1. Introduction
In recent decades, there has been an apparent global increase in the range, intensity, and frequency of harmful algal blooms (HABs) linked to a variety of factors, including range expansions, increases in anthropogenic nutrients into coastal water bodies, and increased aquaculture [
1,
2,
3,
4,
5,
6].
Prorocentrum minimum is a planktonic marine dinoflagellate that forms HABs and is found commonly in temperate estuarine and coastal waters [
7].
P. minimum blooms are most common in eutrophic coastal waters of the northern hemisphere; however, they have also been reported in tropical and subtropical regions globally [
1,
7,
8,
9,
10]. Although few studies have been conducted on
P. minimum in Australia, it is known to occur in high abundances in some regions, with frequent blooms in the Hawkesbury River in New South Wales (NSW) [
11]. In line with the global increase of HABs,
P. minimum appears to have expanded its geographical range over the last 40 years [
1,
10,
12].
P. minimum usually blooms in warm brackish waters that are heavily impacted by excess nutrients, which has led to its presence being used as an indicator of eutrophication in water bodies in the northern hemisphere [
1,
11,
13].
The abundance and even dominance of
P. minimum in dynamic estuarine and coastal systems may be due to its broad salinity tolerance range of 5–17 PSU [
9,
14] and broad temperature tolerance range of 3–30 °C [
1,
15,
16].
P. minimum typically blooms in low-turbulence environments during periods of high irradiance levels [
1]; however, it has been demonstrated that the species can survive complete darkness for extended periods [
17], which may allow it to survive in ship ballast waters.
P. minimum is considered to be a mixotroph, able to supplement its nutrient intake due to feeding on smaller microbes, such as
Cryptomonas spp., in response to depleted nutrients in the water [
7,
11,
18]. Despite the ability to survive with low nutrients,
P. minimum preferentially grows in water bodies with high nutrient loadings, typical of eutrophic water bodies [
1,
9].
P. minimum growth has been found to be associated with high inorganic nitrogen (N) and phosphorus (P), strongly linked to anthropogenic sources, such as fertilisers [
7].
P. minimum blooms have been associated with several different marine biotoxins [
19,
20,
21,
22]; however, the identities of the compounds and their modes of toxicity are debated.
P. minimum blooms have shown toxic effects on shellfish, including mortality, poor development, and altered behaviours [
1,
23,
24,
25]. Recently, a
P. minimum bloom has been associated with the neurotoxin tetrodotoxin (TTX) [
19,
26,
27], possibly due to bacterial species associated with
P. minimum [
26,
28]. It has been suggested that
P. minimum toxicity is variable depending on the strain of the species and the environmental conditions under which it is grown [
1,
26]. Due to incidences of toxin accumulation in shellfish and the impacts on shellfish growth of
P. minimum toxins, it is an important HAB species to monitor in shellfish-harvesting regions [
1,
11,
23,
24,
26].
Until recently, light microscopy has been the only routine method available to identify and manually count HAB species [
29,
30,
31,
32]. However, this method is relatively time-consuming, requires a very high level of taxonomic expertise, and is not able to identify cryptic species, which may appear morphologically indistinguishable from one another despite toxicological differences. For these reasons, alternative methods of monitoring HABs have been developed. Molecular genetic techniques can provide rapid and sensitive HAB monitoring [
29,
30,
33,
34]. Two molecular genetic methods used are quantitative polymerase chain reaction (qPCR) and molecular barcoding. Molecular barcoding, also referred to as amplicon sequencing, is becoming invaluable in studying marine ecological assemblages, as it allows for uncultured cells in samples to be identified [
29,
35,
36]. However, due to the existence of variability in the copy numbers of genes among microalgal species, particularly in dinoflagellates [
37,
38,
39], the number of gene copies amplified may not reflect the relative abundance of species in the sample. There is also a bias introduced with the use of broad-range primers, which can lead to certain sequences being preferentially amplified, giving a skewed proportional abundance of target species [
29,
40]. For this reason, the quantification of dinoflagellate species using amplicon sequencing is uncertain and not accurate when compared with other methods, including qPCR and light microscopy [
29,
41]. The use of molecular barcoding, which provides an overview of the genetic composition of microbial communities, in conjunction with qPCR, may improve the quantitative assessment of the impact of HAB species in the context of, for example, seasonal changes in the wider microbial community.
The aim of the study was to develop and assess new molecular genetic approaches to investigate the dynamics, community, and environmental drivers of P. minimum in an Australian estuary. To do this, a local isolate of P. minimum from Australia was established, and qPCR approaches were designed and tested for the detection and quantification of P. minimum. In addition, 18S rRNA amplicon sequences from estuarine water samples were examined to compare the specificity, detection limits, and quantification accuracy of the methods. Environmental samples, including physico-chemical data, were collected monthly for 14 months from 2016 to 2017 from two sites in Botany Bay, an estuary in southeast Australia. Data of the entire microbial community, the abundance of P. minimum, and the corresponding physico-chemical variables were examined to assess the factors impacting the presence and abundance of P. minimum in an Australian estuary.
4. Discussion
Prorocentrum minimum is a marine dinoflagellate that commonly occurs throughout the world and can form HABs [
1,
7,
10]. HABs due to this species often occur in estuarine and coastal waters where aquaculture takes place, and in relation to that, death of shellfish has been reported [
1,
10].
P. minimum has been reported to show physiological flexibility with a global distribution across a range of conditions from temperate to subtropical [
1,
7,
67]. It has been reported to produce TTX, a harmful neurotoxin [
19,
68,
69,
70]. Due to the potential harmful impacts of
P. minimum on shellfish aquaculture, in this study, we aimed to develop new methods of investigating this species and apply them to environmental samples. In this study, a new culture of
P. minimum was successfully isolated from Berowra Creek, Hawkesbury River, Australia.
P. minimum has been linked to the production of TTX after a bloom in Vistonikos Bay, Greece, was positively correlated with TTX [
27]. TTX was not detected in our strain. Genetic variability among strains may influence the toxicity of
P. minimum, as well as the environmental conditions under which it is grown [
1,
24,
26]. Due to the reported variability in toxicity in this species, future studies will be required to evaluate the toxicity of strains of
P. minimum. As the alga is now successfully in culture with the Cawthron Institute Culture Collection, it allows future studies to look at more in depth toxin profiles, including how different environmental stressors and relationships with other known toxic algae or bacteria influence its toxicity.
qPCR assays have been developed over the past ~15 years for the detection and monitoring of HAB species [
51,
71,
72]. qPCR has advantages over traditional light microscopy methods in that it is sensitive and rapid and allows for possible future automation. In the development of qPCR assays for the detection and quantification of specific taxa in environmental DNA, the most important considerations are the specificity of the assay in that it amplifies only the species of interest, and the amplification efficiency of assays with an efficiency of less than 90% will not give quantitative results across its full detection range [
53,
73]. Assays with an efficiency greater than 110% are considered to show significant inhibition to PCR amplification [
74]. Amplification greater than 100% can be due to contamination in the sample, pipetting errors, inaccurate dilution series, and primer dimers [
29]. For this study, a previously published qPCR assay developed for
P. minimum was originally tested [
51], which targeted a fragment of the small subunit ribosomal (SSU/18S) RNA gene. However, it was found to amplify several other nontarget
Prorocentrum spp. and have a low efficiency (
Table 5, 70%). Therefore, new primer sets were designed to develop a new qPCR assay for
P. minimum with the aim of being specific, sensitive, and efficient. Twenty-two unique primer sets were designed and tested with variable results (
Table 5). Only one of the primer sets was found to be sufficiently specific and efficient and was used to examine environmental samples for the presence of
P. minimum (primer set 20,
Table 5). The newly designed assay was based on the ITS rDNA gene region, which is more variable and faster evolving than the SSU rDNA gene among dinoflagellate species [
75,
76]. The assay did have a low level of cross-reactivity with the most genetically similar species,
P. cf.
balticum. However,
P. cf.
balticum could be distinguished from
P. minimum due to a higher temperature on the melt curve profile. Several studies have used melt curve differences to discriminate similar species [
77,
78]. When analysed for efficiency, the new primer set showed E = 101% (
P. minimum DNA) and E = 99% (gBlock synthetic DNA) (
Table 5). The new assay amplifies a much shorter fragment than the previously published assay (71 bp compared with 325 bp), and this may account for its greater amplification efficiency [
79]. The qPCR assay developed for
P. minimum is more sensitive than most light microscopy counting methods, with a reliable detection limit of 13 cells L
−1 [
65,
66].
Molecular barcoding, or amplicon sequencing, which involves PCR amplification of environmental DNA (eDNA) and then sequencing of short (~600) [
80] “barcoding” gene regions using high-throughput sequencing (HTS), is another molecular genetic method that has begun to be used in phytoplankton research [
81,
82,
83]. Amplicon sequencing uses broad-range primers designed to amplify conserved regions across whole domains of life—in this case, eukaryotes [
56,
84]. A major problem with the use of amplicon sequencing as a tool for quantifying microbial eukaryotes is that the “barcoding” genes may be present in multiple copies that can be variable among microalgal populations and species, meaning that sequenced gene amplicons may not reflect the true abundance of a species in the sample. qPCR is not immune to this problem; however, the impact is minimised by designing primers that amplify gene regions only present in a specific species and using a standard curve with known amounts of target.
However, in this study, the sequencing of amplicons of eukaryotic V4 regions of SSU rDNA from samples from Botany Bay did not show a significantly different quantity of
P. minimum compared with the quantification based on qPCR (
Figure 3 and
Figure 4 and
Table 6). In addition, the results of this method have shown a previously unknown level of phytoplankton diversity in Botany Bay, detecting over 600 eukaryotic microbial ASVs between the two sites in Botany Bay. Previously, phytoplankton identification using light microscopy had led to the detection of only ~100 species or fewer in 10 years of phytoplankton monitoring at Botany Bay [
85,
86]. In this study, only 43% of all phytoplankton ASVs were able to be classified to species level using the 18S V4 primer set and the PR2 database [
59]. Further development of reference databases of 18S V4 sequences from reference “voucher” specimen taxa curated by taxonomists is an important factor in the future of HTS to enable a more complete and accurate picture of microbiome species composition [
87,
88]. Another possible approach that may lead to a more specific identification of taxa is the use of other primer pairs that amplify other amplicon regions, such as the LSU rDNA region in dinoflagellates, the SSU (16S) plastid genes, CO1, cytochrome
b, or other mitochondrial gene regions [
82,
83,
89,
90]. In previous studies, it was found that some groups of taxa, such as dinoflagellates, can be identified more readily using LSU rDNA regions, rather than SSU rDNA [
83].
The collection and preservation of water samples for the identification and manual counting of cells with light microscopy has been the “gold standard” method used to study phytoplankton abundances [
91,
92,
93]. The accuracy of light microscope-based microalgal enumeration is highly variable depending on the particular technique chosen, the counting effort, and the taxonomic expertise of the technician [
81]. Compared with light microscopy enumeration, amplicon sequencing has been shown to be extremely sensitive and has the capacity to identify all phytoplankton species in a sample without requiring any taxonomic expertise. qPCR is an optimal technique for the enumeration of a particular target species, as the limit of detection is low, and the accuracy of the method is independent of the effort or taxonomic skills of the operator. It is relatively inexpensive, rapid, sensitive, and specific and, therefore, is highly suited to adoption for ongoing monitoring programs. qPCR can also be completed in situ at the time of sampling to get rapid results and can be used by trained shellfish producers to get results of HABs on-site.
The light microscopy counting method utilised in this study had a larger-than-average error rate, a high detection threshold of
P. minimum (500 cells L
−1 compared with 13 cells L
−1 with qPCR), and comparatively fewer data points when compared with the molecular methods. Adoption of other light microscope counting methods, like the Utermöhl counting chamber [
94], and the use of a DNA-based stain (i.e., a fluorescence in situ hybridisation (FISH) probe [
95]) may have led to more accurate assessments of the abundance of
P. minimum and the detection of cryptic species. For research into HAB ecology, a combination of the use of amplicon sequencing, to first determine the diversity of phytoplankton, particularly cryptic and small species, and then qPCR, to quantify the exact cell abundance, would appear to give optimal information for ecological inference and understanding of co-occurrence patterns. This two-step molecular pathway appears to be the most appropriate method for future development [
29,
35,
71,
92,
96].
Botany Bay is an estuary in southeast Australia that is extensively modified [
43], containing an international shipping port (Port Botany), an oil fuelling station, recreational beaches, industrial estates, and urban developments [
97]. The bay is a highly populated area and is impacted by freshwater flows from the Georges and Cooks Rivers, both also extensively modified and surrounded with urban developments [
97]. Despite the modification, Botany Bay is also home to a Ramsar wetland and one remaining oyster farm, both at Towra Point [
98]. Thus, it is an important site for ongoing monitoring of HAB species, as they can impact not only the shellfish production but also the quality of the water for recreational and industrial purposes. Botany Bay and the Georges River have both previously experienced HABs, including
Noctiluca scintillans,
Alexandrium pacificum, other
Alexandrium spp., and
Heterocapsa spp. [
99].
Two sites in Botany Bay were sampled from April 2016 to June 2017: Bare Island and Towra Point (
Figure 1). Due to the extensively modified nature of the bay and its surroundings, and the nutrient input that can occur in relation to this land runoff, it was expected that
P. minimum may be abundant at these sites. It was also expected that
P. minimum may be particularly high in abundance at Towra Point, which is impacted by freshwater flows, as this species has been shown to flourish in low salinities with high nutrient freshwater inputs [
1,
70].
P. minimum was found to be in low abundance for most of the sampling period at both sites, detected at ~30 cells L
−1 at both sites for most of the year of sampling (
Figure 3 and
Figure 4). There was one peak in the abundance of
P. minimum (8000–14,000 cells L
−1) at both sites, on 7 June 2016 (
Figure 3 and
Figure 4). This could still be considered a low value for
P. minimum, which has been detected at “bloom” levels upwards of 10 million cells
−L [
11]. The low presence of
P. minimum is an important current baseline for monitoring the health of Botany Bay and other southeast Australian estuaries.
The abundance of
P. minimum was found to have a significant positive relationship with total CO
2, contrary to a previous finding that found that increased CO
2 had no relationship with the abundance of
P. minimum [
100].
P. minimum was also found to have a weak but significant negative relationship with salinity, which supports previous findings that
P. minimum grows preferentially in decreasing salinities [
1,
70,
101].
P. minimum was not found to have a significant relationship with any other environmental variables; however, it is likely that there may be a time lag between an environmental change and increase or decrease in
P. minimum [
102,
103]. Incorporating a measure of exposure of
P. minimum to environmental variables would require a higher temporal sampling frequency than what was undertaken in the present study. However, the correlations we found (+ve CO2 and −ve salinity) between
P. minimum and the environmental variables measured are hypothesis forming and should be further investigated in Australian waters.
An analysis of phytoplankton species that significantly co-occurred with
P. minimum in Botany Bay is useful, as in the past, toxicity attributed to blooms of
P. minimum may have been also associated with the presence of
Dinophysis spp., which are the main causative agents of diarrhetic shellfish poisoning (DSP), even when present in low abundances, such as ~100 cells L
−1 [
1,
21,
22,
104]. Due to the potential uncertainties of amplicon sequencing-based estimates of the absolute abundance of cells in a sample, the data were analysed as a presence/absence matrix across all sampling dates [
29,
37].
P. minimum at Towra Point was found to significantly co-occur with 12 other phytoplankton species and at Bare Island with 4 other phytoplankton species. Of all the co-occurring species, only 1 is a toxin-producing species,
Alexandrium pacificum. A. pacificum is an important HAB species due to the severity of bloom impacts in Australia, New Zealand, Korea, Japan, and other countries [
72,
105,
106,
107].
A. pacificum produces paralytic shellfish toxins (PSTs).
P. minimum blooms have previously been associated with symptoms characteristic of PSTs [
68,
69]; however, there is still a possibility that it can produce other toxins not yet classified [
19,
20].