Microbial Ecology

, Volume 55, Issue 1, pp 81–93

Desmids and Biofilms of Freshwater Wetlands: Development and Microarchitecture

Authors

    • Department of BiologySkidmore College
  • Catherine Rogers Domozych
    • Department of BiologySkidmore College
Article

DOI: 10.1007/s00248-007-9253-y

Cite this article as:
Domozych, D.S. & Domozych, C.R. Microb Ecol (2008) 55: 81. doi:10.1007/s00248-007-9253-y

Abstract

Freshwater wetlands constitute important ecosystems, and their benthic, attached microbial communities, including biofilms, represent key habitats that contribute to primary productivity, nutrient cycling, and substrate stabilization. In many wetland biofilms, algae constitute significant parts of the microbial population, yet little is known about their activities in these communities. An analysis of wetland biofilms from the Adirondack region of New York (USA) was performed with special emphasis on desmids, a group of evolutionarily advanced green algae commonly found in these habitats. Desmids constituted as much as 23.7% of the total algal and cyanobacterial flora of the biofilms during the July and August study periods. These algae represented some of the first eukaryotes to colonize new substrates, and during July their numbers correlated with fluctuations in general biofilm parameters such as biofilm thickness and dry weight as well as total carbohydrate. Significant numbers of bacteria were associated with both the EPS sheaths and cell wall surfaces of the desmids. Colonization of new substrates and development of biofilms were rapid and were followed by various fluctuations in microbial community structure over the short- and long-term observations. In addition to desmids, diatoms, filamentous green algae and transient non-motile phases of flagellates represented the photosynthetic eukaryotes of these biofilms.

Introduction

Biofilms represent complex consortia of surface-dwelling microorganisms ensheathed in a highly hydrated matrix of exopolymers called the extracellular polymeric substance or EPS [4, 37, 42, 55, 66, 67, 71, 73]. In aquatic ecosystems, biofilms provide resident organisms with favorable conditions such as mechanical stability, frictional, hydraulic, and diffusion resistance and greater sorption capacity. The microarchitectural design of the heterogenous mosaic of microbes within the complex EPS matrix provides for an enhancement of metabolic activities and interspecies communication networks. The recent proliferation of biofilm research and the resultant biofilm paradigm have greatly enhanced our understanding of microbial ecology and thus significantly impacted applications in water treatment, medicine, industry, and environmental science [12, 13].

Research dealing with the microbial communities and biofilms of aquatic ecosystems has focused on two major habitat types, riparian systems, and large bodies of relatively deep water including lakes and ponds. It is surprising to note, however, that very little is known about the biofilm communities of freshwater wetlands [28, 3840]. Freshwater or inland wetlands constitute approximately 6% of global land surface, include a diverse array of bogs, fens, swamps and marshes, play vital roles in local and global hydrogeology, and are profoundly important for biological processes such as primary productivity, nutrient cycling, and biodiversity [1416, 29, 30, 49, 68]. In previous studies of wetland primary productivity, aquatic macrophytes have often been the key focal points. However, because of the large expanse of shallow, well-lit waters with ample substrate that typify wetlands, algae also contribute significantly to wetland productivity and biogeochemical processes therein. For example, algal primary productivity in wetlands can range up to 65% of the total (i.e., greater than aquatic macrophyte productivity [28]. Wetland algae are important in substrate stabilization, serving as critical carbon substrates for heterotrophs and creating key sinks for phosphorus and various metals. However, there is a noticeable lack of data concerning wetland algae, their specific niches and interactions with other biota. This represents a major void in our understanding of total wetland dynamics [28, 38].

In many freshwater wetlands, desmids (Conjugatophyceae, Chlorophyta) are common and sometimes dominant algal residents (e.g., oligotrophic and dystrophic waterways [26, 27]. These advanced green algae are found in both planktonic and benthic periphyton communities where they contribute to food chain dynamics [7, 10]. In periphyton habitats, desmids are most often associated with biofilms and possess features that are well-adapted to life in the biofilm. For example, recent research has highlighted the ability of desmids to produce prolific amounts of complex, polysaccharide-based EPS in both natural and laboratory biofilm conditions [1719, 43, 54, 56]. Desmid EPS has been shown to have an adhesive tensile strength similar to commercial adhesives [1], serves as a matrix for extracellular enzymes [11, 64, 65], possesses metal-binding characteristics [25, 44, 45], and serves as a major locus for specific associations with diverse groups of prokaryotes [2224]. Nevertheless, we know virtually nothing about the distribution of desmids in wetland biofilms, their roles in biofilm community structure and the role of desmid-derived EPS in food chains and interspecies networks with other biofilm microbes.

We are engaged in an ongoing study of wetland biofilms and, in particular, those where desmids are quantitatively important members in the biofilm community. Previous research [6], along with our initial sampling of Adirondack (NY, USA) wetlands during the summer months, indicated that desmids are key components of benthic attached communities of low conductance/low nutrient or “softwater” wetlands. Our goals in this study are: a) to assess the population of desmids in relation to other photosynthetic organisms in biofilms in both short-term and long-term studies, b) to provide a profile of water chemistry, nutritional values, and general biofilm structural details of whole biofilms containing desmids, c) to determine the efficacy of desmids as pioneer species upon new substrates, and d) to assess the quantitative distribution of other biofilm residents, especially bacteria, in physical relationships with desmids. This study represents an initial ecological survey of desmid-rich biofilms and is part of a larger effort to elucidate desmid biology as it relates to biofilms.

Methods

General

The study site selected for this project was the Thrailkill Pond complex, a series of eight beaver-engineered wetlands derived from a tributary of the Kayaderosseras Creek of Porter Corners, NY (USA). To evaluate artificial substrates for biofilm formation-efficacy, a preliminary survey of materials including glass, Plexiglas, and sandpaper was performed. No significant differences were noted in algal populations on the various substrates. The final substrate chosen for this study was a thin, lightweight, 11 × 11 cm Plexiglas sheet. All sheets were thoroughly washed in a dilute solution of Alconox detergent (Fisher Scientific, Pittsburg, PA, USA) and repeatedly rinsed in deionized water before emplacement in the wetland site. For quantitative dry weight and biochemical studies that required complete removal of biofilm from the substrate, empty sheets were used. For microscopy-based studies, sheets containing attached 2-cm2 glass/plastic coverslips (washed as above) and/or 10-mm-diameter Magna R nylon (Fisher Scientific) discs were used. These substrates were affixed to the Plexiglas sheet using a double-sticky, nontoxic tape. The sheets were attached via plastic paper clips to nylon string tied to braces traversing a 1 × 1 m PVC frame. These lightweight frames were underlined by a nylon window screen to help support the sheets and allow for confluence of the water column with the substrate below. The Plexiglas sheets were positioned 50 cm apart on the frame. Throughout this study, the frames were submerged at the study site at an average depth of 25 cm and the Plexiglas sheets were positioned 50 cm apart on the frame.

Biofilm Analyses

For short-term analysis of biofilms, the Plexiglas sheets were collected on 1, 2, 4, 8, 14, and 28 days after being seeded in the wetland during July and August, 2004. For the long-term study, the sheets were seeded on May 1 and collected at 15, 30, 60, and 90 days thereafter. Substrates used for dry weight and biochemical analyses were briefly submerged in wetland water to remove any gross attached materials not part of the biofilm. The biofilms were then scraped off the Plexiglas sheets with a Teflon razor blade into 15-mL sterile plastic centrifuge tubes. Microscopic examination using an Olympus IX-70 inverted microscope (×40 objective, fluorescence optics, Plexiglas surface labeled with SYTO9) of the Plexiglas surfaces after scraping revealed that most, if not all, of the biofilm organisms were removed. The samples were stored on ice and transported back to the laboratory for subsequent processing. At least five sheets were processed for each collection period. Sheets containing biofilms for microscopy-based studies were placed in wetland water in tupperware plastic containers and transported immediately to the laboratory.

Dry Weight and Biochemical Analyses

Samples containing biofilms were centrifuged with an IEC clinical centrifuge (5,000×g for 5 min) to separate biofilm mass from wetland water. The water was decanted off and the samples with the biofilm pellets were plunge-frozen in liquid nitrogen (LN2). The samples were stored at −80°C until further use. For dry weight analyses, the tubes were freeze-dried and weighed. Freeze-dried biofilms were also assayed for chlorophyll a [35], total carbohydrate (glucose was used as a standard [20]), and total protein (Bio-Rad kit 500-0001, Hercules, CA; bovine serum albumin was used as standard). Biofilms from at least five Plexiglas sheets were used for all collection periods for each dry weight and biochemical assays.

Water Chemistry

Commercial water testing kits from Lamotte (Cherstertown, PA) or Hach (Ames, IA) were used for testing for total hardness, total silica, total phosphate, and total nitrate. Temperature and dissolved oxygen analyses employed a Fisher Accumet AP74 meter; pH was measured with a Fisher Accumet AP61 meter, and conductance was monitored with an Oaklon CE conductivity meter (Fisher). Except for temperature, at least three independent readings were made per collection period.

Light (LM) and Confocal Laser Scanning Microscopic (CLSM) Analyses

Coverslips containing biofilms were analyzed using standard fluorescence microcopy and CLSM. For the former, an Olympus BX-60 light microscope (LM) was used for general and fluorescence microscopy and images were captured using an Olympus DP70 camera. For CLSM, an Olympus BX-61 light microscope with an attached Fluoview 300 software was used. To evaluate bacterial distribution and numbers, coverslips were placed in a 1-mL solution of a 1-μg/mL SYTO 9 dye (Molecular Probes, Eugene, Oregon, USA) in 0.45-μm filter sterilized wetland water. After 1 min incubation, the coverslip was removed, gently washed with sterile water, and viewed with the CLSM. To assay bacterial numbers and eukaryotic algae (via autofluorescence of chlorophyll), an argon laser (blue light laser; 488 nm) and fluorescein isothiocyanate (FITC) filter set were used and for 3-dimensional constructs, 1- to 2-μm slices were captured and digitally stacked using the Fluoview software.

For algal census studies and identification work, biofilms were removed from the plastic substrates, flash frozen in LN2, subsequently freeze-dried and weighed. Before microscopic analyses, the biofilm was rehydrated in 2 mL of water and vortexed for 60 min to separate cells and other materials. Fifty-microliter samples were then removed, placed on a glass slide, and viewed with an Olympus IX-70 inverted microscope (×10 objective). A built-in stage micrometer that identified 1 mm2 “transects” was used for the counts. Ten different areas were examined for each slide and duplicate samples were run per specific biofilm. Total enumeration was described as “units”, which consisted of all eukaryotic algae and cyanobacteria. A filamentous or colonial alga was considered to be 1 unit. Specific unit counts were also made of all desmids, all other non-desmid green algae, diatoms, cyanobacteria, and “other”, which included mostly euglenoids and dinoflagellates. All counts were averaged. Identification of desmids employed standard taxonomic texts [5761] and an internet resource (AlgaeBASE at http://www.algaebase.org).

We counted the total number of algae seen, calculated the percentage each constituted of the total unit count, and grouped them all into five categories, which were then used to generate histograms and diversity indices. We used the Shannon–Wiener Index [63] to assess the order (or disorder) observed within the biofilm community. This order is characterized by the number of individuals observed for each of the five groups in the sample plot. In short, the following were used: the Shannon–Wiener value was calculated as H = −sum (Pilog[Pi]) (for any base) and the Evenness value was calculated as H/log(S) (for any base). Pi is the number of a given taxon divided by the total number of organisms observed. An online internet resource was used for the calculations (http://www.mdsg.umd.edu/Education/biofilm/diverse.htm#3).

Variable Pressure Scanning Electron Microscopy (VPSEM)

For VPSEM, nylon discs containing biofilms of various ages were immediately plunge-frozen into LN2 after collection. The discs were placed on an LN2-cooled cryostub (JEOL, Peabody, MA), and viewed with a JEOL 6480 VPSEM at 10–40 Pa and 5–20 kV (with backscattered electron detector).

Bacterial Enumeration in Biofilms

To study the distribution and density of bacteria, wetland biofilms of different ages were labeled with SYTO 9. CLSM-based 3-dimensional constructs of the biofilms were then made and 4 cm2 squares were digitally generated on each image using Image ProPlus software (Media Cybernetics, Silver Springs, MD). Placement of the squares was upon zones containing the EPS sheaths of desmids and zones removed from desmids and EPS sheaths. At least five zones were analyzed per biofilm desmid with five biofilms analyzed per collecting period. Coverage of white fluorescence generated by SYTO 9 labeling of bacteria versus the black background of the substrate was calculated. Averages of percent coverage of bacteria were then generated. To ascertain the bacterial flora on desmids in the absence of EPS, individual desmids were isolated from biofilms using a pulled micropipette and placed into a 1.5-mL microcentrifuge tube containing 0.5 mL of sterile wetland water. The tube was vigorously shaken for 1 min and then centrifuged at 1000×g for 5 min on a Fisher Micro 14 microcentrifuge (Pittsburg, PA). The supernatant was removed and cells were resuspended in 0.5 mL sterile wetland water, shaken and centrifuged two more times. The cells were resuspended in 1 mL of sterile wetland water, labeled with SYTO 9 (1 μl stock SYTO 9 to 1 mL of cell suspension) and then viewed with CLSM as previously described.

Results

A general survey of the basic water chemistry, biofilm biochemistry, and biofilm characteristics was performed in this study and is presented in Table 1. Over the course of the long-term study (i.e., 15–90 days), several distinct observations were made. Water temperature varied from 8.6°C in early June to 20.6°C in early July and subsequently stabilized for the rest of the study period. The pH was notably constant ranging from 6.4 to 6.8 and little or no nitrate (N), phosphate (P), or conductance were detected. Dissolved oxygen (DO) levels were highest during the spring season with low points during the mid-summer. Total algal and cyanobacterial unit counts increased throughout the study and peaked at 90 days. The total number of units increased by a factor of 1.50 times from 15 to 30 days, then by a factor of 2.45 times from 30 to 60 days and finally by a factor of 1.73 times from 60 to 90 days. Biofilm dry weight levels increased from 15 days to 30/60 days but decreased significantly at 90 days. Chlorophyll a peaked at 30 days, then decreased at 60 days where it remained steady throughout the 90-day sampling period. Carbohydrate levels, protein levels, and bacterial numbers also experienced peaks during the mid-phases of this study with subsequent drop-offs.
Table 1

General chemical and physical characteristics of study site and inclusive biofilms

D

T

pH

DO

P

N

Si

H

C

Chl

DW

Bt

Cb

Pr

TC

Bc

Long-term 90-day study

15

16.5

6.5

8.5

0

0

0

18

0.1

0.6

8.9

54.5

35.7

3.2

173

14.3

30

8.6

6.4

11.9

0

1

2

22

0.1

1.5

12.5

37.5

61.1

6.5

255

32.0

60

20.6

6.8

4.8

0

0

2

26

0.1

0.7

21.8

43.1

58.1

1.9

625

27.1

90

18.8

6.7

4.8

0

1

4

28

0.1

0.8

9.8

45.6

55.3

2.4

1,086

25.0

Short-term study—July

1

15.7

6.6

6.1

0

0

2

20

0

0.1

0.4

-

7.0

0.5

77

1.5

2

15.7

6.6

6.1

1

0

2

18

0.1

0.2

5.0

2.2

20.3

1.9

218

4.5

4

15.7

6.6

6.3

0

1

2

24

0.1

0.2

3.6

21.8

25.0

3.3

208

12.5

8

15.8

6.6

6.1

1

0

4

20

0.1

0.6

11.2

25.3

29.4

2.6

261

17.0

14

21.1

6.7

4.1

0

0

4

28

0.1

0.5

8.1

19.8

24.1

3.1

408

23.0

28

15.4

6.5

7.5

0

1

4

28

0.1

1.0

16.3

20.8

29.5

2.8

268

21.0

Short-term study—August

1

20.7

6.7

5.0

0

0

4

28

0

0.3

0.6

-

5.0

0.5

5

3.2

2

20.4

6.8

4.6

0

0

6

28

0.1

0.3

2.6

1.8

19.5

3.2

18

4.4

4

18.8

6.8

4.2

0

0

6

24

0

0.2

4.5

22.1

28.8

2.8

123

9.5

8

19.5

6.7

4.8

0

0

6

24

0

0.5

10.1

28.1

40.6

3.1

89

17.2

14

18.0

6.8

7.5

0

0

8

28

0.4

0.8

11.0

67.0

42.1

2.9

76

21.3

28

15.1

7.0

8.0

0

1

16

28

0.1

0.9

19.3

39.0

58.8

3.4

110

25.0

D = day of sampling; T = temperature in °C.; DO = dissolved oxygen in mg/L, ±0.8; P = total phosphate in mg/L, ±0.5; N = total nitrate in mg/L, ±0.5; Si = total silica in mg/L, ±1; H = total hardness in mg/L, ±2; C = conductance in μs, ±0.1; Chl = chlorophyll a amount in mg/121 cm2 of biofilm, ±0.1; DW = dry weight of biofilm in mg/121 cm2 substrate, ±2; Bt = biofilm thickness in μm based on CLSM measurements, ±4; -=non-measurable; Cb = average percent carbohydrate of biofilm, ±3; Pr = average percent protein of biofilm, ±0.4; TC = average total count of algal and cyanobacterial units/cm2 of biofilm, ±10; Bc = bacterial counts based on Syto 9 labeling × 1000, in 1 cm2 of biofilm, ±0.2.

Two 28-day-long (i.e., short-term) studies were also performed to elucidate biofilm changes over shorter time periods and to focus on the time when desmid numbers were highest in biofilms (i.e., based on previous years’ surveys). In our July study (Table 1), water temperature, DO and pH were quite constant with the only noticeable changes occurring at 14 days. Conductivity, nitrate and phosphate levels were consistently low. Chlorophyll a increased slowly during the first phases of this study (1–4 days) and then in greater increments from 8 days onward. Dry weight amounts peaked at 8 and 28 days, whereas biofilm thickness peaked at 8 days with small decreases thereafter. Bacterial numbers increased throughout the study period with a slight decrease at 28 days. Carbohydrate levels peaked at 8 days and protein levels peaked at 4 days. Total algal and cyanobacterial unit numbers increased to a peak at 14 days and then experienced a 35% drop off by 28 days.

Throughout our August study (Table 1), water temperature and pH varied slightly. DO levels were relatively constant early on (i.e., up to 8 days) followed by a significant increase at 14 and 28 days. Conductivity and nitrate and phosphate levels were also low. Silica levels increased from 4 mg/L at 1 day to a high of 16 mg/L at 28 day. Chlorophyll a levels displayed two peaks with highs at 2 days and 14/28 days. Dry weight amounts and bacterial numbers increased throughout the study period, whereas biofilm thickness dramatically peaked at 14 days with a subsequent drop-off at 28 days. Carbohydrate levels increased significantly throughout the August study. Protein levels increased significantly from 1 to 2 d and then remained relatively constant. Our algal and cyanobacterial census during August was remarkably different than that of July. First, the total number of units counted was much lower during August. However, more colonial and filamentous forms were observed in August and counted as only one unit while occupying a larger volume on the biofilm surface. To illustrate, at 28 days the total number was only 41% of that recorded at 28 days for July. At 14 days, the total counted was only 18.6% of that recorded for 14 days in July.

Distribution and numbers of algal units were monitored over the course of the long-term study (Fig. 1). For algae and cyanobacteria, except diatoms, peak numbers were observed during the first 60 days of study, with a subsequent drop-off at the last sampling period. Desmid levels peaked at 60 days where they constituted 11.2% of the total algal and cyanobacterial unit numbers. Desmid numbers increased 4.6 times from the 30-day period to the 60-day period. This period of time corresponded with a 3.4 times increase in other green algae (e.g., Bulbochaete, Mougeotia, Spirogyra, Coleochaete). However, their numbers and percentage per total units counted dropped notably at 90 days. Diatom numbers increased significantly throughout the sampling period and peaked dramatically at the termination point of this study. For example, after 60 days, all non-diatom taxa constituted 34.2% of the total units counted whereas by 90 days, they constituted only 9.8% of the total counted units.
https://static-content.springer.com/image/art%3A10.1007%2Fs00248-007-9253-y/MediaObjects/248_2007_9253_Fig1_HTML.gif
Figure 1

Long-term study. Y-axis = % “units “ counted. X-axis = % “units” counted of each group per day of the study.

In the short-term July study (Fig. 2), diatom numbers were highest of any group and their percentage ranged from 61.0 to 69.7% of the total taxa counted. Desmid numbers increased from days 1 to 8 where they peaked at 24.0% of the total taxa counted. This was followed by a slight drop-off at 14 days, but a subsequent increase at 28 days. Cyanobacterial numbers and non-desmid green algal numbers were consistently low during the entire month. It is interesting to note that those algae grouped in the “other category” were quite significant in the early sampling period (e.g., 24.6% of the taxa counted at 1 day and 23.4% at 2 days), but dropped off notably in the latter stages of the study. Euglenoids and dinoflagellates comprised most of the algae found in the other category.
https://static-content.springer.com/image/art%3A10.1007%2Fs00248-007-9253-y/MediaObjects/248_2007_9253_Fig2_HTML.gif
Figure 2

Short-term July 2004 study. Y-axis = % “units” counted. X-axis = % “units” counted during each day of the study.

In the August study (Fig. 3), desmids and non-desmid green algae constituted the majority of units counted up till 8 days and then dropped to 36.8% and 44.5% of the total count for the last two sampling days. Desmids were one of the few organisms noted during the first day of sampling and ultimately constituted between 16.8% and 23.6% of the total units counted. Diatom numbers were significantly lower during the month in comparison to July. For example, at 14 days, diatom numbers were 6.5 times higher in July than in August. The units counted as “other” were insignificant at first, but made up 14.5% of the units counted at 28 days. Finally, cyanobacteria numbers became noticeable at 28 days.
https://static-content.springer.com/image/art%3A10.1007%2Fs00248-007-9253-y/MediaObjects/248_2007_9253_Fig3_HTML.gif
Figure 3

Short-term August 2004 study. Y-axis = % “units” counted. X-axis = percent “units” counted per group each day of the study.

Two desmid genera, Closterium and Cosmarium, constituted 75% of all desmids counted in July and 56% of those counted in August. Pleurotaenium species became a major desmid taxa in August, which constituted 26% of all desmids counted. Other desmid genera that were noted during this study include: Euastrum, Staurastrum, Penium, Netrium, Micrasterias, Arthrodesmus, Bambusina, Desmidium, Onychonema, Gonatozygon, Hyalotheca, Spondylosium, and Xanthidium. During late August, when non-desmid green algae were also found in noticeable numbers, the genera found included Bulbochaete, Coleochaete, Mougeotia, and Spirogyra.

For the long -term study, the Shannon–Wiener Index values indicated greatest diversity and evenness at 15 days, with the least diversity and least evenness at 90 days. H values ranged from a low of 0.447 at 90 d to a high of 1.483 at 15 days (Fig. 1) as diatoms predominate by 90 days. For the short-term study in July, diversity was highest at 1 day and lowest at 14 days. The greatest evenness was seen at 4 days (Fig. 2). In August, a very different pattern emerged. The greatest diversity was seen at 28 days with an H value of 1.552, with the next highest H value of 1.307 at 4 days. The lowest H value was 0.673 at 1 day (Fig. 3).

Previous research has shown that desmids produce large amounts of EPS [16, 17] and that distinct microbial communities are associated with their EPS sheaths [21, 22, 23]. In this study, we wanted to elucidate community structure of biofilms containing desmids and their large EPS “sheaths” as well as the changes in microbe distribution and numbers over time. We performed coverage analysis of bacteria in the biofilm using CLSM-generated images of SYTO-9-labeled biofilms collected throughout the study period. First, we undertook a general study of bacterial growth and biofilm development. Figure 4 illustrates this development in our 28-day July study. Bacterial colonization of new substrates occurs quickly within the first 2 days. By 4 and 8 days, thick bacterial lawns cover the substrate. This corresponds with significant increases in both bacterial counts and biofilm thickness during this period of study (Table 1). By 14 days, the density of bacteria remains high, but clumping and patchiness, i.e., regions with dense bacterial growth interspersed with less densely populated zones, are also found. This patchiness was often associated with algae that were part of the biofilm community. By 28 days, bacterial distribution was less dense, but still considerable. Comparable profiles of bacterial numbers over time were also noted for our August and long-term study.
https://static-content.springer.com/image/art%3A10.1007%2Fs00248-007-9253-y/MediaObjects/248_2007_9253_Fig4_HTML.gif
Figure 4

CLSM profiles of bacterial growth upon the Plexiglas substrate over the 28-day July study period. Within 1 day, bacteria begin colonizing the substrate surface (arrows). At 2 days, bacterial numbers grow and eukaryotic microbes enter the developing biofilm (arrows). By day 4, bacterial numbers increase significantly and rich lawns of bacteria develop. Dense bacterial lawns interspersed in some zones around eukaryotes (arrows) are also apparent at 8 and 14 days. By 28 days, the density of bacteria decreases and patchiness in distribution is clearly seen. Bar for all images = 50 μm.

Second, we performed similar imaging protocols specifically in biofilm zones containing desmids. In this article, a greater density of bacterial numbers occurred around the desmids, apparently in their EPS sheaths. In a 1-day-old biofilm, bacterial numbers are clearly larger around the desmid, Pleurotaenium sp., than in zones removed from the desmid (Fig. 5). The EPS secreted from one pole of the cell has a rich assortment of bacterial residents. Likewise, in older biofilms, a rich bacterial flora was also located surrounding the common desmid, Cosmarium reniforme, ensheathed within the biofilm matrix. A simple post-imaging analysis of bacterial coverage was employed to provide semiquantitative comparative data of bacterial distribution around desmids versus zones removed from desmids. In an analysis of over 30 biofilms ranging in age from 1 to 28 days, bacterial numbers were 300–1,100% higher in zones immediately outside desmids (e.g., EPS sheaths) than in zones 50 μm or more removed.
https://static-content.springer.com/image/art%3A10.1007%2Fs00248-007-9253-y/MediaObjects/248_2007_9253_Fig5_HTML.gif
Figure 5

Bacterial associations with biofilm desmids. Subpanel a demonstrates the notable association of bacteria (arrows) around the periphery of the desmid, Pleurotaenium, from a 1-day-old biofilm. Note that bacterial numbers are noticeably larger around the desmid than in regions removed away from the cell. Bar = 20 μm. Subpanel b is a magnified view of the bacterial association with the recently secreted EPS of Pleurotaenium (*). Note the close association of the bacteria with the EPS sheath (arrows). Bar = 25 μm. Subpanel c is a profile of an 8-day-old biofilm containing the desmid, Cosmarium reniforme (*). Note the high density of bacteria just outside the cell in the EPS sheath (arrow). Bar = 50 μm. Subpanels d–f illustrate the bacterial flora found upon the cell walls of various desmids that were removed from the biofilms and washed free of EPS. They include Cosmarium sp. (d), Euastrum sp. (e) and Pleurotaenium sp. (f). (Note the distinct coating of bacteria found upon the surface of each desmid arrows). In f, note the numerous linear associations of bacteria upon the desmid cell wall. Bar in d = 8 μm, e  = 7 μm, f = 3.5 μm.

We analyzed the distribution of bacteria upon the surface of desmids removed from biofilms and washed free of EPS (Fig. 5d–f). Rich bacterial flora were found attached to the cell wall surfaces of all 50 desmid samples, which represented over 10 different genera. Noticeably large numbers of bacteria were especially apparent on the surfaces of desmids found in 4-day or older biofilms.

VPSEM and cryofixation-based preparation of field-collected biofilms was a valuable and efficient tool in imaging the early stages of biofilm development as well as the general microbial diversity in native biofilms. This technology allows for rapid processing of biofilms without the laborious and sometimes damaging effects of more conventional techniques like chemical fixation, dehydration, critical-point drying, and/or lyophilization. During peak times for desmids within the biofilms (e.g., July, 8 days), a large assortment of Closterium species can be found within the biofilm complex (Fig. 6). In addition to larger-sized desmid taxa, biofilms also contain small, filamentous taxa. More importantly, desmids are early colonizers of new substrates as exemplified by one of the most common taxa in our biofilms, Cosmarium reniforme. This desmid moves upon the substrate by a gliding mechanism that is generated by the localized secretion of EPS (Fig. 7). Once positioned on the substrate, the cell secretes the EPS in a more generalized fashion resulting in an ensheathment of the cell onto the substrate. C. reniforme is also one of the desmids that is a frequently encountered resident of older biofilm communities. In 28-day-old biofilms, it was commonly found among the biotic and abiotic constituents of biofilm communities.
https://static-content.springer.com/image/art%3A10.1007%2Fs00248-007-9253-y/MediaObjects/248_2007_9253_Fig6_HTML.gif
Figure 6

VPSEM analysis of desmid distribution. Subpanel a demonstrates the large numbers of desmids (arrowheads) within an 8-day-old biofilm. Bar = 150 μm. In b, although large desmids like Closterium (*) are most noticeable, small desmids including many filamentous forms (arrows) are found within the biofilm matrix. Bar = 35 μm.

https://static-content.springer.com/image/art%3A10.1007%2Fs00248-007-9253-y/MediaObjects/248_2007_9253_Fig7_HTML.gif
Figure 7

VPSEM analysis of substrate colonization by Cosmarium reniforme. Subpanel a illustrates the surface of a 1-day-old biofilm. Note that in addition to other microbes, the desmid, C. reniforme (arrows) also colonizes the new substrate. Bar = 55 μm. Subpanel b demonstrates the secretion of EPS during gliding by C. reniforme in a 2-day-old bioflm. Bar = 23 μ. Subpanel c shows that once gliding has stopped, a more generalized secretion of EPS covers the cell surface (arrows) and ensheaths the cell to the substrate. Bar = 10 μm. Subpanel d represents a 28-day-old biofilm. Note that C. reniforme (arrow) is part of the complex microbial community. Bar = 25 μm.

Discussion

Freshwater wetlands are habitats of profound importance in the context of both local and global ecology [16, 36]. They support a rich assortment of photosynthetic organisms including microorganisms such as algae. Wetland algae appear to play a major role in ecosystem dynamics [28], but we are only just beginning to elucidate the specific algal residents and their activities in the context of the “biofilm” community [2]. Previous research on these microbial communities and subsequent development of biofilm models has been especially difficult in that wetlands are dynamic and shallow ecosystems, which experience rapid fluctuations in water level, water chemistry, temperature, O2 levels, and other parameters.

In this study, we initiated an analysis of wetland biofilms of the Adirondack region of NY with special emphasis on desmids, a group of green algae commonly encountered in these wetlands. Our study site could be characterized as a circa-neutral fen with low P, N, and conductivity, i.e., typical of wetlands that contain large and diverse desmid populations [6]. Desmids constituted significant portions of the total algal and cyanobacterial flora of the biofilms. In the short-term July study, desmids constituted 23.7% and 22.7% of the total units counted at 8 and 28 days, respectively. In August, desmids constituted up to 23.6% of the total units counted at 28 days. In addition, desmids were found to be some of the first algae to colonize new substrates. These data support previous studies that also highlighted the large numbers of desmids in benthic attached communities, periphyton assemblages and biofilms of aquatic ecosystems. Wantanabe et al. [70] reported large desmid populations in biofilms of an acidic mire. In the month of June, desmids constituted 63% of the total algal flora. Woelkerling and Gough [74, 75] also showed significant desmid populations in softwater systems where in 6 out of 35 acid bogs of a Wisconsin study site, desmids accounted for over 20% of the algal flora (with a high of 27.5%). In addition, as many as 16,300 desmid epiphytes were counted per milligram of the aquatic macrophyte, Utricularia. In a summertime study of benthic attached algae of a softwater stream, Burkholder and Sheath [9] showed that as many as 1,700 desmids were found per square centimeter of substrate. In the Everglades of Florida (USA), periphyton biofilms may be responsible for 80% of the primary productivity in shallow oligotrophic areas. In the northern Everglades region of Florida (e.g., the Loxahatchee National Wildlife Reserve), desmid-rich assemblages characterize the biofilms [8, 4648]. All of these results indicate that desmids may represent significant portions of the photosynthetic microbe population of wetland biofilm communities, particularly those characterized by low conductance and low nutrient waters. These wetland types are very common in the Northern Hemisphere and include many bogs, fens, tundra, taiga kettle holes, and cedar bogs. Our study adds to this pool of data amd specifically shows that desmids are important in biofilms of rapid-successional wetland habitats, such as those engineered by beavers [50, 51, 62, 76]. These unique wetlands are of particular interest because they are increasing in number as beavers return to areas where they had been previously extirpated (e.g., the Adirondacks). Research on these wetlands provides insight not only into the basic ecology of a wetland type that was very common in the North American landscape before the mid-1500s, but also affords us a rare snapshot of microbial primary productivity and food chains in our modern era where the need to preserve all types of wetlands is recognized.

Eukaryotic algae that are residents of wetland biofilms might be expected to have particular adaptive features necessary for biofilm existence. For instance, one might hypothesize that photosynthetic eukaryotes have morphological adaptations that enhance light absorption in benthic locations, i.e., zones located on submerged substrates where light quantity and quality are quickly filtered out in the water column. The morphology of desmids exemplifies some of these adaptations. For example, desmids are relatively large algae with sizes ranging from 20 to 40 μm to well over 300 μm. They also display distinct shapes that maximize surface area, including elongate cylinders (e.g., Closterium, Penium), compressed semicircles (e.g., Cosmarium), or highly dissected, flattened forms (e.g., Micrasterias). In addition, most desmids, including those found in biofilms, possess large and extensively lobed chloroplasts that fill much of the inner volume of the cell. These features would provide desmids with maximum plastid surface area that allow for enhanced light-capturing efficiency for photosynthesis.

One might further hypothesize that biofilm desmids, like all biofilm organisms, would be expected to possess structural and biochemical features and mechanisms that effectively allow them to attach to surfaces and ultimately incorporate into the biofilm community. In this study, VPSEM analysis of biofilm formation over time revealed that, in fact, desmids (e.g., Cosmarium reniforme) are able to colonize new substrates quickly and become part of the biofilm community. Previous research has shown that desmid EPS possesses significant adhesion properties [1, 17], which allows the organism to quickly and effectively attach to a substrate. Once attached, desmids often glide, a process whereby hygroscopic swelling of locally secreted EPS generates the force necessary for movement [3, 1719, 54] upon a solid substrate. This phototaxis-based motility mechanism [3133] allows the desmid to position itself favorably upon the surface in relation to light. Finally, a more generalized secretion of EPS occurs along the entire cell surface to ensheath the cell to the substrate. Despite our fundamental understanding of EPS secretion in desmids, the actual mechanisms responsible for a desmid “sensing” a new substrate and colonizing it remain unknown. In studies of bacterial biofilms of aquatic ecosystems, it is generally assumed that the ubiquity of bacteria in the planktonic realm allows for a ready source of potential colonizers of new surfaces and the ultimate formation of biofilms. That is, the large numbers of bacteria found throughout the water column significantly increase the frequency of contact between the microorganisms and their substrate. However, this is not true for many large eukaryotes like desmids, whose numbers are much lower than that of prokaryotes. In addition, many of the biofilm desmids from our study site are not commonly found in the planktonic realm. It is possible that physical disturbances within the wetland could dislodge many desmids from pre-existing biofilms and place them into the water column (i.e., planktonic realm). Subsequent sinking might then place the desmid onto new substrates. However, it is also possible that desmids may move onto new substrates via their gliding mechanism. Desmids have been shown to glide considerable distances both upon surfaces [54] (e.g., centimeter distances over short periods of time) and vertically within soft substrates [6, 7, 27]. This motility mechanism is similar to those reported in other photosynthetic eukaryotes found in biofilms (e.g., diatoms) where gliding mechanisms sometimes account for significant lateral and vertical movements through and upon substrates and biofilms [21, 34]. The stimuli and signal transduction mechanisms that would be involved in recognizing suitable substrates and the accompanying mechanisms that direct gliding movements toward that substrate are not known and require further research. This would include investigations of chemotactic stimuli such as conditioning films or chemicals and thigmotactic signals regulating gliding.

Recent research has shown that benthic attached algae of aquatic ecosystems may excrete 5–40% of their total photosynthate, of which 30–90% may be subsequently used by heterotrophic bacteria [5]. This demonstrates that benthic algal exudates are important to fixed carbon flow and food chain dynamics [42]. With respect to biofilm desmids, photosynthate may be expressed in two forms, cell exudates and whole cell mass. For the former, EPS is the major secreted material (e.g., microgram per cell quantities have been demonstrated in laboratory biofilm cultures) and is used for major biofilm-related behaviors [1719]. In this study, we have demonstrated distinct populations of bacteria concentrated in and on the EPS sheaths as well as upon the cell wall surfaces of desmids, especially in biofilms of 14 days or older. This supports previous research that has also reported distinct and diverse bacterial flora associated with the EPS sheaths of desmids from wetlands [2224] including taxa from the subdivisions of the Proteobacteria and the Flexibacter/Cytophaga/Bacteroides phyla, Rubrivivax/Rhodocyclus group—non-sulfur purple photosynthetic bacteria, N2-fixers, and facultative H2-oxidizing bacteria. Whereas desmid EPS may simply serve as a physical matrix for microbial colonization, it may also serve as a major carbon substrate for wetland heterotrophs. In many ways, this would parallel the bacteria-rich biofilms that are found in the rhizosphere of higher plants, a rich organic “oasis” of root-derived secretory materials [69]. Further research dealing with carbon flow and food chain dynamics is required to elucidate the role of desmids in wetland food chains.

Our structural and biochemical analyses of the wetland biofilms highlighted some unusual features. During July, noticeable parameter fluctuations occurred including: increased biofilm thickness and dry weight up to 8 days, followed by modest decreases in biofilm thickness, dry weight, and carbohydrate content at 14 days, rebounding of dry weight, biofilm thickness, and carbohydrate levels at 28 days and increased desmid numbers at 4 and 8 days, followed by a modest decrease at 14 days and a subsequent increase at 28 days. These results suggest that fluctuations in biofilm structure and biochemistry may correlate with concurrent changes in desmid numbers and desmid EPS production in the biofilms. Previous studies [40, 41] have shown that green algae contain distinctly higher levels of carbohydrate than other types of algae and that desmids produce prolific amounts of EPS [17, 18, 54]. The results of this study suggest that increasing desmid numbers may indeed change the structural, biochemical, and nutritional quality of biofilms. However, in our short-term August study, major fluctuations were not noted as dry weight, biofilm thickness, and carbohydrate levels all increased throughout this study period. It is also important to note that the numbers of algae and cyanobacteria were much lower in August than in July, but that bacterial numbers were slightly higher. These results may indicate that during August, nutrient limitation (i.e., little nutrient influx from surrounding terrestrial zones during low rainfall periods) and higher temperatures may keep algal populations down, but would not necessarily affect bacterial numbers that could be using pre-existing biofilm material.

Distinct fluctuations of several parameters were also noted in our long-term biofilm study including an increase in most biofilm parameters during the early part of the study and subsequent and significant decreases in biofilm thickness from 15 to 30 days. Chlorophyll a and protein levels increased from 30 to 60 days. Bacterial numbers increased from 30 to 90 days and dry weight also increased from 60 to 90 days. These results may reflect two possible situations. First, as biofilms grow in size and change in composition over time, physical pressures may cause the removal of biofilm biomass from its substrate. For example, as a biofilm ages, the distribution and physiological activities of its inclusive microbial residents as well as concurrent biochemical modifications to the EPS-based matrix may weaken the biofilm’s infrastructure. This would then allow for easier fragmentation by physical forces such as water column disturbances around the biofilm. Likewise, as a biofilm ages, there is an increased likelihood that macroinvertebrate grazing over time would remove significant biofilm biomass.

During the course of monitoring long-term biofilms, it is very apparent that diatoms become the main photosynthetic eukaryotes within the biofilm community. For example, desmid numbers peaked at 60 days (11.2% of the total units counted), but decreased significantly in 90-day-old biofilms where they constituted only 2.8% of the units. In these older biofilms, diatoms clearly represented the dominant algae. It is also important to note that diatom numbers are usually much higher in biofilms with high microbial population densities (e.g., long-term study and the July study) versus biofilms with lower numbers of microbial residents (e.g., August study). Diatom population increases in biofilms may very well correspond with increased levels of available silica found in the water during late summer. For example, during this latter part of the terrestrial growing season when surrounding grasses die and decompose, their cell-wall-based silica might be expected to leach out into the wetland water column (see August silica levels, Table 1). Diatom cell walls or frustules are also silica-complexed. The increased availability of silica during the late summer from terrestrial “runoff” may subsequently stimulate population growth of biofilm diatoms. In addition, the complex motility mechanisms of diatoms [21, 34] that allow for relatively quick and efficient movement upon substrates may also contribute to their competitive edge over other microorganisms for space in the biofilm matrix.

In addition to desmids and diatoms, the wetland biofilms displayed a rich and diverse assortment of other algae. First, in both our initial May collection (15-day long-term study) and in the last sampling time (28 days) of our August short-term study, filamentous green algae such as Bulbochaete and the conjugalean taxa Mougeotia and Spirogyra, were commonly observed. The distinct time periods during which these green algae were found in large numbers correspond to their seasonal blooms. For example, metaphyton blooms of filamentous conjugalean green algae often form large and attached metaphyton flocs during spring and fall seasons in wetlands [72]. These organisms rapidly die out during the summer months and often reappear in the fall. Our observations may very well have captured these organisms’ population spikes and their attachment to our biofilms in the pre- and post-summer seasons. Second, in many biofilm samples, we noted the presence of unicellular flagellated algae that dropped their flagella and incorporated into the biofilm matrix. This included euglenoids and dinoflagellates whose numbers represented significant percentages of the total counted units during early biofilm formation in July (e.g., 1 and 2 days, the beginning of our long-term study) and in late biofilms of our August study (28 days). However, during other times of our study, their numbers were very low. These organisms most likely represent transient visitors to the biofilms.

Finally, our study showed that new substrates inserted into our wetland study site were quickly colonized by microbes and measurable biofilms were established within a few days. For example, our CLSM analyses clearly showed bacterial attachment to our artificial substrates within 1 day and biofilms with thicknesses of 20 μm or more were established by 4 days. These results support previous research with aquatic biofilms where colonization and biofilm establishment occurred in short periods of time [52, 53]. Also, we have shown that biofilm thickness does not necessarily correspond to numbers of inclusive microbial flora. For example, in August, we measured biofilms with an average thickness of 67.0 μm. However, this was a time when algal and cyanobacterial numbers were relatively low and bacterial numbers were roughly equivalent to levels seen in much thinner films. It is likely that the microbial flora of this August biofilm produced significantly larger amounts of EPS that caused the spike in thickness, but outside events such as the deposition of extra-biofilm material in the water column (e.g., detritus) may have contributed as well. It is known that the EPS produced by many microbes consists of macromolecules that are charged (e.g., polyanionic polysaccharides). These charged polymers may act as effective “sponges” for removing materials from the water column.

We are only just beginning to recognize the significance of eukaryotic microorganisms in biofilms. In freshwater wetlands, a variety of photosynthetic eukaryotes contribute significantly to biofilms including desmids. The elucidation of these microorganisms in biofilm structure and metabolism will ultimately provide a more complete understanding of the microbial dynamics of these important ecosystems.

Acknowledgements

Special thanks to Kelly Dempsey-Little for her help in this project. Part of this study was supported by NSF grant 0419131 and Faculty Development Grants from Skidmore College.

Copyright information

© Springer Science+Business Media, LLC 2007