Ecotoxicology

, Volume 24, Issue 5, pp 1131–1140 | Cite as

Copper tolerance and distribution of epibiotic bacteria associated with giant kelp Macrocystis pyrifera in southern California

  • Julia Busch
  • Juliana Ribeiro Nascimento
  • Ana Carolina Rubem Magalhães
  • Bas E. Dutilh
  • Elizabeth Dinsdale
Open Access
Article

Abstract

Kelp forests in southern California are important ecosystems that provide habitat and nutrition to a multitude of species. Macrocystis pyrifera and other brown algae that dominate kelp forests, produce negatively charged polysaccharides on the cell surface, which have the ability to accumulate transition metals such as copper. Kelp forests near areas with high levels of boating and other industrial activities are exposed to increased amounts of these metals, leading to increased concentrations on the algal surface. The increased concentration of transition metals creates a harsh environment for colonizing microbes altering community structure. The impact of altered bacterial populations in the kelp forest have unknown consequences that could be harmful to the health of the ecosystem. In this study we describe the community of microorganisms associated with M. pyrifera, using a culture based approach, and their increasing tolerance to the transition metal, copper, across a gradient of human activity in southern California. The results support the hypothesis that M. pyrifera forms a distinct marine microhabitat and selects for species of bacteria that are rarer in the water column, and that copper-resistant isolates are selected for in locations with elevated exposure to transition metals associated with human activity.

Keywords

Copper tolerance Bacterial isolate Kelp forest 

Introduction

Kelp forests are one of the most widespread and productive ecosystems. Found in shallow temperate coastal waters as well as deeper tropical waters, kelp forests are dominated by brown algae from the genus Laminaria (Santelices 2007), such as Macrocystis pyrifera (giant kelp) and their high productivity provide habitats and nutrients to an entire ecosystem (Egan et al. 2008; Goecke et al. 2010). The nutrient-rich surface of the macroalgae provides a desirable location for epibiotic colonization resulting in higher abundance and diversity of microbial populations (Egan et al. 2008; Blight and Thompson 2008; Armstrong et al. 2001; Lachnit et al. 2011). Kelp produces a matrix of acidic polysaccharides that have a high binding affinity for transition metals such as copper (Davis et al. 2003; Andrade et al. 2004, 2010). While the algal surface is desirable for colonization, in locations where the amount of transition metals is increased by human activity, the bioaccumulation of metals by the kelp may create a stressful environment for bacteria.

Urbanization, industrialization and boating activity can increase the levels of transition metals entering the ocean. In particular, copper-based paint used to prevent marine fouling on boat hulls contributes to elevated levels of transition metals in marinas and bays around the world (Dobalian and Arias 2005; Biggs and D’Anna 2012; Schiff et al. 2006). In 2004 the average copper level in San Diego Bay was 8 ppb (Johnson and Gonzalez 2005), a 36-fold increase of the natural level of 0.225 ppb in ocean water (Blossom 2007). While copper is an essential micronutrient, in high levels it is lethal. The distribution of marine cyanobacteria cells is adversely affected by copper exposure and strains from populated coastal areas are more tolerant to elevated copper levels than those from the open ocean (Stuart et al. 2009; Moffett et al. 2012). However, little research has been conducted on the microbes associated with the kelp forest and the effects of transition metal accumulation. In southern California, the combination of high boating activity and the close proximity of kelp forests to marinas may increase the bioaccumulation of transition metals. To examine whether there was an effect of bioaccumulation of transition metals on microbial populations, we isolated bacteria from the kelp surface and surrounding water at three locations in southern California and described their taxonomic distribution and effect of copper on their growth.

Materials and methods

Bacteria were cultured from blades of M. pyrifera and surface seawater from three different kelp forest locations in southern California: The Point Loma kelp forest (32°39′59.56″N, 117°14′50.62″W) close to the port of San Diego, La Jolla kelp forest (32°50′3.71″N, 117°16′54.57″W) located adjacent to an urban area but with minimal boating activity, and Santa Catalina Island (33°27′1.89″N, 118°29′12.32″W) located adjacent a marine protected area approximately 70 km from the mainland (Fig. 1). Bacteria associated with the surface of the algae were obtained by washing the blade in 0.02 μm-filtered seawater to remove loosely associated bacteria. After washing, the kelp was wiped across the media plates transferring the viscous surface mucus and associated microbes onto the agar. Bacteria closely associated with the kelp blade tissue were isolated using a technique adapted from Burke et al. (2009) in which kelp surface-associated bacteria were eliminated before plating. The kelp blade pieces were washed in a 1 % bleach solution with gentle agitation for 30 s followed by two washes in 0.02 μm filtered seawater and the blade pieces were applied directly onto the agar. Additional bacterial isolates were obtained by crushing the kelp blade using a sterile mortar and pestle and a 100 μl sample was pipetted onto the media. Bacteria from the water column were obtained from around the kelp blades and 100 μl plated. The media used included TCBS (Neogen Cat. No. 7210A) and Marine Broth (Difco Laboratories Cat. No. 2013-10-31) with 15 g/l agar added. Bacteria were cultured at room temperature and streaked until purification. A total of 600 bacterial isolates were grown and a random set (98) were selected for RFLP analysis and copper tolerance assays.
Fig. 1

a A map of the San Diego Bay labeled with the areas with high boating activity that contribute to the copper load in the water. a Navy Submarine Base; b Shelter Island; c Fisherman’s Landing; d Harbor Island; e US Coast Guard; f B St. Cruise Ship Terminal, Broadway Pier, USS Midway; g Tuna Basin; h Embarcadero Marina; i US Naval Station; j Chula Vista Yacht Club; k South Bay Power Plant; l Coronado Cays; m Fiddler’s Cove; n Naval Amphibious Base Coronado; o Glorietta Bay; p Naval Air Station North Island. b Sampling locations off the coast of southern California. La Jolla (32°50′3.71″N, 117°16′54.57″W); Pt. Loma (32°39′59.56″N, 117°14′50.62″W); Catalina (33°27′1.89″N, 118°29′12.32″W)

Bacterial DNA was extracted using nuclei lysis buffer and protein precipitate solution. 16S PCR amplification was performed in which the DNA was denatured for 5 min at 94 °C and amplified through 30 cycles of 94 °C for 1 min, 65 °C for 1 min with 0.5 °C touchdown, 72 °C for 3 min, with a final extension of 72 °C for 10 min. The amplified 16S rDNA was detected by electrophoresis on a 1 % agarose gel with 0.5 μg/ml ethidium bromide in 1× TBE buffer.

To investigate the types of bacteria that were associated with either kelp or the water column at the three locations we conducted a restriction fragment length polymorphism (RFLP) analysis. The 16S sequences of 98 randomly selected bacteria were purified and digested using three restriction endonucleases: ScaI, HaeIII, and MspI. The digested 16S rDNA was run on a 2 % agarose gel followed by staining in ethidium bromide with a final concentration of 0.5 μg/ml. The 16S rDNA PCR product of one isolate from each RFLP group was sequenced using Sanger sequencing. Two sequences were sequenced from Pattern A to investigate whether the restriction pattern were unique to a species. The sequences were aligned with MUSCLE (Edgar 2004) and a maximum likelihood phylogenetic tree was created using RAxML (Stamatakis 2014) with standard settings and 100 bootstrap repetitions. Reference sequences for the phylogenetic analysis where chosen from NCBI after the alignment to identify the type strain of the closes matches to our 16S sequences. The type strains used were of Alteromonas addita, Photobacterium phosphoreum, Pseudoalteromonas arctica, Pseudoalteromonas citrea, Pseudoalteromonas espejiana, Psychroserpens burtonensis, Ruegeria mobilis, Shewanella fidelia, Vibrio cyclitrophicus, Vibrio mediterranei, Vibrio parahaemolyticus, Vibrio splendidus, and Vibrio wodanis (accession numbers: AY682202, PHR16SRD4, DQ787199, X82137, X82143, PBU62913, AB255401, AF420312, AM162656, X74710, AF388386, AJ515230, AJ132227 respectively).

The accumulation of copper in kelp was compared for Pt. Loma, near the highest human occupation and Catalina near the lowest level of human occupation in this study. The amount of copper in M. pyrifera blades were quantified using an acid digestion method modified from Warnau et al. (1995). The digested samples were analyzed on a Perkin–Elmer DV 4300 inductively coupled plasmid atomic emissions spectrometer (ICP-AES) and conversion calculations were performed to determine the for Cu concentrations. Differences between the copper concentrations at Catalina and Pt. Loma were compared using a student t test. Water quality measurements, including temperature, salinity, pH, dissolved oxygen, and chlorophyll-a were taken at surface (~0.5 m) and benthos (~12 m) for both the Pt. Loma and Catalina Island kelp forest collection sites with a MANTA-2 multiprobe (Measurement Specialties, VA, USA). A two sample two sided t test for each of the water quality measurements was use to compare the mean from three replicate measurements.

Growth inhibition assays were conducted in order to identify whether the bacteria isolated from different kelp forests had varying tolerance to copper. The median inhibitory concentration (IC50) of a bacterial isolate was defined as the lowest concentration of Cu at which growth was half the control value. IC50 assays were performed on ten bacterial isolates from Pt. Loma, ten bacterial isolates from La Jolla, and eight bacterial isolates from Catalina. A stock copper solution of CuSO4·5H2O in deionized water was prepared and sterilized. Bacterial isolates were grown from 1:1000 inoculations of overnight cultures in sterile Zobell Marine Broth 2216 (Difco™) with ranging concentrations of copper. Five replicates of each copper concentration were incubated for 18–20 h at room temperature. Optical density was measured at 550 nm to compare bacterial growth inhibition. To explore the relationship of kelp-associated and seawater isolates and growth inhibition by copper, a one-way ANOVA test was conducted, with the effect of geographic location and sample type (i.e., kelp or seawater) being tested.

Results

To investigate whether sample type (i.e., kelp tissue or water) or geographic location affected bacterial distribution, a random selection of 98 isolates was classified and 18 different RFLP patterns (labeled A through R) were identified (Table 1). Of the 98 isolates, 57 were Vibrio species and were found in both kelp and water from all three locations (Table 1, 2). Bacteria with RFLP Pattern A—most similar to V. splendidus were the most abundant and widespread, while Pattern B—most similar to V. cyclitrophicus was found across all three locations but not all sample types (Table 1). In addition to geographical differences, eight bacterial species most similar to Ruegeria sp., three Pseudoalteromonas spp., Alteromonas sp., two additional Vibrio spp., and Psychroserpens sp. were specific to kelp (n = 18 isolates), and two groups of bacteria were found exclusively in water (n = 7 isolates) including species similar to V. wodanis and P. phosphoreum.
Table 1

Sizes of RFLP fragments obtained from PCR amplified 16S rDNA of marine bacterial isolates

Stock number

Location

ScaI

HaeIII

MspI

Restriction pattern

156

Catalina kelp

1100, 475

310, 290, 210, 190, 150

550, 520, 125

A

157

Catalina kelp

1100, 475

310, 290, 210, 190, 150

550, 520, 125

A

260

Catalina kelp

1100, 475

310, 290, 210, 190, 150

550, 520, 125

A

330

Catalina kelp

1100, 475

310, 290, 210, 190, 150

550, 520, 125

A

131

Catalina water

1100, 475

310, 290, 210, 190, 150

550, 520, 125

A

147

Catalina water

1100, 475

310, 290, 210, 190, 150

550, 520, 125

A

149

Catalina water

1100, 475

310, 290, 210, 190, 150

550, 520, 125

A

189

Catalina water

1100, 475

310, 290, 210, 190, 150

550, 520, 125

A

3

La Jolla kelp

1100, 475

310, 290, 210, 190, 150

550, 520, 125

A

4

La Jolla kelp

1100, 475

310, 290, 210, 190, 150

550, 520, 125

A

5

La Jolla kelp

1100, 475

310, 290, 210, 190, 150

550, 520, 125

A

592

La Jolla kelp

1100, 475

310, 290, 210, 190, 150

550, 520, 125

A

593

La Jolla kelp

1100, 475

310, 290, 210, 190, 150

550, 520, 125

A

594

La Jolla kelp

1100, 475

310, 290, 210, 190, 150

550, 520, 125

A

595

La Jolla kelp

1100, 475

310, 290, 210, 190, 150

550, 520, 125

A

596

La Jolla kelp

1100, 475

310, 290, 210, 190, 150

550, 520, 125

A

598

La Jolla kelp

1100, 475

310, 290, 210, 190, 150

550, 520, 125

A

599

La Jolla kelp

1100, 475

310, 290, 210, 190, 150

550, 520, 125

A

582

La Jolla water

1100, 475

310, 290, 210, 190, 150

550, 520, 125

A

584

La Jolla water

1100, 475

310, 290, 210, 190, 150

550, 520, 125

A

585

La Jolla water

1100, 475

310, 290, 210, 190, 150

550, 520, 125

A

586

La Jolla water

1100, 475

310, 290, 210, 190, 150

550, 520, 125

A

590

La Jolla water

1100, 475

310, 290, 210, 190, 150

550, 520, 125

A

591

La Jolla water

1100, 475

310, 290, 210, 190, 150

550, 520, 125

A

642

La Jolla water

1100, 475

310, 290, 210, 190, 150

550, 520, 125

A

400

Pt. Loma water

1100, 475

310, 290, 210, 190, 150

550, 520, 125

A

401

Pt. Loma water

1100, 475

310, 290, 210, 190, 150

550, 520, 125

A

403

Pt. Loma water

1100, 475

310, 290, 210, 190, 150

550, 520, 125

A

404

Pt. Loma water

1100, 475

310, 290, 210, 190, 150

550, 520, 125

A

406

Pt. Loma water

1100, 475

310, 290, 210, 190, 150

550, 520, 125

A

407

Pt. Loma water

1100, 475

310, 290, 210, 190, 150

550, 520, 125

A

252

Catalina water

1100, 475

310, 290, 210, 190, 150

550, 520

B

411

Catalina water

1100, 475

310, 290, 210, 190, 150

550, 520

B

653

La Jolla kelp

1100, 475

310, 290, 210, 190, 150

550, 520

B

643

La Jolla water

1100, 475

310, 290, 210, 190, 150

550, 520

B

478

Pt. Loma kelp

1100, 475

310, 290, 210, 190, 150

550, 520

B

479

Pt. Loma kelp

1100, 475

310, 290, 210, 190, 150

550, 520

B

480

Pt. Loma kelp

1100, 475

310, 290, 210, 190, 150

550, 520

B

482

Pt. Loma kelp

1100, 475

310, 290, 210, 190, 150

550, 520

B

501

Pt. Loma kelp

1100, 475

310, 290, 210, 190, 150

550, 520

B

129

Catalina kelp

Uncut

310, 290, 210, 190, 150

550, 520

C

136

Catalina kelp

Uncut

310, 290, 210, 190, 150

550, 520

C

517

Catalina kelp

Uncut

310, 290, 210, 190, 150

550, 520

C

120

Catalina water

Uncut

310, 290, 210, 190, 150

550, 520

C

121

Catalina water

Uncut

310, 290, 210, 190, 150

550, 520

C

122

Catalina water

Uncut

310, 290, 210, 190, 150

550, 520

C

137

Catalina water

Uncut

310, 290, 210, 190, 150

550, 520

C

515

Catalina water

Uncut

310, 290, 210, 190, 150

550, 520

C

635

Catalina water

Uncut

310, 290, 210, 190, 150

550, 520

C

522

Pt. Loma kelp

Uncut

500, 290, 190, 125

600, 475, 275, 150

D

523

Pt. Loma kelp

Uncut

500, 290, 190, 125

600, 475, 275, 150

D

524

Pt. Loma kelp

Uncut

500, 290, 190, 125

600, 475, 275, 150

D

525

Pt. Loma kelp

Uncut

500, 290, 190, 125

600, 475, 275, 150

D

526

Pt. Loma kelp

Uncut

500, 290, 190, 125

600, 475, 275, 150

D

527

Pt. Loma kelp

Uncut

500, 290, 190, 125

600, 475, 275, 150

D

575

Pt. Loma kelp

Uncut

500, 290, 190, 125

600, 475, 275, 150

D

576

Pt. Loma kelp

Uncut

500, 290, 190, 125

600, 475, 275, 150

D

583

La Jolla water

Uncut

500, 290, 190, 125, 100

500, 375, 190, 125

E

644

La Jolla water

Uncut

500, 290, 190, 125, 100

500, 375, 190, 125

E

645

La Jolla water

Uncut

500, 290, 190, 125, 100

500, 375, 190, 125

E

646

La Jolla water

Uncut

500, 290, 190, 125, 100

500, 375, 190, 125

E

483

Pt. Loma kelp

Uncut

500, 290, 190, 125, 100

500, 375, 190, 125

E

484

Pt. Loma kelp

Uncut

500, 290, 190, 125, 100

500, 375, 190, 125

E

485

Pt. Loma kelp

Uncut

500, 290, 190, 125, 100

500, 375, 190, 125

E

521

Pt. Loma kelp

Uncut

500, 290, 190, 125, 100

500, 375, 190, 125

E

588

La Jolla kelp

1100, 475

500, 290, 190, 125

550, 520

F

654

La Jolla kelp

1100, 475

500, 290, 190, 125

550, 520

F

657

La Jolla kelp

1100, 475

500, 290, 190, 125

550, 520

F

658

La Jolla kelp

1100, 475

500, 290, 190, 125

550, 520

F

659

La Jolla kelp

1100, 475

500, 290, 190, 125

550, 520

F

581

La Jolla water

1100, 475

500, 290, 190, 125

550, 520

F

132

Catalina water

1100, 475

310, 290, 210, 190, 150

510, 480

G

187

Catalina water

1100, 475

310, 290, 210, 190, 150

510, 480

G

253

Catalina water

1100, 475

310, 290, 210, 190, 150

510, 480

G

412

Catalina water

1100, 475

310, 290, 210, 190, 150

510, 480

G

639

La Jolla water

1100, 475

310, 290, 210, 190, 150

510, 480

G

127

Catalina kelp

1100, 475

310, 290, 250, 210, 190, 150

550, 520

H

133

Catalina kelp

1100, 475

310, 290, 250, 210, 190, 150

550, 520

H

134

Catalina kelp

1100, 475

310, 290, 250, 210, 190, 150

550, 520

H

135

Catalina kelp

1100, 475

310, 290, 250, 210, 190, 150

550, 520

H

641

La Jolla water

1100, 475

310, 290, 250, 210, 190, 150

550, 520

H

329

Catalina kelp

Uncut

310, 290, 210, 190, 150

510, 480, 125

I

416

Catalina kelp

Uncut

310, 290, 210, 190, 150

510, 480, 125

I

139

Catalina water

Uncut

310, 290, 210, 190, 150

510, 480, 125

I

248

Catalina kelp

1100, 475

500, 290, 210, 190, 125

510, 375, 180

J

647

La Jolla water

1100, 475

500, 290, 210, 190, 125

510, 375, 180

J

587

La Jolla kelp

850, 475, 225

500, 290, 190, 125

550, 520

K

589

La Jolla kelp

850, 475, 225

500, 290, 190, 125

550, 520

K

656

La Jolla kelp

Uncut

310, 290, 190, 150

550, 520

L

486

Pt. Loma kelp

Uncut

310, 290, 190, 150

550, 520

L

142

Catalina kelp

Uncut

500, 290, 190, 125

510, 375, 180

M

143

Catalina kelp

Uncut

500, 290, 190, 125

510, 375, 180

M

651

La Jolla kelp

850, 650

310, 290, 190, 150

510, 480

N

413

Catalina kelp

Uncut

310, 290, 210, 175, 150

550, 520

O

124

Catalina water

Uncut

390, 310, 290, 190, 150

550, 520

P

637

Pt. Loma kelp

Uncut

500, 290, 190, 125, 100

550, 520, 125

Q

481

Pt. Loma kelp

Uncut

650, 290,250, 190

800, 450, 125

R

Bold numbers indicate sequenced Isolates from each group

Table 2

BLASTn results of 16S rDNA sequences from each RFLP pattern

Sequenced stock number

RFLP pattern

number of isolates

Percent found in kelp

Class

Family

Hypothesized classification

Accession number

ED401

A

31

45.16

Gamma proteobacteria

Vibrionaceae

Vibrio sp.

KF442438

ED478

B

9

66.6

Gamma proteobacteria

Vibrionaceae

Vibrio sp.

KF442439

ED121

C

9

33.3

Gamma proteobacteria

Vibrionaceae

Vibrio sp.

KF442440

ED575

D

8

100

Alpha proteobacteria

Rhodobacteraceae

Ruegeria sp.

KF442441

ED583

E

8

50

Gamma proteobacteria

Alteromonadaceae

Pseudoalteromonas sp.

KF442442

ED588

F

6

83.3

Gamma proteobacteria

Alteromonadaceae

Pseudoalteromonas sp.

KF442443

ED187

G

5

0

Gamma proteobacteria

Vibrionaceae

Photobacterium sp.

KF442444

ED641

H

5

80

Gamma proteobacteria

Vibrionaceae

Vibrio sp.

KF442445

ED329

I

3

66.6

Gamma proteobacteria

Shewanellaceae

Shewanella sp.

KF442446

ED248

J

2

50

Gamma proteobacteria

Alteromonadaceae

Pseudoalteromonas sp.

KF442447

ED589

K

2

100

Gamma proteobacteria

Alteromonadaceae

Pseudoalteromonas sp.

KF442448

ED486

L

2

100

Gamma proteobacteria

Alteromonadaceae

Alteromonas sp.

KF442449

ED142

M

2

100

Gamma proteobacteria

Alteromonadaceae

Pseudoalteromonas sp.

KF442450

ED651

N

1

100

Gamma proteobacteria

Vibrionaceae

Vibrio sp.

KF442451

ED413

O

1

100

Gamma proteobacteria

Vibrionaceae

Vibrio sp.

KF442452

ED124

P

1

0

Gamma proteobacteria

Vibrionaceae

Vibrio sp.

KF442453

ED637

Q

1

100

Gamma proteobacteria

Alteromonadaceae

Pseudoalteromonas sp.

KF442454

ED481

R

1

100

Flavobacteria

Flavobacteriaceae

Psychroserpens sp.

KF442455

Total

18

97

 

3

5

18

 
Isolates from geographic locations and sample types were distributed evenly throughout the phylogenetic tree (Fig. 2). The Vibrio and Pseudoalteromonas species clustered into two separated groups and bacteria isolates that had the RLFP Pattern D were not similar to anything in the database and appeared as an outlier in the phylogenetic analysis, suggesting it is a new species based on the 16S rDNA sequence (Fig. 2).
Fig. 2

Phylogenetic tree of 16S rDNA sequences created from a MUSCLE alignment using RAxML. Bootstrap values (100 iterations) are indicated at the nodes. The scale bar indicates 0.2 mutations per site

The concentration of copper found in the kelp from Pt. Loma was greater with a mean of 6.21 (± 0.19) ppm compared with the kelp from Catalina which had a mean of 4.33 (±0.08) ppm (Tdf=4 = 10.334, p < 0.001). No data was collected from La Jolla. Water quality measurements indicate that the Pt. Loma and Catalina Island kelp forests were similar at surface depths for dissolved oxygen levels (p = 0.345) and salinity (p = 0.107), and similar at the benthos for chlorophyll-a (p = 0.891). In contrast, temperature was higher at the surface in Pt. Loma compared to Catalina [x̄PtLoma = 16.807 (±0.039) °C, x̄Catalina = 16.593 (±0.009) °C] (p = 0.027), yet lower in the benthos of Pt. Loma compared to Catalina [x̄PtLoma = 12.210 (±0.015) °C, x̄Catalina = 15.167 (±0.003) °C] (p < 0.001). In addition, pH was higher in Catalina at surface [x̄PtLoma = 8.073 (±0.055), x̄Catalina = 8.847 (±0.009)] (p = 0.004) and benthos [x̄PtLoma = 8.397 (±0.003), x̄Catalina = 8.783 (±0.007)] (p < 0.001). Furthermore, chlorophyll-a was higher at the surface of Pt. Loma [x̄PtLoma = 1.347 (±0.015) mg/l, x̄Catalina = 1.253 (±0.018) mg/l] (p = 0.016), while dissolved oxygen was lower [x̄PtLoma = 6.123 (±0.023) mg/l, x̄Catalina = 8.880 (±0.010) mg/L] (p < 0.001) and salinity was higher [x̄PtLoma = 45.361 (±0.026) ppt, x̄Catalina = 41.770 (±0.027) ppt] (p < 0.001) at the benthos of Pt. Loma compared to Catalina (Table 3).
Table 3

Water quality measurements taken with a MANTA-2 multiprobe

 

Depth (m)

Temp (°C)

Salinity (ppt)

pH

Chlorophyll-a (mg/l)

Dissolved oxygen

(% sat)

Dissolved oxygen

(mg/l)

Pt. Loma

0.607 ± 0.047

16.807 ± 0.039

40.308 ± 0.039

8.073 ± 0.055

1.347 ± 0.015

118.067 ± 2.034

9.367 ± 0.153

13.950 ± 0.183

12.210 ± 0.015

45.361 ± 0.026

8.397 ± 0.003

3.853 ± 0.116

70.267 ± 0.260

6.123 ± 0.023

Catalina

0.230 ± 0.006

16.593 ± 0.009

40.205 ± 0.010

8.847 ± 0.009

1.253 ± 0.018

115.000 ± 0.808

9.177 ± 0.064

11.490 ± 0.074

15.167 ± 0.003

41.770 ± 0.027

8.783 ± 0.007

3.973 ± 0.766

108.233 ± 0.167

8.880 ± 0.010

Data was logged at surface and benthos depths for each water collection site; Pt. Loma and Catalina Island, California kelp forests. Values displayed are the mean for three replicate measurements, with standard error of the mean (SEM)

For all kelp forest bacterial isolates, the lowest concentrations of copper in the media resulted in a slight decrease in optical density. After the decrease, bacterial growth was maintained until a steep drop was observed in the graph of absorbance versus copper concentration suggesting the copper concentration had reached an inhibitory level (Fig. 3a–c). The IC50 data for Pt. Loma isolates (n = 10) showed higher copper tolerance with the majority of growth inhibition occurring between 200 and 250 ppm Cu (Fig. 3a). The bacterial isolates from La Jolla (n = 10) showed growth inhibition occurring between 150 and 275 ppm Cu (Fig. 1b). Catalina isolates (n = 9) showed growth inhibition occurring between 100 and 225 ppm of Cu (Fig. 3c).
Fig. 3

a The concentration of copper that inhibited bacterial growth in isolates from Pt. Loma was found to be above 200 ppm in all but one sample. b The concentration of copper that inhibited bacterial growth in isolates from La Jolla was found to range from 150 to 275 ppm. c The concentration of copper that inhibited bacterial growth in isolates from Catalina was found to range from 100 to 225 ppm. d Mean copper inhibitory growth concentrations of M. pyrifera (green) and seawater (blue) associated bacterial isolates across three geographic locations. e The copper MIC assay data for bacterial species classified as similar to V. cyclitrophicus by 16S rDNA isolated from Pt. Loma (red), La Jolla (blue), and Catalina (green). The isolates from Pt. Loma were the most resistant to copper while the one collected from Catalina was the least copper tolerant. Isolate numbers used in the MIC are from: Pattern A—ED400, ED3, ED4, ED5, ED582, ED147,ED149, Pattern B—D8, ED478, ED482, ED395, ED653, ED643, Pattern C—ED129, Pattern E—ED583, ED484, Pattern G—ED639, Pattern H—ED2, ED133, Pattern I—ED416, ED139, Pattern O—ED413 (Color figure online)

The minimal inhibitory concentration varied with location (Fdf=2 = 41.215 and p < 0.001), sample type (Fdf=1 = 4.018 and p = 0.047) and there was an interaction between location and sample type (Fdf=2 = 11.394 and p < 0.001) (Fig. 3d). The minimal inhibitory concentration was highest at Pt. Loma and the kelp associated bacteria had a higher IC50 than the water associated bacteria. The second highest IC50 values were found at La Jolla, but the water associated bacteria had higher IC50 than the kelp associated bacteria. Catalina showed the lowest IC50 and once again the kelp associated bacteria had higher IC50 than the water associated bacteria. Trends of the copper tolerance of the community of bacteria were observed in specific bacterial species where five isolates similar to V. cyclitrophicus from M. pyrifera and seawater showed different values for copper tolerance across the three geographic locations (Fig. 3e). The copper tolerance of isolates was negatively correlated with distance from human urbanization and marinas.

Discussion

High levels of urbanization, boating and the resultant copper introduced into the water column from antifouling paint may be capable of affecting the structure and function of bacterial populations in kelp forests. Bacterial isolates from Pt. Loma, the location with highest measured copper levels, were able to tolerate copper concentrations significantly higher than bacteria isolated from La Jolla and Catalina, locations with less human activity (Fig. 3d). We showed that the input of transient metals causes a physiological change (tolerance to copper) in the bacteria and the ramifications of these changes on the kelp health will be the goals of future research. Because a range of bacterial species were tested, we show that the development of copper tolerance was location rather than species specific, suggesting that bacteria that lived in a higher copper habitat develop the ability to withstand higher levels of copper.

In addition to the positive correlation between copper pollution and bacterial copper resistance across different species, we observed that strains of a single species, similar to V. cyclitrophicus, had higher copper tolerance from Catalina to La Jolla, and the greatest tolerance was found in isolates from Pt. Loma (Fig. 3e). The results of increased copper tolerance in bacteria from areas with higher levels of copper are consistent with findings that Enterobacteriaceae strains from marine sediments and seawater have copper resistance directly influenced by pollution (Altug and Balkis 2009). These results suggest that some marine bacteria are able to increase their tolerance to higher levels of copper when living in locations with elevated inputs of transient metals.

Copper in the oceanic water column is reported to be 0. 225 ppb (Blossom 2007) and levels in San Diego Bay was ~8 ppb (Johnson and Gonzalez 2005; Neira et al. 2011), and we found that the copper in the kelp at Pt. Loma and Catalina was high, a mean 6.21 and 4.5 ppm respectively, showing the kelp is concentrating the copper from the water column. The amount of copper in an organism depends on the diffusion across tissue boundaries and the amount to which it is retained in the organism (Neira et al. 2011). The switch in higher copper tolerance in the microbes isolated from the water column compared to the kelp at La Jolla, suggests that at this location the copper may be returning to solution at a faster rate than at Pt. Loma or Catalina. Although the concentrations of copper at which bacterial growth was inhibited for the majority of isolates were above the levels found in M. pyrifera from Pt. Loma and Catalina, our results suggests that the transient metals in the environment are correlated with a phenotypic change in the bacteria. In addition previous studies have shown that the brown macroalgae Ascophyllum sp. and Laminaria japonica can accumulate 80.06 and 100.03 ppt of Cu respectively (Mehta and Gaur 2005),which is 1000 fold greater than the concentrations shown to inhibit growth of the bacterial isolates (Fig. 3). If the antifoulant paint used on boats in the San Diego Bay and other marinas worldwide continue to contribute copper to the water, in addition to other anthropogenic inputs, the algae in adjacent kelp forests such as Pt. Loma are capable of accumulating lethal concentrations of copper for even the most resistant bacterial isolate and these changes in epibiosis may affect kelp health.

A distinct bacterial communities associated with giant kelp were identified. Of the 18 RFLP patterns observed, ten were found in a particular location and ten are specific to either kelp or seawater. These results support the hypothesis that kelp forms a distinct microhabitat that selects for certain species of bacteria that are rare in the surrounding water and we are currently investigating the genotypes of these species to identify why they a specific to either kelp or water column. Vibrio species were the dominant organism cultured from the kelp and surrounding water (Table 1). These gram negative, motile rod-shaped species are ubiquitous in marine environments and are found in especially high amounts on eukaryotic host organisms, such as coral and algae (Thompson et al. 2004). This is consistent with our findings as bacterial species from the Vibrio genus were the most widespread in terms of location as well as the most abundant. Pseudoalteromonas is another well-represented marine Gamma proteobacteria and representatives from this genus have been isolated from both brown and green algae as well as seawater (Egan et al. 2008; Sawabe et al. 1998; Staufenberger et al. 2008). Shewanella species have previously been isolated from both seawater and marine organisms and also have the ability to digest agar (Ivanova et al. 2001, 2004). While some species were isolated from both sample types across all sampling sites, many were exclusive to geographical location. Catalina and Pt. Loma, the sites with the least and greatest anthropogenic effects, respectively, have more species that appear to be specific to the locations than La Jolla. Our current research is investigating the non-cultured organisms associated with the kelp using next generation DNA sequencing of metagenomes, which will provide a more complete picture of the kelp microbiome. By culturing bacteria we are able to conduct phenotypic analysis, which are not possible with metagenomics. In addition, we are sequencing the genomes of the cultured bacteria to understand the underpinnings of the phenotypic differences in copper tolerance of the bacteria.

The results of this study suggest that marine microbes are subject to selection at the species level across different geographical locations with varying anthropogenic and industrial influences. Furthermore, kelp appears to select for certain species of bacteria, creating a diverse and distinct microbiome on the algal surface. Kelp forests near industrial areas appear to be decreasing in cover and to date, have failed to recover to pre-1950 levels (Coleman et al. 2008; Connell et al. 2008; Foster and Schiel 2010). The consequences of an altered microbial community could have dramatic effects on the health of the ecosystem and warrants further investigation.

Notes

Acknowledgments

Dinsdale is supported by NSF Division of Undergraduate Education #1323809 and Division of Microbial Biology #1330800, and FIPSE-CAPES USA-Brazil Higher Education Consortia Program. Nascimento and Magalhães are supported by CAPES-FIPSE. Dutilh is supported by the Dutch Science foundation (NWO) Veni Grant 016.111.075.

Conflict of interest

The authors declare that they have no conflict of interest.

References

  1. Altug G, Balkis N (2009) Levels of some toxic elements and frequency of bacterial heavy metal resistance in sediment and sea water. Environ Monit Assess 149:61–69. doi:10.1007/s10661-008-0183-z CrossRefGoogle Scholar
  2. Andrade LR, Salgado LT, Farina M et al (2004) Ultrastructure of acidic polysaccharides from the cell walls of brown algae. J Struct Biol 145:216–225. doi:10.1016/j.jsb.2003.11.011 CrossRefGoogle Scholar
  3. Andrade LR, Leal RN, Noseda M et al (2010) Brown algae overproduce cell wall polysaccharides as a protection mechanism against the heavy metal toxicity. Mar Pollut Bull 60:1482–1488. doi:10.1016/j.marpolbul.2010.05.004 CrossRefGoogle Scholar
  4. Armstrong E, Yan L, Boyd KG et al (2001) The symbiotic role of marine microbes on living surfaces. Water 461:37–40Google Scholar
  5. Biggs TW, D’Anna H (2012) Rapid increase in copper concentrations in a new marina, San Diego Bay. Mar Pollut Bull 64:627–635. doi:10.1016/j.marpolbul.2011.12.006 CrossRefGoogle Scholar
  6. Blight AJ, Thompson RC (2008) Epibiont species richness varies between holdfasts of a northern and a southerly distributed kelp species. J Mar Biol Assoc UK 88:469–475. doi:10.1017/S0025315408000994 CrossRefGoogle Scholar
  7. Blossom N (2007) Copper in the Ocean Environment. Am Chemet Corp 1–8Google Scholar
  8. Burke C, Kjelleberg S, Thomas T (2009) Selective extraction of bacterial DNA from the surfaces of macroalgae. Appl Environ Microbiol 75:252–256. doi:10.1128/AEM.01630-08 CrossRefGoogle Scholar
  9. Coleman MA, Kelaher BP, Steinberg PD, Millar AJK (2008) Absence of a large brown macroalgae on urbanized rocky reefs around Sydney, Australia, and evidence for historical decline 1. J Phycol 44:897–901. doi:10.1111/j.1529-8817.2008.00541.x CrossRefGoogle Scholar
  10. Connell S, Russell B, Turner D et al (2008) Recovering a lost baseline: missing kelp forests from a metropolitan coast. Mar Ecol Prog Ser 360:63–72. doi:10.3354/meps07526 CrossRefGoogle Scholar
  11. Davis TA, Volesky B, Mucci A (2003) A review of the biochemistry of heavy metal biosorption by brown algae. Water Res 37:4311–4330. doi:10.1016/S0043-1354(03)00293-8 CrossRefGoogle Scholar
  12. Dobalian L, Arias CM (2005) Total maximum daily load for dissolved copper in Shelter Island Yacht Basin. San Diego Regional Water Quality Board, Executive Officer's Report, 16 ppGoogle Scholar
  13. Edgar RC (2004) MUSCLE: multiple sequence alignment with high accuracy and high throughput. Nucleic Acids Res 32:1792–1797. doi:10.1093/nar/gkh340 CrossRefGoogle Scholar
  14. Egan S, Thomas T, Kjelleberg S (2008) Unlocking the diversity and biotechnological potential of marine surface associated microbial communities. Curr Opin Microbiol 11:219–225. doi:10.1016/j.mib.2008.04.001 CrossRefGoogle Scholar
  15. Foster MS, Schiel DR (2010) Loss of predators and the collapse of southern California kelp forests (?): alternatives, explanations and generalizations. J Exp Mar Bio Ecol 393:59–70. doi:10.1016/j.jembe.2010.07.002 CrossRefGoogle Scholar
  16. Goecke F, Labes A, Wiese J, Imhoff J (2010) Chemical interactions between marine macroalgae and bacteria. Mar Ecol Prog Ser 409:267–299. doi:10.3354/meps08607 CrossRefGoogle Scholar
  17. Ivanova EP, Sawabe T, Gorshkova NM et al (2001) Shewanella japonica sp. nov. Int J Syst Evol Microbiol 51:1027–1033CrossRefGoogle Scholar
  18. Ivanova EP, Gorshkova NM, Bowman JP et al (2004) Shewanella pacifica sp. nov., a polyunsaturated fatty acid-producing bacterium isolated from sea water. Int J Syst Evol Microbiol 54:1083–1087. doi:10.1099/ijs.0.02993-0 CrossRefGoogle Scholar
  19. Lachnit T, Meske D, Wahl M et al (2011) Epibacterial community patterns on marine macroalgae are host-specific but temporally variable. Environ Microbiol 13:655–665. doi:10.1111/j.1462-2920.2010.02371.x CrossRefGoogle Scholar
  20. Mehta SK, Gaur JP (2005) Use of algae for removing heavy metal ions from wastewater: progress and prospects. Crit Rev Biotechnol 25:113–152. doi:10.1080/07388550500248571 CrossRefGoogle Scholar
  21. Moffett JW, Brand LE, Croot PL et al (2012) Cu speciation and cyanobacterial distribution in harbors subject to anthropogenic Cu inputs. Limnol Oceanogr 42:789–799CrossRefGoogle Scholar
  22. Neira C, Mendoza G, Levin LA, Zirino A, Delgadillo-Hinojosa F, Porrachia M, Deheyn DD (2011) Macrobenthic community response to copper in Shelter Island Yacht Basin, San Diego Bay, California. Mar Poll Bull 62:701–717CrossRefGoogle Scholar
  23. Santelices B (2007) The discovery of kelp forests in deep-water habitats of tropical regions. Proc Natl Acad Sci USA 104:19163–19164. doi:10.1073/pnas.0708963104 CrossRefGoogle Scholar
  24. Sawabe T, Makino H, Tatsumi M et al (1998) Pseudoalteromonas bacteriolytica sp. nov., a marine bacterium that is the causative agent of red spot disease of Laminaria japonica. Int J Syst Bacteriol 48:769–774CrossRefGoogle Scholar
  25. Schiff K, Brown J, Diehl D (2006) Extent and magnitude of copper contamination in Marinas of the San Diego Region, California. Water Res 54:322–328Google Scholar
  26. Stamatakis A (2014) RAxML version 8: a tool for phylogenetic analysis and post-analysis of large phylogenies. Bioinformatics 30:1312–1313. doi:10.1093/bioinformatics/btu033 CrossRefGoogle Scholar
  27. Staufenberger T, Thiel V, Wiese J, Imhoff JF (2008) Phylogenetic analysis of bacteria associated with Laminaria saccharina. FEMS Microbiol Ecol 64:65–77. doi:10.1111/j.1574-6941.2008.00445.x CrossRefGoogle Scholar
  28. Stuart RK, Dupont CL, Johnson DA et al (2009) Coastal strains of marine Synechococcus species exhibit increased tolerance to copper shock and a distinctive transcriptional response relative to those of open-ocean strains. Appl Environ Microbiol 75:5047–5057. doi:10.1128/AEM.00271-09 CrossRefGoogle Scholar
  29. Thompson FL, Iida T, Swings J (2004) Biodiversity of Vibrios. Microbiol Mol Biol Rev 68:403–431. doi:10.1128/MMBR.68.3.403 CrossRefGoogle Scholar
  30. Warnau M, Ledent G, Temaral A, Jangouxr M (1995) Experimental cadmium contamination of the echinoid Paracentrotus lividus: influence of exposure mode and distribution of the metal in the organism. Mar Ecol Prog Ser 116:117–124CrossRefGoogle Scholar
  31. Johnson LT, Gonzalez JA (2005) Nontoxic antifouling? Demonstrating a solution to copper boat bottom paint pollution. In Proceedings of the 14th Biennial Coastal Zone Conference Google Scholar

Copyright information

© The Author(s) 2015

Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.

Authors and Affiliations

  • Julia Busch
    • 1
    • 2
  • Juliana Ribeiro Nascimento
    • 3
  • Ana Carolina Rubem Magalhães
    • 4
  • Bas E. Dutilh
    • 5
    • 6
  • Elizabeth Dinsdale
    • 1
  1. 1.Biology DepartmentSan Diego State UniversitySan DiegoUSA
  2. 2.Scripps Institution of Oceanography, University of CaliforniaSan DiegoUSA
  3. 3.Marine Biology Institute-CCS, Universidade Federal do Rio de JaneiroRio de JaneiroBrazil
  4. 4.Biology DepartmentFederal Fluminense UniversityRio de JaneiroBrazil
  5. 5.Computer Science DepartmentSan Diego State UniversityCaliforniaUSA
  6. 6.Centre for Molecular and Biomolecular Informatics, CMBI, Nijmegen Centre for Molecular Life SciencesRadboud University Nijmegen Medical CentreNijmegenThe Netherlands

Personalised recommendations