations) was 4 to 17 0/00, whereas the salinity of water at stations from which V. cholerae organisms were not isolated was less than 4 or greater than 17 0/00. From results of statistical analyses, no correlation between incidence of fecal coliforms and V. cholerae could be detected, whereas incidence of Salmonella species, measured concurrently, was clearly correlated with fecal coliforms, with Salmonella isolated only in areas of high fecal coliform levels. A seasonal cycle could not be determined since strains of V. cholerae were detectable at low levels (ca. 1 to 10 cells/liter) throughout the year. Although none of the Chesapeake Bay isolates was agglutinable in V. cholerae O group 1 antiserum, the majority for Y-1 adrenal cells. Furthermore, rabbit ileal loop and mouse lethality tests were also positive for the Chesapeake Bay isolates, with average fluid accumulation in positive ileal loops ranging from 0.21 to 2.11 ml/cm. Serotypes of the strains of V. cholerae recovered from Chesapeake Bay were those of wide geographic distribution. It is concluded from the data assembled to date, that V. cholerae is an autochthonous estuarine bacterial species resident in Chesapeake Bay.     

Multi-scale AM–FM analysis for the classification of surface electromyographic signals

In this work, multi-scale amplitude modulation–frequency modulation (AM–FM) features are extracted from surface electromyographic (SEMG) signals and they are used for the classification of neuromuscular disorders. The method is validated on SEMG signals recorded from a total of 40 subjects: 20 normal and 20 abnormal cases (11 myopathy, and 9 neuropathy cases), at 10%, 30%, 50%, 70% and 100% of maximum voluntary contraction (MVC), from the biceps brachii muscle. For the classification, three classifiers are used: (i) the statistical K-nearest neighbor (KNN), (ii) the self-organizing map (SOM) and (iii) the support vector machine (SVM). For all classifiers, the leave-one-out methodology is used to validate the classification of the SEMG signals into normal or abnormal (myopathy or neuropathy). A classification success rate of 78% for the AM–FM features and SVM models was achieved. These results also show that SEMG can be used as a non-invasive alternative to needle EMG for differentiating between normal and abnormal (myopathy, or neuropathy) cases.     

Succession of microbial communities during hot composting as detected by PCR-single-strand-conformation polymorphism-based genetic profiles of small-subunit rRNA genes

A cultivation-independent technique for genetic profiling of PCR-amplified small-subunit rRNA genes (SSU rDNA) was chosen to characterize the diversity and succession of microbial communities during composting of an organic agricultural substrate. PCR amplifications were performed with DNA directly extracted from compost samples and with primers targeting either (i) the V4-V5 region of eubacterial 16S rRNA genes, (ii) the V3 region in the 16S rRNA genes of actinomycetes, or (iii) the V8-V9 region of fungal 18S rRNA genes. Homologous PCR products were converted to single-stranded DNA molecules by exonuclease digestion and were subsequently electrophoretically separated by their single-strand-conformation polymorphism (SSCP). Genetic profiles obtained by this technique showed a succession and increasing diversity of microbial populations with all primers. A total of 19 single products were isolated from the profiles by PCR reamplification and cloning. DNA sequencing of these molecular isolates showed similarities in the range of 92.3 to 100% to known gram-positive bacteria with a low or high G+C DNA content and to the SSU rDNA of gamma-Proteobacteria. The amplified 18S rRNA gene sequences were related to the respective gene regions of Candida krusei and Candida tropicalis. Specific molecular isolates could be attributed to different composting stages. The diversity of cultivated bacteria isolated from samples taken at the end of the composting process was low. A total of 290 isolates were related to only 6 different species. Two or three of these species were also detectable in the SSCP community profiles. Our study indicates that community SSCP profiles can be highly useful for the monitoring of bacterial diversity and community successions in a biotechnologically relevant process.     

Improved serial analysis of V1 ribosomal sequence tags (SARST-V1) provides a rapid, comprehensive, sequence-based characterization of bacterial diversity and community composition

Serial analysis of ribosomal sequence tags (SARST) is a recently developed technology that can generate large 16S rRNA gene (rrs) sequence data sets from microbiomes, but there are numerous enzymatic and purification steps required to construct the ribosomal sequence tag (RST) clone libraries. We report here an improved SARST method, which still targets the V1 hypervariable region of rrs genes, but reduces the number of enzymes, oligonucleotides, reagents, and technical steps needed to produce the RST clone libraries. The new method, hereafter referred to as SARST-V1, was used to examine the eubacterial diversity present in community DNA recovered from the microbiome resident in the ovine rumen. The 190 sequenced clones contained 1055 RSTs and no less than 236 unique phylotypes (based on > or = 95% sequence identity) that were assigned to eight different eubacterial phyla. Rarefaction and monomolecular curve analyses predicted that the complete RST clone library contains 99% of the 353 unique phylotypes predicted to exist in this microbiome. When compared with ribosomal intergenic spacer analysis (RISA) of the same community DNA sample, as well as a compilation of nine previously published conventional rrs clone libraries prepared from the same type of samples, the RST clone library provided a more comprehensive characterization of the eubacterial diversity present in rumen microbiomes. As such, SARST-V1 should be a useful tool applicable to comprehensive examination of diversity and composition in microbiomes and offers an affordable, sequence-based method for diversity analysis.     

SlimPLS: A Method for Feature Selection in Gene Expression-Based Disease Classification

A major challenge in biomedical studies in recent years has been the classification of gene expression profiles into categories, such as cases and controls. This is done by first training a classifier by using a labeled training set containing labeled samples from the two populations, and then using that classifier to predict the labels of new samples. Such predictions have recently been shown to improve the diagnosis and treatment selection practices for several diseases. This procedure is complicated, however, by the high dimensionality if the data. While microarrays can measure the levels of thousands of genes per sample, case-control microarray studies usually involve no more than several dozen samples. Standard classifiers do not work well in these situations where the number of features (gene expression levels measured in these microarrays) far exceeds the number of samples. Selecting only the features that are most relevant for discriminating between the two categories can help construct better classifiers, in terms of both accuracy and efficiency. In this work we developed a novel method for multivariate feature selection based on the Partial Least Squares algorithm. We compared the method''s variants with common feature selection techniques across a large number of real case-control datasets, using several classifiers. We demonstrate the advantages of the method and the preferable combinations of classifier and feature selection technique.     

Partially supervised learning using an EM-boosting algorithm

Training data in a supervised learning problem consist of the class label and its potential predictors for a set of observations. Constructing effective classifiers from training data is the goal of supervised learning. In biomedical sciences and other scientific applications, class labels may be subject to errors. We consider a setting where there are two classes but observations with labels corresponding to one of the classes may in fact be mislabeled. The application concerns the use of protein mass-spectrometry data to discriminate between serum samples from cancer and noncancer patients. The patients in the training set are classified on the basis of tissue biopsy. Although biopsy is 100% specific in the sense that a tissue that shows itself to have malignant cells is certainly cancer, it is less than 100% sensitive. Reference gold standards that are subject to this special type of misclassification due to imperfect diagnosis certainty arise in many fields. We consider the development of a supervised learning algorithm under these conditions and refer to it as partially supervised learning. Boosting is a supervised learning algorithm geared toward high-dimensional predictor data, such as those generated in protein mass-spectrometry. We propose a modification of the boosting algorithm for partially supervised learning. The proposal is to view the true class membership of the samples that are labeled with the error-prone class label as missing data, and apply an algorithm related to the EM algorithm for minimization of a loss function. To assess the usefulness of the proposed method, we artificially mislabeled a subset of samples and applied the original and EM-modified boosting (EM-Boost) algorithms for comparison. Notable improvements in misclassification rates are observed with EM-Boost.     

Incidence of Vibrio parahaemolyticus in Chesapeake Bay

A Bay-wide survey of the distribution of Vibrio parahaemolyticus was carried out in Chesapeake Bay during May 1972, to determine whether the annual cycle of V. parahaemolyticus which was observed to occur in the Rhode River subestuary of Chesapeake Bay took place in other parts of Chesapeake Bay. In an earlier study, April to early June, when the water temperature rises from 14 to 19 C, was found to be a critical period in the annual cycle of the organism in the Rhode River, since this is the time period when the annual cycle is initiated. Results of this study, however, revealed that V. parahaemolyticus could not be found in the water column during May 1972. Nevertheless, several samples of sediment and plankton yielded V. parahaemolyticus isolates. Comparison of data with those for the Rhode River area examined in the earlier studies of the annual cycle of V. parahaemolyticus suggests that the time of initiation of the annual cycle of V. parahaemolyticus in the open Bay proper may be influenced by various factors such as temperature and salinity, i.e., deeper water locations may show initiation of the V. parahaemolyticus annual cycle later than shallow areas. Confirmation of the presence of the organisms in the samples studied was accomplished using numerical taxonomy with 19 reference strains also included in the analyses.     

Depth-related patterns in benthic community condition along an estuarine gradient in Chesapeake Bay,USA

We tested whether macrobenthic community condition varies significantly with water depth in a variety of regions of Chesapeake Bay, USA. Benthic community condition was characterized using the Benthic Index of Biotic Integrity (B-IBI) previously developed for the Bay. We applied two water depth thresholds intended to emphasize the ecological importance and/or anthropogenic impacts upon shallow-water regions. The first threshold of 2 m emphasizes restoring and supporting submerged aquatic vegetation while the second threshold of 4 m emphasizes the zone of maximum anthropogenic impact upon natural ecosystem functions. An a priori expectation is that benthic community condition may worsen with increasing depth, specifically in regions (1) where water column stratification at depth results in prolonged low dissolved oxygen levels or (2) where net deposition at depth results in higher levels of hydrophobic, sediment-bound contaminants. Samples collected from a major tributary of Chesapeake Bay, the York River estuary, spanned the entire salinity range from tidal freshwater to polyhaline. We also tested the shallow-water depth thresholds using data from the Virginia Mainstem of Chesapeake Bay and the Southern Branch of the Elizabeth River. These two polyhaline regions are characterized as having the best and worst benthic community condition in Chesapeake Bay. At the scale of the entire tidal York River system, there were no significant differences in benthic community condition with water depth. However, two salinity regions, low mesohaline and polyhaline, had significant depth effects with the shallowest water depth zone significantly different from the other two depth regions. For the low mesohaline region benthic community condition was worse at the shallowest depth and for the polyhaline region the shallowest depth was better comparing the three depth regions. No depth-related differences in the B-IBI were found for the two additional Chesapeake Bay strata, the Virginia Mainstem characterized with the lowest levels of benthic community degradation and for the Southern branch of the Elizabeth River, characterized by the highest levels of benthic community degradation. We conclude that the ecological state of Chesapeake Bay subtidal benthic communities is adequately characterized by randomly sampling all depths without further stratification into shallow and deeper regions.     

Predictability of Vibrio cholerae in Chesapeake Bay

Vibrio cholerae is autochthonous to natural waters and can pose a health risk when it is consumed via untreated water or contaminated shellfish. The correlation between the occurrence of V. cholerae in Chesapeake Bay and environmental factors was investigated over a 3-year period. Water and plankton samples were collected monthly from five shore sampling sites in northern Chesapeake Bay (January 1998 to February 2000) and from research cruise stations on a north-south transect (summers of 1999 and 2000). Enrichment was used to detect culturable V. cholerae, and 21.1% (n = 427) of the samples were positive. As determined by serology tests, the isolates, did not belong to serogroup O1 or O139 associated with cholera epidemics. A direct fluorescent-antibody assay was used to detect V. cholerae O1, and 23.8% (n = 412) of the samples were positive. V. cholerae was more frequently detected during the warmer months and in northern Chesapeake Bay, where the salinity is lower. Statistical models successfully predicted the presence of V. cholerae as a function of water temperature and salinity. Temperatures above 19°C and salinities between 2 and 14 ppt yielded at least a fourfold increase in the number of detectable V. cholerae. The results suggest that salinity variation in Chesapeake Bay or other parameters associated with Susquehanna River inflow contribute to the variability in the occurrence of V. cholerae and that salinity is a useful indicator. Under scenarios of global climate change, increased climate variability, accompanied by higher stream flow rates and warmer temperatures, could favor conditions that increase the occurrence of V. cholerae in Chesapeake Bay.     

Uncertainty in Model Predictions of Vibrio vulnificus Response to Climate Variability and Change: A Chesapeake Bay Case Study

The effect that climate change and variability will have on waterborne bacteria is a topic of increasing concern for coastal ecosystems, including the Chesapeake Bay. Surface water temperature trends in the Bay indicate a warming pattern of roughly 0.3–0.4°C per decade over the past 30 years. It is unclear what impact future warming will have on pathogens currently found in the Bay, including Vibrio spp. Using historical environmental data, combined with three different statistical models of Vibrio vulnificus probability, we explore the relationship between environmental change and predicted Vibrio vulnificus presence in the upper Chesapeake Bay. We find that the predicted response of V. vulnificus probability to high temperatures in the Bay differs systematically between models of differing structure. As existing publicly available datasets are inadequate to determine which model structure is most appropriate, the impact of climatic change on the probability of V. vulnificus presence in the Chesapeake Bay remains uncertain. This result points to the challenge of characterizing climate sensitivity of ecological systems in which data are sparse and only statistical models of ecological sensitivity exist.