Since the 1960s, use has been made, primarily of plant tissues, to act as baits to attract oomycete spores out of bodies of water via chemo-attraction and chemotaxis. Many different nuances of the baiting approach exist, and the earliest of these techniques used for Phytophthora species have been comprehensively catalogued by Ribeiro (1978). As they are low in cost, straightforward, and highly adaptable, baiting techniques are still widely developed and successfully used for the capture of oomycetes in water. The baits selected for an assay strongly influence the resulting catch: for example, autoclaved insect parts are likely to attract Saprolegniales and some Pythium species (Sarowar et al., 2013), whereas fresh, whole rhododendron leaves work well for Phytophthora species in temperate streams and hardy nursery-stock irrigation systems (Themann and Werres, 1998; Themann et al., 2002). Green et al. (2020) effectively deployed a mix of different fresh leaf species as baits in a study investigating the range of Phytophthora species present in public parks, whilst in comparative trials in Australia, primarily looking at soils, a multiple bait system used by the Centre for Phytophthora Science and Management was the most effective baiting procedure (Burgess et al., 2021). The age, quality, and physiological state of the plant tissues being used as baits are hugely influential on the results obtained (Themann et al., 2002; Hüberli et al., 2013; Werres et al., 2014), and inclusion of dead tissues within a bait mix can increase the number of species captured (Wielgoss et al., 2009; Aram and Rizzo, 2018; Sarker et al., 2023a). When baiting is deployed in comparative experiments over time, variability can be reduced by using more controllable tissues such as seedlings or seedling parts (Banihashemi and Mitchell, 1975; Erwin and Ribeiro, 1996; Pettitt et al., 1998). Perhaps the ultimate in reproducible baits use just chemical attractants and do not have to rely on maintaining consistency of plant material. This principle was successfully deployed using phenols, alcohols, and amino acids in the development of chemo-attractive dipsticks to detect viable Phytophthora cinnamomi, and later P. nicotianae, zoospores (Cahill and Hardham, 1994a and Cahill and Hardham, 1994b; Gautam et al., 1999). Nevertheless, plant tissue-based baits continue to predominate and be widely and successfully deployed for the capture of oomycetes from water.

Irrespective of the materials used for baits, there are two main approaches to baiting water; in situ and ex situ (Werres et al., 2014). In situ baiting is probably the most frequently deployed and involves placing baits directly within the bodies of water being monitored. Examples include floating bait leaves (Huai et al., 2013; Matsiakh et al., 2016), or fruits (Rodríguez et al., 2021) on pond, stream, or river surfaces for 3-10 days before collection and isolating from lesions. Ex situ baiting, on the other hand, involves the placement of baits in collected and contained samples away from the sampling location, ideally under standardized incubation conditions to encourage zoospore release/re-release and infectivity (Sarker et al., 2021). Redekar et al. (2020) used ex situ baiting as a method for viable propagule capture as part of their sampling routine for metabarcoding analysis of irrigation water at a horticultural nursery in southern California, USA. They placed fresh leaves of Rhododendron catawbiense ‘Grandiflorum’ in 1L water samples and extracted DNA for analysis from lesions formed following incubation for 3 days at 18-22°C. Looking for lesions on harvested baits moves from using baits for capture to the first stages of detection especially of potentially phytopathogenic oomycete species. Only selecting bait lesions runs the risk of bias in favor of detecting only species capable of rapid infection and symptom induction in the bait species used, and when assessed, asymptomatic baits are often found to have captured a wider range of oomycete species (Sarker et al., 2023b). In situ and ex situ baiting of bodies of water differ both temporally and volumetrically. They are temporally different in that ex situ baits test samples taken from locations at specific times that consequentially represent very short time periods (‘freeze frames’). On the other hand, in situ baits might be considered to be sampling their locations for as long as they remain in situ, although their attractiveness and capacity for inoculum capture will likely vary over time depending on their physical condition. Being limited to restricted to pre-set sample volumes, ex situ baiting differs volumetrically to in situ where baits are likely to be capturing inoculum from unconfined and, depending on the movement of the sampled water body, more variable and potentially of much larger volume. In soils, recent work has shown that ex situ baiting of many small samples as opposed to the generally-preferred large pooled samples can increase the diversity of oomycete species captured (Sarker et al., 2023a; Sarker et al., 2023b). This result is thought to be due to the dilution effect of smaller samples reducing the probability of slower-growing or less aggressive species being out competed in bait infection. A similar effect might be observed in ex situ baiting of water samples where it is often assumed that the bulk of inoculum present is motile and infective already and not subject to the limitations of sporulation and release associated with soil. Nevertheless, variation in the aggressiveness of this water-borne inoculum is likely. Additionally, possible bait-induced release of zoospores from caducous sporangia and viable cysts and even the activation of mycelium or oospores in debris, make similar studies with ex situ water baiting worthy of investigation.

In soils, the rates of sporangium formation and zoospore release of different species of Phytophthora have been demonstrated to affect the efficacy of their detection by baits (Sarker et al., 2021). This process might be reflected in water, where inoculum release from infected plants is often in short-term surges (Hallett and Dick, 1981; Pettitt et al., 2015). These inoculum surges are often initiated by changes in the environment such as increased irrigation frequency or by rainfall (Ristaino, 1991; Café-Filho et al., 1995), but can also result from subtle changes in cultural practice, for example sudden reductions in the root zone temperature in hydroponics crops (Kennedy and Pegg, 1990). There are large seasonal variations in the numbers of oomycete propagules seen in irrigation water in the UK (Pettitt et al., 2015), some likely driven by seasonal changes in water temperature (Pettitt and Skjøth, 2016). A fairly consistent double peak in general oomycete CFU counts is seen in clinic samples from UK nurseries, with larger counts in late Spring and in late Summer-Early Autumn and a decline in mid-Summer (Pettitt et al., 2015). This result is in contrast to total filamentous fungal counts and the numbers of Fusarium CFU, which generally reach a single peak in August/September. A similar distribution with distinct peaks of detected CFU in Spring and Autumn was observed in citrus orchard soils (Cacciola and Magnano di San Lio, 2008), and interestingly this represents the annual progress curves of two different species of Phytophthora (P. citrophthora and P. nicotianae).

The main alternative to using baits for capture, is to extract oomycete propagules from water by filtration. This approach was initially developed as a precursor to plating viable propagules onto semi-selective agar plates for colony-forming unit (CFU) counts (Ali-Shtayeh et al., 1991; Hong et al., 2002; Pettitt et al., 2002). Filtration to capture dispersed inoculum in water is now also used, often in parallel with baiting, to capture oomycete biomass ready for detection and identification by DNA extraction and nucleotide assays, especially with the increasing use of metabarcoding techniques (Scibetta et al., 2012; Català et al., 2015; Green et al., 2021). To capture viable propagules for colony-plating, membrane filters are used, most frequently of pore size 3 or 5 µm and 47 mm diameter. Once filtration has been completed, membranes can be plated directly onto selective agar media (Davidson et al., 2005; Calvo-Bado et al., 2006; Davidson et al., 2008). Alternatively, the inoculum caught on the filter membrane surface can be released into a small volume of a resuspension medium consisting of 0.09% agar solution (Ali-Shtayeh et al., 1991), ideally containing selective antibiotics at the same concentrations as the plating medium (Pettitt et al., 2002; Büttner et al., 2014). A membrane fabric that readily and reliably releases caught viable propagules is desirable for colony-plating procedures. Hong et al. (2002) investigated a range of membranes and found the most efficient recoveries in terms of filtration speed and colony recoveries were on 5 µm polyvinylidene fluoride (‘Durapore 5’) membranes, although the only cellulose nitrate filters they investigated were of 0.45 µm pore size, which was the size and type originally used by Ali-Shtayeh et al. (1991). This pore size was originally used at Horticulture Research International Efford (UK) but was found to be too fine and too readily blocked for efficient extraction of oomycetes from irrigation water samples. An increase in pore size to 5 µm cellulose nitrate (Whatman International Ltd, Maidstone, UK) gave faster filtration and consistently high colony recoveries of Phytophthora and especially Pythium zoospores (Pettitt et al., 2002). The pore size was later reduced to 3 µm following the findings of Hwang et al. (2008), although later comparisons in 2012 between 3 and 5 µm cellulose nitrate membranes and 3 µm polycarbonate (Cyclopore™ track-etched, Whatman International Ltd, Maidstone, UK) membranes showed no significant differences in colony recoveries from irrigation reservoir samples (Pettitt et al., 2015).

When filtration has been deployed for the capture stage for genomic studies of oomycetes in water, effective use has been made of mixed cellulose ester filters of pore size 1.2 µm to capture target biomass (Scibetta et al., 2012), although again this size has proved too fine for efficient filtration of field samples, and in more recent studies an increased pore size of 5 µm has been used (Mora-Sala et al., 2022). Interestingly, these workers refined their oomycete extractions by using a baiting technique, placing filter membrane halves into longitudinal cuts in apple fruits, a widely used bait, especially for baiting Pythium and Phytophthora species from soil samples and decaying, diseased plant tissues (Campbell, 1949). Cuts containing membrane halves were then sealed with parafilm, and molecular identifications of cultures isolated from the resulting lesions were carried out.

Membrane filtration of water samples taken to the laboratory is normally carried out using bottle-top filter funnels such as the reusable, readily-cleaned, and autoclavable Nalgene™ Polysulfone units (Pettitt, 2016), and passing water through the filter using suction from a vacuum pump. In the field, in the absence of an available electricity supply, this method can still be applied using a hand-held suction pump (Matsiakh et al., 2016; Pettitt et al., 2018). Hand-operated suction pumps are compact and work well, but are tiring to the forearms of the operator after a short time. An alternative approach developed by Scibetta et al. (2012) is to push the water through the filter mounted in an in-line polypropylene filter cartridge using an adapted knapsack sprayer (Cooper Pegler CP3). This method is effective, especially for filtering larger volumes of water (2-5 L), although cleaning the apparatus between samples, particularly when collecting samples for eDNA analysis, is time-consuming and requires a comparatively large amount of oomycete-free water. A refinement of this basic concept was developed during the ‘Phytothreats’ project (Green et al., 2021), using a cyclist’s track pump to apply air pressure to a simple pressure vessel that then drives one or more water sample(s) from their connected collection bottles out through up to 3 parallel filter cartridges. This process greatly reduces potential cross-contamination and the of cleaning needed between samples is reduced to just the filter cartridges and their supply manifold, which can be flushed through with cleaning agent and then rinsed with small volumes of sample water before new filter membranes are mounted.

There can be discrepancies in the numbers and abundance of oomycete species captured by filtration and baiting methods (Sarker et al., 2023a), especially with ex situ baiting which in comparisons, tends to capture fewer species and less biomass than filtration (Pettitt et al., 2002; Redekar et al., 2020). Recent studies looking at assay conditions and sample sizes for in situ baiting of soils are closing this gap (Sarker et al., 2023a; Sarker et al., 2023b), but further studies are required to fully understand the differences seen in water tests.

Culture-based detection and identification techniques either consist of plating of captured propagules from baits or directly from membrane filters onto semi-selective antibiotic-amended agar (Tsao, 1970), or of direct observation of signature symptoms and/or sporulation on baits. Depending on the numbers of propagules present in a sample and the condition of the water, culture-based techniques can be highly sensitive with membrane filtration-colony plating potentially resulting in one colony per viable spore (Pettitt et al., 2002). Rapid identifications to genus, either directly from observations of colony morphology on agar plates (O’Brien et al., 2009) or from sporulation on infected baits, are possible (Werres et al., 2014). Culture- plating methods tend to lack specificity and this can be increased by altering the components of the of the mix of fungicides and antibiotics used in the semi-selective medium. Additions of hymexazole to media improves selectivity for many Phytophthora spp. (Tsao, 1983; Erwin and Ribeiro, 1996), but this selectivity often comes at the price of reduced sensitivity as a consequence of inhibitory effects of the antimicrobial mix (Tsao, 1970; Sarker et al., 2020). Nevertheless, isolations and/or further culturing to encourage sporulation and properly observe colony morphology are, more often than not, required to achieve a diagnosis using identification keys, a process requiring patience, practice and skill. Nevertheless, extractions of detected oomycetes can be made from colony picks or from recovered baits for examination and further identification by either immunodiagnostic or nucleotide-based assays.

Baiting techniques can be used as a capture and initial detection step prior to testing by enzyme-linked immunosorbent assays (ELISA) or lateral flow devices (LFDs) (Wedgwood, 2014), whilst immunodiagnostic dipstick assays can carry out the baiting capture step without the need for live plant tissues and on collection can progress straight to a diagnostic staining process (Cahill and Hardham, 1994b; Wilson et al., 2000). Immunodiagnostic assay, both ELISAs and LFDs, have also been used directly on extracts of inoculum captured on membrane filters with some success (Wedgwood, 2014). Trapped propagules can also be subjected to direct immunostaining procedures in situ on a membrane filter and observed by stereo microscope or a hand lens using the zoospore trapping immunoassay (ZTI), a procedure whereby captured, viable zoopsores and cysts are encouraged to germinate in situ prior to immuno-staining (Wakeham et al., 1997; Pettitt et al., 2002). The range of immunodiagnostic procedures of potential use in the testing and monitoring of irrigation water for oomycetes has been more fully reviewed by O’Brien et al. (2009) and more recently by Wakeham and Pettitt (2017).

In recent years, molecular nucleotide assays essentially based around polymerase chain reaction (PCR) have become the preferred method for laboratory-based detection and identification of oomycetes (Schena et al., 2008; Wakeham and Pettitt, 2017: La Spada et al., 2022). For detection of oomycetes in water, PCR tests are used in conjunction with either a filtration or baiting capture technique, and baits have been widely use as a ‘biological amplification’ stage (e.g., Shrestha et al., 2013; Sena et al., 2018: Kunadiya et al., 2019). Over the last 20 years, basic PCR, quantitative PCR and isothermal amplification techniques have been improved, tested, and refined for oomycete detection and identification. Much of this work has focused on Phytophthora because of both the importance of this genus commercially, and the devastating impact of species like P. cinnamomi (Carter, 2004; Shearer et al., 2004; Scott et al., 2013; Hardham and Blackman, 2017) and P. ramorum (Rizzo et al., 2005; Grünwald et al., 2008) on the environment. There are some excellent reviews of these successful, still developing methodologies (Lévesque and DeCock, 2004; Cooke et al., 2007; O’Brien et al., 2009; Robideau et al., 2011; Martin et al., 2012; Schroeder et al., 2013; Magray et al., 2019).

Using DNA sequence data is a precise method of achieving identifications, but there is sometimes a concern about accuracy in that there are many misidentified sequences in public sequence databases, for example for certain species of Phytophthora (Abad et al., 2023). New identifications/diagnoses should be carried out with caution and, where possible, be based on both molecular and morphological characters with close alignment with ex-type sequence data before assigning to species. A good example of this practice is illustrated in the protocol described by Mora-Sala et al. (2022), where new isolates were assigned to a species when the identity was above the 99% cut-off in respect to the ex-type isolates recommended by the IDphy online resource (see Abad et al., 2023).

The large amounts of oomycete genomic data that have been generated, the protocols and PCR primers used to access and collate it, and the identification of so many new taxa (e.g. over 100 new Phytophthora species alone in the last twenty or so years) can be bewildering to the non-specialist. Nevertheless, robust oomycete phylogenies are now building on the early molecular studies of Cooke et al. (2000) and Lévesque and DeCock (2004). These phylogenies consider both morphological and physiological traits as well as nucleotide sequence data. In addition to the originally-used Internal Transcribed Spacer (ITS) region sequence data from other DNA regions including mitochondrially encoded cytochrome oxidase, cox I and II (Martin and Tooley, 2003; Robideau et al., 2011) and β-tubulin (Villa et al., 2006) are now used. Rigorous databases for this information have also been established to guide accurate species identifications for Phytophthora at least (e.g. The Phytophthora Database: http://www.phytophthoradb.org, IDphy: https://idtools.org/phytophthora/).

More recently, increasingly widespread use of metabarcoding technologies has greatly broadened the scope of oomycete community studies and has been successful in a number of studies looking at either oomycete (Redekar et al., 2019; Redekar et al., 2020; Foster et al., 2022) or primarily Phytophthora species (Català et al., 2015; Redondo et al., 2018; Riddell et al., 2020; Green et al., 2021) in water. The wider metagenomic approach also allows broader communities to be considered, meaning that oomycetes in water can be assessed within a broader context (Marčiulynas et al., 2020), an area that needs further exploration. The different ‘next generation’ sequencing (NGS) technologies and the interpretation software used in metabarcoding techniques are beyond the scope of this review. Each method has its own rates of error, and balancing the use of these together with determining the most effective sequence information to read for accurate species identifications is an ongoing, rigorous process of analysis and critique (Riddell et al., 2019), and as more studies are reported, comparisons and improvements can be shared (Burgess et al., 2022). The NGS metabarcoding approach allows the spectrum of species present in samples to be determined and therefore the possible interrelationships to be explored in a previously-unfeasible context. As broad sweeps are possible, samples can be screened for both familiar endemic and novel species all at once. This capacity makes these methods powerful tools for identifying, monitoring, and tracing new and potentially invasive disease threats. Indeed, a large proportion of oomycete studies testing water samples reported so far have, rightly, been primarily focused on biosecurity.

For the first category of monitoring, probably the best approach to assessing the efficacy of water treatments in situ is the use of resuspension-colony plating and counting CFU (Büttner et al., 2014), as there is a need for sensitivity but not for a high degree of specificity. Treated and untreated water samples could also be sensitively monitored using ZTI or immuno-dipstick assays using more generic anti-oomycete antibodies than those used by Cahill and Hardham (1994a) or Gautam et al. (1999), (Pettitt et al., 2002).

The second category of monitoring presents a more complex situation regarding the need for an initial diagnosis for in situ monitoring programs, which would pragmatically still involve a combination of cultural and molecular (metabarcoding) approaches to cast a broad net and minimize presumptive bias (La Spada et al., 2022; Mora-Sala et al., 2022; Sarker et al., 2023a). Depending on the complexity of the diagnosis, it may then be possible to proceed to quantification of inoculum in relation to the specific problem. Again, CFU counts can be used here if colony morphologies are distinctive enough, and ZTI or immuno-dipstick approaches may be workable, depending on the specificity of available antibodies. In many instances in this category of monitoring, it may not be essential to continuously monitor viability. In this case, depending on the availability of suitable primers, qPCR or even multiplex qPCR (Schena et al., 2006) would provide excellent, specific determinations of inoculum concentrations in samples. The assumption here is that this type of monitoring would require large numbers of samples, and currently the costs of using a metagenomics approach for this may prove cost-prohibitive. However, this situation should change as costs will likely reduce with the pace of developments.

The third category of monitoring is generally concerned with sites like horticultural nurseries or gardens that can have complex irrigation rigs (or maybe just gutters for collecting rainwater) that need to be routinely examined for oomycete and other potential pathogens. This is a situation where metabarcoding, combined with either baiting or filtration-colony plating, would provide excellent information, but at the moderate to high intensity of sampling needed, again, might not be economic. Currently, this kind of monitoring is carried out using the same kind of culture-based methodologies as the first category of monitoring, often with the inclusion of generic culture media such as basic potato dextrose agar in addition to oomycete-selective media to crudely assess the ‘background’ microbiota as well as oomycetes (Büttner et al., 2014).

The fourth category of monitoring was, for a long time, only feasible using baiting and culturing methods (Erwin and Ribeiro, 1996). These were later supplemented and greatly improved by augmenting morphologically-based identifications with PCR and sequencing (Ghimire et al., 2009; Ghimire et al., 2011; Riolo et al., 2020). This methodology was further adapted for the monitoring of Phytophthora species in water by the extraction and nested PCR procedure of Scibetta et al. (2012) and has subsequently been revolutionized by the use metabarcoding and metagenomics procedures which offer high levels of accuracy, sensitivity and some measure of abundance or relative abundance of detected taxa (Green et al., 2020; Green et al., 2021).