Academia.eduAcademia.edu

Figure 11 – uploaded by Michael Mallin

See full PDFdownloadDownload figure
Fig. 11. Principal component plots of biota abundance considering paired tanks with similar ballasting history (A, phytoplankton; C, bacteria) vs. “unpaired”’ tanks with different ballasting history (B, phytoplankton; D, bacteria). Paired tanks were from voyages 1, 2,3, 4, 7,9, 10, 12, 14, 15, 18, 22, and 28; voyage numbers are shown, with the second member of each pair in red. For (B and D), the following tanks were included: voyages 11, 16, 19, 20, 25, and 26, each with two tanks; voyage 24 with three tanks; voyages 17 and 27 each with 4 tanks; and voyage 6 with 8 tanks. In (B), the lower right cluster includes six tanks from voyage 6, four tanks from voyage 17, two tanks from voyage 27, and one tank each from voyages 16, 20, and 26. In (D), the upper left cluster includes four tanks from voyage 6, two from voyage 24, one from voyage 26, and four from voyage 27; the lower left cluster includes four tanks from voyage 6, two from voyage 11, four from voyage 17, and one each from voyages 18, 19, and 20.

Figure 11 Principal component plots of biota abundance considering paired tanks with similar ballasting history (A, phytoplankton; C, bacteria) vs. “unpaired”’ tanks with different ballasting history (B, phytoplankton; D, bacteria). Paired tanks were from voyages 1, 2,3, 4, 7,9, 10, 12, 14, 15, 18, 22, and 28; voyage numbers are shown, with the second member of each pair in red. For (B and D), the following tanks were included: voyages 11, 16, 19, 20, 25, and 26, each with two tanks; voyage 24 with three tanks; voyages 17 and 27 each with 4 tanks; and voyage 6 with 8 tanks. In (B), the lower right cluster includes six tanks from voyage 6, four tanks from voyage 17, two tanks from voyage 27, and one tank each from voyages 16, 20, and 26. In (D), the upper left cluster includes four tanks from voyage 6, two from voyage 24, one from voyage 26, and four from voyage 27; the lower left cluster includes four tanks from voyage 6, two from voyage 11, four from voyage 17, and one each from voyages 18, 19, and 20.

Related Figures (27)

Fig. 1. Ballast tank sampling decision tree.
Fig. 1. Ballast tank sampling decision tree.
Ballast water management summary for the voyages and tanks examined
Ballast water management summary for the voyages and tanks examined
Real-time PCR assays used to screen for potentially harmful microalgae and cyanobacteria in ballast water, including target species, target loci, loci descriptions, PCR techniques used, and lower limits of detection * See Bergholtz et al. (2006—K. veneficum), Marshall et al. (2006—P. shumwayae).
Real-time PCR assays used to screen for potentially harmful microalgae and cyanobacteria in ballast water, including target species, target loci, loci descriptions, PCR techniques used, and lower limits of detection * See Bergholtz et al. (2006—K. veneficum), Marshall et al. (2006—P. shumwayae).
* Probe is specific for pathogenic Leptospira spp. > Probe detects Mycobacterium marinum, M. ulcerans, M. chesapeakii, M. pseudoshottsii, and M. liflandii. © Probe detects Vibrio spp. Probe detects toxic strains of Vibrio cholerae. Real-time PCR assays used to screen for potentially harmful eubacteria, including target species, target loci, loci descriptions, PCR techniques used, and lower limits of detection
* Probe is specific for pathogenic Leptospira spp. > Probe detects Mycobacterium marinum, M. ulcerans, M. chesapeakii, M. pseudoshottsii, and M. liflandii. © Probe detects Vibrio spp. Probe detects toxic strains of Vibrio cholerae. Real-time PCR assays used to screen for potentially harmful eubacteria, including target species, target loci, loci descriptions, PCR techniques used, and lower limits of detection
Fig. 2. Distribution of tanks based on ballast water management status. U = “unknown” tanks with insufficient records to determine whether exchange had occurred; N = no exchange was conducted; XU = ex- change or open-ocean dilution occurred, but to an undetermined extent of tank volume exchange; X1 = | tank-volume exchange; X2 = 2 tank- volume exchanges; X3 = 3 tank-volume exchanges. Appropriate determinative analyses were applied to further characterize relationships among the measured variables. These included correlation analyses and a standard set of general linear models (GLM) analyses (linear regressions, ANOVAs) to examine relationships among response variables (phytoplankton abundance, phytoplankton species richness, bacterial abundance) and explanatory variables (physical/chemical data, ballast management), including log and other transfor- mations. It should be noted that four tanks (Voyage 3, tank pair; Voyage 25, both tanks) were eliminated from statistical analyses involving coastal versus non-coastal source waters because there were no ballast records available. Various transformations and different data
Fig. 2. Distribution of tanks based on ballast water management status. U = “unknown” tanks with insufficient records to determine whether exchange had occurred; N = no exchange was conducted; XU = ex- change or open-ocean dilution occurred, but to an undetermined extent of tank volume exchange; X1 = | tank-volume exchange; X2 = 2 tank- volume exchanges; X3 = 3 tank-volume exchanges. Appropriate determinative analyses were applied to further characterize relationships among the measured variables. These included correlation analyses and a standard set of general linear models (GLM) analyses (linear regressions, ANOVAs) to examine relationships among response variables (phytoplankton abundance, phytoplankton species richness, bacterial abundance) and explanatory variables (physical/chemical data, ballast management), including log and other transfor- mations. It should be noted that four tanks (Voyage 3, tank pair; Voyage 25, both tanks) were eliminated from statistical analyses involving coastal versus non-coastal source waters because there were no ballast records available. Various transformations and different data
* Note that of the 62 tanks assessed, 4 (from voyages 3 and 25) could not be classified as coastal or open-ocean because there were no ballast records. Age distribution of water (days) in ballast tanks that contained at least a portion of coastal source(s) water*
* Note that of the 62 tanks assessed, 4 (from voyages 3 and 25) could not be classified as coastal or open-ocean because there were no ballast records. Age distribution of water (days) in ballast tanks that contained at least a portion of coastal source(s) water*
Fig. 3. (A) Temperature and salinity and (B) pH and dissolved oxygen in each ballast water tank, considered by major source region (note that U = unidentified major source region). Within each major source region, tanks are arranged by age in ascending order from left to right. Data are given as means + | standard deviation (S.D.) for each tank.
Fig. 3. (A) Temperature and salinity and (B) pH and dissolved oxygen in each ballast water tank, considered by major source region (note that U = unidentified major source region). Within each major source region, tanks are arranged by age in ascending order from left to right. Data are given as means + | standard deviation (S.D.) for each tank.
Salinity, temperature, pH, and dissolved oxygen (DO) in the ballast water of sampled tanks Each parameter was averaged over all depths for each sampling event, and then an overall mean was calculated for the ballast water from each ocean. Data are given as means + 1S.D. (n = number of replicates); coefficients of variation (CVs) are also indicated. Table 5
Salinity, temperature, pH, and dissolved oxygen (DO) in the ballast water of sampled tanks Each parameter was averaged over all depths for each sampling event, and then an overall mean was calculated for the ballast water from each ocean. Data are given as means + 1S.D. (n = number of replicates); coefficients of variation (CVs) are also indicated. Table 5
* TKN values estimated (below detection limit). Nutrient concentrations and TN:TP ratios (molar basis) in the ballast water of sampled tanks (means + 1S.E., with S.D.s given in parentheses; data available for 56 of 62 tanks)*
* TKN values estimated (below detection limit). Nutrient concentrations and TN:TP ratios (molar basis) in the ballast water of sampled tanks (means + 1S.E., with S.D.s given in parentheses; data available for 56 of 62 tanks)*
Phytoplankton taxa found in ballast water samples, also indicating potential harmful traits
Phytoplankton taxa found in ballast water samples, also indicating potential harmful traits
Taxa are marine/estuarine unless otherwise indicated (note: temp—temperate, trop—tropical, subtrop—subtropical, frw—freshwater, ‘’—distribu- tion uncertain). References consulted for harmful traits include Burkholder (1998), Rhodes et al. (1998), Rengefors and Legrand (2001), Landsberg (2002), Hallegraeff et al. (2003) and Salas et al. (2005). Potentially harmful taxa are indicated in bold. Harmful taxa were defined as photosynthetic and heterotrophic algae capable of directly or indirectly causing disease or death of humans or beneficial aquatic life.
Taxa are marine/estuarine unless otherwise indicated (note: temp—temperate, trop—tropical, subtrop—subtropical, frw—freshwater, ‘’—distribu- tion uncertain). References consulted for harmful traits include Burkholder (1998), Rhodes et al. (1998), Rengefors and Legrand (2001), Landsberg (2002), Hallegraeff et al. (2003) and Salas et al. (2005). Potentially harmful taxa are indicated in bold. Harmful taxa were defined as photosynthetic and heterotrophic algae capable of directly or indirectly causing disease or death of humans or beneficial aquatic life.
Fig. 4. Examples of phytoplankton taxa in the ballast tanks, as scanning electron micrographs (SEMs) or light micrographs (LMs), and indicating whether the species was successfully cultured from ballast tank waters: (A and E) cultured Heterocapsa rotundata—SEM and LM, respectively; (B and F) cultured harmful species Karlodinium veneficum—SEM and LM, respectively; (C) cultured harmful species Karlodinium australe—SEM; (D) cultured Lingulodinium polyedrum—SEM; (G) Dinophysis caudata—LM of two flagellate cells; (H) Dictyocha speculum—LM; (I) cultured Heterosigma akashiwo—SEM;; (J) Ceratium tripos—LM; (K) Ceratium furca—LM,; (L) Dinophysis acuminata—LM; Chaetoceros didymus—LM; and (N, O) Pseudo-nitzschia fraudulenta—LM (two cells in center) and SEM, respectively.
Fig. 4. Examples of phytoplankton taxa in the ballast tanks, as scanning electron micrographs (SEMs) or light micrographs (LMs), and indicating whether the species was successfully cultured from ballast tank waters: (A and E) cultured Heterocapsa rotundata—SEM and LM, respectively; (B and F) cultured harmful species Karlodinium veneficum—SEM and LM, respectively; (C) cultured harmful species Karlodinium australe—SEM; (D) cultured Lingulodinium polyedrum—SEM; (G) Dinophysis caudata—LM of two flagellate cells; (H) Dictyocha speculum—LM; (I) cultured Heterosigma akashiwo—SEM;; (J) Ceratium tripos—LM; (K) Ceratium furca—LM,; (L) Dinophysis acuminata—LM; Chaetoceros didymus—LM; and (N, O) Pseudo-nitzschia fraudulenta—LM (two cells in center) and SEM, respectively.
Fig. 5. Neighbor-joining analysis using the Jukes-Cantor evolutionary model. Analysis is based on a 149-bp region of the 3’-end of the 18S rRNA dinoflagellate locus. Sequences available in GenBank with >90% similarity to ballast clones were included. In addition, 29 representative sequences from other dinoflagellate genera were included: Alexandrium, Amyloodinium, Amphidinium, Dinophysis, Gonyaulax, Gymnodinium/Gyrodinium, Heterocapsa, Oxyrrhis, Oxytricha, Peridinium, Prorocentrum, Scrippsiella, Strombiodinium, Takayama and Noctiluca (the latter, as the outgroup).
Fig. 5. Neighbor-joining analysis using the Jukes-Cantor evolutionary model. Analysis is based on a 149-bp region of the 3’-end of the 18S rRNA dinoflagellate locus. Sequences available in GenBank with >90% similarity to ballast clones were included. In addition, 29 representative sequences from other dinoflagellate genera were included: Alexandrium, Amyloodinium, Amphidinium, Dinophysis, Gonyaulax, Gymnodinium/Gyrodinium, Heterocapsa, Oxyrrhis, Oxytricha, Peridinium, Prorocentrum, Scrippsiella, Strombiodinium, Takayama and Noctiluca (the latter, as the outgroup).
Fig. 6. Relative abundance of viable phytoplankton as major groups, for: (A) All ballast water tanks collectively, all tanks with negligible (<5%) coastal water sources, and tanks with coastal sources (dino- flag., dinoflagellates; cyano., cyanobacteria); and (B) ballast tanks that had 1-3 complete exchanges (X1, X2, X3) vs. tanks that received no exchange (N), tanks that were exchanged but to an unknown extent (XU = 22 of 62 tanks), and “‘unknown”’ tanks with insufficient records to determine whether exchange had occurred (U). Data are given as means + 1S.E.
Fig. 6. Relative abundance of viable phytoplankton as major groups, for: (A) All ballast water tanks collectively, all tanks with negligible (<5%) coastal water sources, and tanks with coastal sources (dino- flag., dinoflagellates; cyano., cyanobacteria); and (B) ballast tanks that had 1-3 complete exchanges (X1, X2, X3) vs. tanks that received no exchange (N), tanks that were exchanged but to an unknown extent (XU = 22 of 62 tanks), and “‘unknown”’ tanks with insufficient records to determine whether exchange had occurred (U). Data are given as means + 1S.E.
Fig. 7. Total and viable (W) phytoplankton abundance for each ballast tank, quantitatively collected from plankton net tows (20-~m mesh) (means + 1S.D., n = 2). The data are grouped by major source region (U = unidentified major source) and arranged by ascending age proceeding from left to right within each major ocean source. The age (in days) of the ballast water (time held in tanks), based upon the most recent documented ballasting event prior to sampling, is shown over each bar.
Fig. 7. Total and viable (W) phytoplankton abundance for each ballast tank, quantitatively collected from plankton net tows (20-~m mesh) (means + 1S.D., n = 2). The data are grouped by major source region (U = unidentified major source) and arranged by ascending age proceeding from left to right within each major ocean source. The age (in days) of the ballast water (time held in tanks), based upon the most recent documented ballasting event prior to sampling, is shown over each bar.
tank and unpaired tank datasets, respectively. For the paired tanks dataset, the dominant variables for PC1 were salinity, DO, and TOC, and the dominant variables for PC2 were ballast water age and temperature. For the tightly clustered in comparison to the unpaired tank ordinations (Fig. 11). The percentage of the variance in the dataset that was explained by PC1 and PC2 was moderate, with means of 58% and 62% for the paired Fig. 8. Total and viable phytoplankton abundances by size class (maximum cell dimension), arranged in order of water age from youngest to oldest within each ocean source, in comparison to proposed regulatory discharge limits (lines) as (A) IMO (2004), 10 viable cells m~*; (B) U.S. Senate (2005), 0.1 viable cell m~*; (C) IMO (2004), 10’ viable cell m~?; (D) U.S. Senate (2005), 10° viable cells m~°.
tank and unpaired tank datasets, respectively. For the paired tanks dataset, the dominant variables for PC1 were salinity, DO, and TOC, and the dominant variables for PC2 were ballast water age and temperature. For the tightly clustered in comparison to the unpaired tank ordinations (Fig. 11). The percentage of the variance in the dataset that was explained by PC1 and PC2 was moderate, with means of 58% and 62% for the paired Fig. 8. Total and viable phytoplankton abundances by size class (maximum cell dimension), arranged in order of water age from youngest to oldest within each ocean source, in comparison to proposed regulatory discharge limits (lines) as (A) IMO (2004), 10 viable cells m~*; (B) U.S. Senate (2005), 0.1 viable cell m~*; (C) IMO (2004), 10’ viable cell m~?; (D) U.S. Senate (2005), 10° viable cells m~°.
Fig. 9. Relationship between phytoplankton abundance (log-transformed data) and ballast water age, showing total cells and viable cells (upper and lower panels, respectively).
Fig. 9. Relationship between phytoplankton abundance (log-transformed data) and ballast water age, showing total cells and viable cells (upper and lower panels, respectively).
Fig. 10. Phytoplankton species richness and harmful algal species richness in each ballast tank (upper and lower panels, respectively), considering total and viable cells from quantitative 20-j.zm-mesh net tows. The data are presented as in Fig. 7.
Fig. 10. Phytoplankton species richness and harmful algal species richness in each ballast tank (upper and lower panels, respectively), considering total and viable cells from quantitative 20-j.zm-mesh net tows. The data are presented as in Fig. 7.
Data summary from molecular screening for pathogenic eubacteria detected in ballast tanks* * All blanks were negative for these taxa; data not available for water from ballast tanks 12-Pla,b and 26-1. Note that Pseudomonas putida was also detected from tank 10-P1b using other techniques (see Table 9).
Data summary from molecular screening for pathogenic eubacteria detected in ballast tanks* * All blanks were negative for these taxa; data not available for water from ballast tanks 12-Pla,b and 26-1. Note that Pseudomonas putida was also detected from tank 10-P1b using other techniques (see Table 9).
Fig. 12. Bacterial abundance, with data given as in Fig. 7. (A) Total bacterial abundance by major source region, age, and status (coastal, open- ocean, uncertain). The presence of copious flocculent materials prevented analysis of the water from tanks 27-P1a,1b. (B) Abundance of Vibrio spp. in each ballast tank by source region, age, and status (coastal, open-ocean, uncertain). The data were obtained using TCBS plating, which is selective for Vibrio but does not permit speciation. Numbers indicate the age of the water based upon the most recent ballasting activity; the asterisk (*) indicates a sample for which Vibrio spp. were too numerous to count. The line (at 1 CFU 100 ml7') indicates the regulatory treatment standards threshold of the IMO Convention (2004) and the U.S. Ballast Water Management Act (2005; see United States Senate 2005) for Vibrio cholerae.
Fig. 12. Bacterial abundance, with data given as in Fig. 7. (A) Total bacterial abundance by major source region, age, and status (coastal, open- ocean, uncertain). The presence of copious flocculent materials prevented analysis of the water from tanks 27-P1a,1b. (B) Abundance of Vibrio spp. in each ballast tank by source region, age, and status (coastal, open-ocean, uncertain). The data were obtained using TCBS plating, which is selective for Vibrio but does not permit speciation. Numbers indicate the age of the water based upon the most recent ballasting activity; the asterisk (*) indicates a sample for which Vibrio spp. were too numerous to count. The line (at 1 CFU 100 ml7') indicates the regulatory treatment standards threshold of the IMO Convention (2004) and the U.S. Ballast Water Management Act (2005; see United States Senate 2005) for Vibrio cholerae.
Prokaryote clones isolated from the ballast water tanks, indicating the taxon, GenBank accession number, vessel and tank number, and the sequence similarity score with a sequence in the RD database
Prokaryote clones isolated from the ballast water tanks, indicating the taxon, GenBank accession number, vessel and tank number, and the sequence similarity score with a sequence in the RD database
Restricted maximum likelihood variance components for phytoplankton and bacterial abundance, expressed as proportions of the overall variance within and across ships, for all tanks (complete data set) and for tank subsets 1-3 lote that the error term for each of the four data set analyses was a very small proportion of the overall variance (<0.01). * Paired subset—subset of ships for which only paired tanks (same ballasting history) were sampled (voyages 1, 2, 3, 4, 10, 12, 14, 15, 18, 22, 28) > Unpaired subset 1—subset of ships for which more than a single set of paired tanks was sampled (voyages 6, 11, 17, 24, 27). © Unpaired subset 2—subset of ships for which multiple pairs or a single set of unpaired tanks was sampled (voyages 6, 11, 16, 17, 20, 24, 26, 27)
Restricted maximum likelihood variance components for phytoplankton and bacterial abundance, expressed as proportions of the overall variance within and across ships, for all tanks (complete data set) and for tank subsets 1-3 lote that the error term for each of the four data set analyses was a very small proportion of the overall variance (<0.01). * Paired subset—subset of ships for which only paired tanks (same ballasting history) were sampled (voyages 1, 2, 3, 4, 10, 12, 14, 15, 18, 22, 28) > Unpaired subset 1—subset of ships for which more than a single set of paired tanks was sampled (voyages 6, 11, 17, 24, 27). © Unpaired subset 2—subset of ships for which multiple pairs or a single set of unpaired tanks was sampled (voyages 6, 11, 16, 17, 20, 24, 26, 27)
Proposed regulatory discharge limits for viable phytoplankton, in comparison to the data from this study Data are given as number (percentage) of ballast tanks.
Proposed regulatory discharge limits for viable phytoplankton, in comparison to the data from this study Data are given as number (percentage) of ballast tanks.
Related topics:
Harmful AlgaeInternational Maritime OrganizationBallast water exchange
580 California St., Suite 400
San Francisco, CA, 94104
© 2025 Academia. All rights reserved