Hindcasting cyanobacterial communities in Lake Okaro with germination experiments and genetic analyses


  • Editor: Riks Laanbroek

Correspondence: Susanna A. Wood, Cawthron Institute, Private Bag 2, Nelson 7042, New Zealand. Tel.: +64 3 548 2319; fax: +64 3 546 9464; e-mail: susie.wood@cawthron.org.nz


Cyanobacterial blooms are becoming increasingly prevalent worldwide. Sparse historic phytoplankton records often result in uncertainty as to whether bloom-forming species have always been present and are proliferating in response to eutrophication or climate change, or if there has been a succession of new arrivals through recent history. This study evaluated the relative efficacies of germination experiments and automated rRNA intergenic spacer analysis (ARISA) assays in identifying cyanobacteria in a sediment core and thus reconstructing the historical composition of cyanobacterial communities. A core (360 mm in depth) was taken in the central, undisturbed basin of Lake Okaro, New Zealand, a lake with a rapid advance of eutrophication and increasing cyanobacteria populations. The core incorporated a tephra from an 1886 volcanic eruption that served to delineate recent sediment deposition. ARISA and germination experiments successfully detected akinete-forming nostocaleans in sediment dating 120 bp and showed little change in Nostocales species structure over this time scale. Species that had not previously been documented in the lake were identified including Aphanizomenon issatschenkoi, a potent anatoxin-a producer. The historic composition of Chrococcales and Oscillatoriales was more difficult to reconstruct, potentially due to the relatively rapid degradation of vegetative cells within sediment.


Cyanobacteria are natural constituents of lentic and lotic waters, but they appear to have become increasingly prominent in recent decades, possibly in association with anthropogenic eutrophication (Mur et al., 1999) and climate change (Paerl & Huisman, 2008). In lentic systems cyanobacteria can proliferate rapidly in response to adequate nutrient supply and elevated water temperature, with stratification being a key part of their surface accumulation (Oliver & Ganf, 2000). Cyanobacterial blooms are aesthetically unpleasant and can have serious environmental impacts (Paerl et al., 2001).

During periods when conditions are unfavourable for planktonic growth, many cyanobacteria persist in lake sediments as resting stages, either as short filaments (hormogonia), akinetes (resting stages) or vegetative cells (Head et al., 1998; Verspagen et al., 2004). Benthic populations that survive winter may provide significant inocula for the development of pelagic cyanobacterial populations (Preston et al., 1980; Brunberg & Blomqvist, 2003; Kim et al., 2005). It has been suggested that large overwintering populations are one reason why cyanobacteria with low specific growth rates (e.g. Microcystis) become dominant in summer phytoplankton communities (Reynolds, 1994). Analyses of surficial lake sediments can provide valuable forecasts of the potential species composition of pelagic blooms (Baker & Bellifemine, 2000; Faithfull & Burns, 2006), while sediment profiles may provide information on historical phytoplankton composition and abundance (e.g. Livingstone & Jaworski, 1980; Dickman & Glenwright, 1997; Tani et al., 2002).

Various techniques have been used to analyse phytoplanktons within sediment cores. Livingstone & Jaworski (1980) showed that akinetes from sediments deposited up to 64 years earlier could be germinated when incubated in culture media. Recently, molecular techniques have been successfully used to detect cyanobacteria (Innok et al., 2005) and cysts of eukaryotic algae (Coyne & Cary, 2005) from sediments. Automated rRNA intergenic spacer analysis (ARISA) is a recently developed DNA finger-printing method (Fisher & Triplett, 1999) that exploits the length heterogeneity of the intergenic spacer (ITS) region between the 16S and 23S rRNA genes. In this study, we used both germination experiments and ARISA assays to investigate the cyanobacterial community composition in layers of a sediment core taken from a eutrophic lake of volcanic origin in the Rotorua district of New Zealand.

A light-colored tephra deposited over the Rotorua district in the 1886 Tarawera volcanic eruption provides highly visual differentiation between lake sediments deposited before and after the Tarawera eruption (Nelson, 1983). Within this period, many Rotorua lake watersheds have been subject to European colonization and changes in land use from native forest and scrub to pastoral farming and plantation forestry. Correspondingly, nutrient loads to many of the lakes have increased and, for some lakes, there is a well-documented history of increasing trophic status, for example Lake Rotorua (White et al., 1985) and Lake Rotoiti (Vincent et al., 1984). However, there is little reliable historic information on phytoplankton species composition. Thus, it is difficult to ascertain if bloom-forming species have always been present in the lakes and have proliferated relatively recently, or if there has been a succession of cyanobacteria with new arrivals through recent history.

Lake Okaro is a small, monomictic, eutrophic lake in the Rotorua district of central North Island of New Zealand. It was formed as a hydrothermal explosion crater c. 900 years before present (Healy, 1964). Pastoral farming proliferated rapidly in the Okaro watershed in the 1950s (Jolly, 1968) and by 1970 around 95% of land use in the watershed had been adapted for pastoral farming (McColl, 1972), similar to present day land use. Compared with other New Zealand lakes, Lake Okaro has a long limnological data record, dating back to the 1950s (e.g. Jolly, 1959; Fish, 1969; McColl, 1972; Flint, 1977; Dryden & Vincent, 1986; Forsyth et al., 1988). This extended record includes a period when the lake changed from a continuously oxygenated hypolimnion during the 8-month seasonal stratification cycle, to being devoid of oxygen for all but 1 month. Correspondingly, there have been large increases in nutrient concentrations, increased relative abundance and biomass of cyanobacteria, and decline in diversity of littoral benthos (Forsyth et al., 1988). Since the 1970s, the lake has had seasonally recurrent cyanobacterial blooms (Dryden & Vincent, 1986). Gall & Downes (1997) investigated fossil pigments in a sediment core from Lake Okaro. The pigments myxoxanthophyll and canthaxanthin are specific to cyanobacteria. Neither of these pigments was detected in the core in the period estimated from 210Pb dating to be before 1900. A small increase in both myxoxanthophyll and canthaxanthin occurred between 1900 and 1950, followed by peaks of both pigments around 1965. The pigment analysis unequivocally showed an increase in cyanobacterial concentrations in the lake; however, it provided no information on possible changes in species composition.

The major objective of this study was to reconstruct the historical composition of cyanobacterial communities in Lake Okaro in order to provide a sedimentary record of their presence and long-term succession through a period of rapid progression of eutrophication in the lake. A secondary objective was to evaluate the relative efficacies of germination experiments and ARISA assays in reconstructing the assemblage of cyanobacteria through the sediment profile.


Sample site and core collection

Lake Okaro is a small (0.33 km2 surface area), shallow (maximum depth 15 m) eutrophic lake in central North Island (38°17′S, 176°23′E) of New Zealand (McColl, 1972). A sediment core of length 360 mm was taken from the deepest, central part of the lake with a cylindrical gravity corer (ø=100 mm) on 22 May 2006. Visual inspection through the acrylic barrel of the core indicated no disturbance of sediments into the overlying surface waters. The sediment was extruded from the core barrel in discrete 20-mm sections that were placed into sterile 50-mL Falcon tubes and stored at 4 °C in darkness.

Age of core layers

The Tarawera tephra was identifiable in the core as a discrete light grey region commencing 320 mm below the sediment surface and extending below 360 mm in our core. Gall & Downes (1997) had previously identified the Tarawera tephra at a depth of 190 mm below the surface in a 1995 sediment core collected from the same central location of Lake Okaro as part of a 210Pb dating and sediment plant pigment study. Alignment of the Tarawera tephra between cores allowed us to extrapolate their 210Pb dating data to our core. The average sedimentation rate between 1995 and the present day is estimated to be 12 mm year−1, allowing approximate calculation of sediment age in the upper section of the core.

Germination experiments

To assess the viability of akinetes and vegetative cells in the sectioned core, 1-g aliquots were taken from each Falcon tube, which contained the 20-mm layer of the core, and resuspended in Erlenmeyer flasks (100 mL) containing 75 mL of MLA medium (Bolch & Blackburn, 1996). The flasks were incubated in a growth cabinet under a light regime of 100 μmol m−2 s−1 with a 12 : 12 h light : dark cycle, at a temperature of 18±1 °C. All germination experiments were undertaken in duplicate. An aliquot (5 mL) of the medium was collected from each flask every 4 days over a duration of 20 days and preserved with Lugol's iodine.

Cyanobacterial identification from germination experiments and ARISA profiles

Identification of cyanobacteria was carried out using an inverted Olympus microscope (IMT-2) and Utermöhl settling chambers (Utermöhl, 1958). Identification to species level was made with taxonomic guides of Baker (1991, 1992), Baker & Fabbro (2002), McGregor & Fabbro (2001), Wood et al. (2004) and McGregor (2007).

Rueckert et al. (2007) and S.A. Wood (unpublished data) established a New Zealand-specific cyanobacterial ITS library allowing phylogenetic information (16S rRNA gene sequences) to be assigned to peaks in ARISA profiles. During germination experiments several species were observed for which no 16S rRNA gene or ITS sequence information was available. In these instances, 2-mL aliquots were collected, frozen and pelleted by centrifugation (18 000 g, 10 min). The supernatant was removed by sterile pipeting. DNA was extracted from the remaining pellets using the Invitrogen Purelink Genomic DNA Kit (Invitrogen, New Zealand) according to the gram-negative bacteria extraction protocol supplied by the manufacturer. The 16S rRNA gene and ITS sequences were determined by cloning and sequencing of PCR products as described in Rueckert et al. (2007). Sequences generated during this work were deposited in NCBI GenBank database under accession numbers EU402396–7.

Isolation of DNA from sediment

Subsamples of sediment taken from each 20-mm layer of the core were centrifuged (5000 g, 5 min) to remove excess water. The supernatant was removed using sterile pipeting and DNA was extracted from c. 0.25 g of sediment using the MoBio Power Soil kit (Carlsbad) according to the manufacturer's protocol.

ARISA fingerprinting and analysis

ARISA PCR reactions were carried out using cyanobacterial-specific primers as described previously (Wood et al., 2008). Amplicons were diluted 1 in 10 with sterile water, and 2 μL of product mixed with 0.25 μL of ROX-labelled genotyping internal size standard ETR900R (GE Healthcare, Auckland, New Zealand). The sample was made up to 10 μL with 0.2 v/v Tween-20 in sterile water. ITS lengths were determined by electrophoresis using the MegaBACE system (Amersham Pharmacia Biotech). Run conditions were 44 °C separation temperature, 10 kV voltage and 120-min separation time.

ARISA fragment lengths (AFLs) were analysed by Genetic Profiler V.2 (GE Healthcare) and data transferred to Microsoft Excel for further processing. All AFL information was transposed to the presence/absence data for further analysis. To account for occasional small shifts in AFL between analyses and ensure that species diversity was not over estimated, AFL that differed ≤2 bp were considered identical. If multiple AFL fell within this range, then only the AFL with the highest fluorescence was maintained (Wood et al., 2008). AFL falling below a threshold of 250 fluorescence units were considered ‘background noise’. AFL of <300 bp were considered to be too short for the ITS to be valid (Wood et al., 2008) and were removed from further analysis.

Anatoxin-a analysis

To investigate potential anatoxin-a production by one of the species identified in the germination experiments, Aphanizomenon issatschenkoi, aliquots (5 mL) of two cultures (from sediment depths 140–160 and 160–180 mm) were collected at day 20 and frozen (−20 °C). These samples were selected as they contained the greatest abundance of A. issatschenkoi at that time during the germination experiment. The samples were subsequently thawed and an equal volume of acetonitrile and formic acid was added to 0.1% v/v, and then extracted using sonication for 10 min. Following centrifugation (3500 g, 10 min) an aliquot of the supernatant was analysed directly for anatoxin-a using liquid chromatography-MS (LC-MS) as described in Wood et al. (2007a).


Germination experiments

A total of eleven different species of cyanobacteria were identified through microscopic examination. The species were from three different orders: Chroocococcales (3), Oscillatoriles (3) and Nostocales (5; Table 1). Cyanobacteria were identified in all sediment layers except the deepest (340–360 mm), which overlapped with the Tarawera tephra, commencing around 320 mm. The highest diversity occurred in the two-surface layers (0–20 and 20–40 mm), where all 11 species were found. Aphanizomenon issatschenkoi was the most common species through the depth profile, found in 15 of the 18 sediment layers (Table 1).

Table 1.   Cyanobacterial species identified in 20-mm sections of a sediment core from Lake Okaro
Sediment depth (cm)0–12–34–56–78–910–1112–1314–1516–1718–1920–2122–2324–2526–2728–3930–3132–3334–35
Estimated sediment date*20062004200220011999199719951987197619651954194319321921191119001886>1886
  • Germination experiments were undertaken in duplicate and samples collected every four days for 20 days, this table shows combined result for each layer.

  • *

    Sediment date estimated by alignment with 210Pb dated core of Gall & Downes (1997).

  • AFL given in parentheses.

  • AFL, ARISA fragment lengths.

  • Germination (x), ARISA (ITS 1 •, ITS 2+)

 Aphanocapsa sp.xx                
 Aphanothece sp.xx  x             
 Microcystis spp. (531)x •x •xx •x •x •            
 Synechocystis sp. (565)             
 Geitlerinema sp.xxxxx             
 Pseudanabaena limnetica (688)xx xx •   xx       
 Anabaena circinalis (460, 672)x •xx •+x •+x •+x •+x •+x •+x ••+•+x ••+ 
 A. lemmermannii (471)     
 Anabaena sp.xxxx xxxxxxxxx    
 Anabaena c.f. sp. Nova (450)xx •x •x •x •xxx •x •xx •x •xx •  
 Aphanizomenon gracile (440)x •x •x •xx •     
 A. issatschenkoi (421, 646)x •x •+x ••+x •+x x •+x •+x •+x •+x •+x •x •+x •+x •+x •+ 
Unassigned AFL
Total number of species via germination10107785244543321310
Total number of AFL876793287868886541

Eight days into the germination experiment, akinetes were visible in A. issatschenkoi, allowing definitive identification of this species (Fig. 1a and b). No akinetes were observed in other Nostocales species. However, distinctive features (i.e. terminal cell shape, spiral breadth) allowed the identification of Anabaena circinalis and Aphanizomenon gracile. A second Anabaena species was tentatively identified as Anabaena sp. Nova (Fig. 1c), as described in Baker & Fabbro (2002). Observations of specimens of this species conformed to the following morphological description: trichomes solitary and regularly spiralled coils of 20–35 μm width and closely compacted; vegetative cells spherical, slightly compacted at the poles, 6–7 μm breadth and with gas vesicles; heterocytes spherical and 7–8 μm breadth. A third much smaller Anabaena sp. was also identified in 14 of 18 sediment layers. The specimens observed conformed to the following morphological description: trichomes irregularly coiled and occasionally entangled; vegetative cells barrel-shaped, 3.5–5 μm in length and 2.5–3 μm in breadth (Fig. 1d).

Figure 1.

 Light photomicrographs of a selection of cyanobacterial species identified in 20-mm sections of a sediment core from Lake Okaro via germination experiments. (a, b) Aphanizomenon issatschenkoi, (c) Anabaena c.f. sp. Nova, (d) Anabaena sp. Scale bar=10 μm. A, akinete; H, heterocyte.

Four of the species identified in our germination experiment (Aphanothece sp., Geitlerinema sp., Anabaena c.f. sp. Nova and A. issatschenkoi) had not previously been documented among the Lake Okaro phytoplankton community (Dryden & Vincent, 1986; Bay of Plenty Regional Council, unpublished data).

Cyanobacterial ITS library

Two 16S rRNA gene and ITS sequences that were not in the New Zealand cyanobacterial ITS library (Rueckert et al., 2007; S.A. Wood, unpublished data) were obtained from the clone libraries. The 16S rRNA gene segment sequences were submitted to blastn (Altschul et al., 1997) to identify other highly homologous sequences. The partial 16S rRNA gene sequence (c. 1200 bp) from clone Okaro10 (GenBank EU402396) matched at >99% sequence homology to Pseudanabaena sp. 1tu24s9 and PCC7408 (GenBank AM259269 and AB039020). The partial 16S rRNA gene sequence (c. 1200 bp) from clone Okaro9 (GenBank EU402397) matched at >99% sequence homology to Anabaena sigmoidea 0tu36s7 and 0tu38s4 (GenBank AJ630434 and AJ630435). The sequence-derived AFL for these species were 688 and 450 bp.

ARISA analysis

Analysis of ARISA data for all samples produced a total of 19 distinct AFL. The number of AFL in each sample ranged from one in the deepest layer (340–360 mm) to nine (60–80 mm; Table 1). Unlike the germination experiment, the diversity did not decrease with depth. Of the 19 AFLs, 10 could be attributed to known planktonic cyanobacterial ITS lengths based on our current ITS library. The most commonly detected AFL (460 bp), identified in 16 samples, was attributed to A. circinalis. This was closely followed by the 421-bp AFL of A. issatschenkoi, detected in 15 samples (Table 1).

Anatoxin-a detection

The two subsamples collected from the germination experiment culture (layers 140–160 and 160–180 mm) tested positive for anatoxin-a using LC-MS.


Historic changes in cyanobacterial composition

A series of limnological observations in Lake Okaro extending back to the 1950s provide quantitative evidence of a progressive decline in water quality to the present time. The data include dissolved oxygen profiles, which show some oxygen remaining in the hypolimnion throughout seasonal stratification (1955–1956; Jolly, 1968) but declining to the point where sometime between 1961–1964, the hypolimnion was anoxic by the end of stratification. By 2005–2006, the hypolimnion was anoxic for all but 1 month of the seasonal stratification cycle (Paul et al., 2008). The phytoplankton community in Lake Okaro has been studied at irregular intervals over the past five decades and the data clearly show a shift in species composition. The most conspicuous of these shifts was the appearance of cyanobacteria. Jolly (1959) observed no colonial cyanobacteria in samples collected from Lake Okaro during 1955–1956. The earliest reports of bloom-forming cyanobacteria in Lake Okaro were in the 1960s when Anabaena was the dominant species (Fish, 1968). This observation corresponds with the findings of Gall & Downes (1997), who measured concentrations of myxoxanthophyll and canthaxanthin, pigments specific to cyanobacteria, in a sediment core taken from the centre of the lake. They found a marked increase in these pigments during the 1960s. In the 1970s and 1980s, Anabaena spiroides, Anabaena flos-aquae and Microcystis aeruginosa were all reported at various times among the species contributing to persistent blooms (McColl, 1972; Flint, 1977; Dryden & Vincent, 1986). Monthly monitoring of Lake Okaro since the 1990s has shown that blooms of Anabaena spp. and Microcystis spp. occur regularly in spring and summer (Bay of Plenty Regional Council, unpublished data; Paul et al., 2008).

The results of the ARISA and germination experiments clearly show that cyanobacteria have been a component of the phytoplankton community in Lake Okaro since the beginning of the century. Cyanobacteria were detected using both methods in the 320–340-mm layer, just above the Tarawera tephra, dating from 1886. The ARISA assay showed greater diversity throughout the sediment layers. One consideration when using molecular techniques such as ARISA, is that these methods detect the presence of genes (or gene fragments) and this does not necessarily correspond to viability (Coyne & Cary, 2005). A further consideration when interpreting the data from both methods is that different species vary in their tolerance to, and persistence in the sediments. Vegetative cells are not preserved well in lake sediments (Räsänen et al., 2006) and this may explain the paucity of Chroococcales and Oscillatoriales in deeper sediment layers. This was particularly apparent in the germination experiments, with the majority of species from these orders were not observed below 120 mm. Therefore, using analysis of sediment cores to document chronological changes for all species may be misleading. Microcystis spp., for example, were not detectable (via either method) below 120 mm, which equates to an approximate sediment date of 1995. However, historic records (Dryden & Vincent, 1986) show that this species had already formed blooms in Lake Okaro by 1979.

Certain cyanobacteria from the Nostocales and Stigonematales orders produce akinetes, i.e., resting spores. These cells have thicker walls making them more resistant to decomposition (Räsänen et al., 2006). In this study, akinetes were germinated from sediments that had been deposited c. 120 years before present, indicating that the viability of akinetes persists for long periods of time.

The identification of A. issatschenkoi in almost all sediment layers and to sediment depths of 340 mm (120 years old), using both ARISA and germination experiments was unexpected. This species has only recently been identified in New Zealand (Wood et al., 2007a) and has never been identified in the phytoplankton community of Lake Okaro (Fish, 1968; McColl, 1972; Flint, 1977; Dryden & Vincent, 1986; Bay of Plenty Regional Council, unpublished data). A possible reason for the abundance of A. issatschenkoi in the sediment core is the production of multiple akinetes by this species. In the germination experiments, this species produced multiple akinetes (up to seven akinetes per filament; Fig. 1b). It is plausible that conditions in Lake Okaro are never optimal for akinete germination; thus, populations may have only ever occurred at low concentrations. Various studies (e.g. van Dok & Hart, 1997; Baker & Bellifemine, 2000) have shown that germination depends on the occurrence of a relatively narrow range of conditions occurring in both the sediment and water column.

Wood et al. (2007a) detected the potent neurotoxin, anatoxin-a, in a culture of A. issatschenkoi isolated from Lake Hakanoa, Waikato, New Zealand. Anatoxin-a has been responsible for multiple animal deaths in New Zealand (Wood et al., 2007b) and worldwide (e.g. Gugger et al., 2005). Anatoxin-a was detected in the two samples collected from the germination experiments. Aphanizomenon issatschenkoi could therefore become a significant health risk and careful examination of phytoplankton samples from Lake Okaro should be undertaken to document both potential risk to lake users as well as any changes in its abundance as water quality restoration is attempted (Paul et al., 2008).

Three other species, Aphanothece sp., Geitlerinema sp. and Anabaena c.f. sp. Nova, were observed in the sediment core, and had not previously been documented in the lake. Aphanothece sp. and Geitlerinema sp. were only recorded in the upper layers of the core. These are small species and may have been overlooked in routine monitoring programmes where low (× 200) magnification was used. Anabaena c.f. sp. Nova was observed in multiple sediment layers in both germination and ARISA experiments. This species is very similar morphologically to A. spiroides. Given the abundance of A. spiroides in historic records (Dryden & Vincent, 1986) it seems probable that this species has been misidentified. The 16S rRNA gene sequence for the Lake Okaro strain showed a very high homology to A. sigmoidea (Rajaniemi et al., 2005) suggesting that the current taxonomic classification of this species may need revision.

There were multiple peaks observed in the ARISA profiles that could not be assigned to species observed in germination experiments. ITS information was not available for all species (e.g. Aphanothece sp., Geitlerinema sp.); therefore, it is likely that these species would account for some of the AFL. Additionally, it has been shown that interoperonic differences in spacer length occur within the genomes of microorganisms (Nagpal et al., 1998), thus a single species may contribute more than one peak to an ARISA profile. Previous studies (e.g. Gugger et al., 2002; Wood et al., 2008) indicate that species of the order Nostocales commonly have two types of ITS regions (i.e. two AFLs), whereas Chroococcales and Oscillatoriales have only one. Only one ITS length has been identified for some of the species (e.g. Anabaena c.f. sp. Nova). The yet to be determined second ITS may account for some of the unidentified AFL. Phylogenetically unrelated species can have identical ITS lengths, therefore ARISA may underestimate species diversity or possibly the wrong species may be assigned to an AFL. For example, we assigned Anabaena lemmermannii to the AFL of 471; however, this species was not present in the germination experiments. Some of the unassigned peaks could also be due to artefacts produced during DNA extraction and PCR (Taton et al., 2006).

Surface sediments

There has been debate on the importance of the role of benthic cyanobacteria in reinoculation of pelagic populations. Some studies have suggested germination of akinetes or recruitment of hormogonia or vegetative cells from the surface sediment plays a critical role in bloom initiation (e.g. Rother & Fay, 1977; Brunberg & Blomqvist, 2003; Kim et al., 2005). Conversely, Reynolds (1975), Karlsson-Elfgren et al. (2004) and Verspagen et al. (2004) found that recruitment of akinetes and vegetative cells from surface sediments had little influence on summer pelagic cyanobacterial populations.

The result from the Lake Okaro sediment core suggests that the importance of sediment recruitment varies between species. One of the surprising findings was the complete absence of Anabaena planktonica from the sediment surface layers. Anabaena planktonica was first detected in New Zealand in 2000 and has rapidly spread throughout the North Island (Wood et al., 2004). Dense populations (at times >80 000 cells mL−1) of A. planktonica have been recorded in Lake Okaro during the past four summers (Bay of Plenty Regional Council, unpublished data). The absence of this species in the surface sediments suggests that it is able to survive in a pelagic vegetative state throughout the winter. Other Anabaena species have been observed to overwinter in a pelagic vegetative state. Head et al. (1999) found A. flos-aquae filaments overwintered in the water column, and surmised that this pelagic population was the primary source of subsequent cyanobacterial growth. In the Okaro core, the highest cyanobacterial diversity was observed in the surface layers (0–40 mm), indicating that overwintering in either a vegetative state or as akinetes is an important adaptive strategy for many species, not including A. planktonica. Further research involving sediment traps and year-round phytoplankton monitoring is required to elucidate the importance of sediment overwintering and determine variables that trigger recruitment in Lake Okaro.


Sediment germination experiments and/or molecular techniques can be used to successfully monitor chronological changes in community structure in most akinete-producing species over long time scales. For Nostocales, the results of this study indicate that there has not been a dramatic change in cyanobacterial species composition in Lake Okaro in the past 100 years. The methods used in this study do not measure quantitative changes, but other studies (e.g. Flint, 1977; Dryden & Vincent, 1986) indicate an increase in planktonic cyanobacterial abundance in response to nutrient enrichment. The absence of A. planktonica in the sediment, demonstrates that overwintering on the sediment is not an important survival strategy for all cyanobacterial species. Analysis of the sediment core revealed the presence of previously unreported cyanobacterial species. Of particular concern is anatoxin-a-producing A. issatschenkoi. This species has recently become dominant in other New Zealand lakes and because of its potential threat to human health, its abundance should be closely monitored.


The authors thank Dennis Trolle (Waikato University) for field assistance, Andrew Selwood (Cawthron Institute) for help with anatoxin-a analysis and Bay of Plenty Regional Council for use of their data. This research was funded by the New Zealand Foundation for Research Science and Technology (UOWX0505) and a postdoctoral fellowship (CAWX0501) to S.A.W.