An exploration of nematode assemblages in the hydroponic media beds of a commercial aquaponics system

Summary – Nematodes perform important ecosystem services in soils and are well studied in traditional farming, but their role in newer soil-less food production systems is unstudied. We determined the nematode assemblages from substrates within the hydroponic media beds of a commercial aquaponics system. Three substrates were studied: coir, compost and lightweight expanded clay aggregate (LECA), with the nematode assemblages ascertained via traditional morphological analysis. All substrates were suitable for nematodes but abundance was lower compared with traditional compost. Compost from within the hydroponic media beds had nematode assemblages with similar maturity indices (MI) to compost external to the system, and both composts housed nematodes with a high proportion of bacterivores, a high degree of enrichment and no structure. LECA that were regularly disturbed and had little organic build up resembled compost – highly enriched and lacking structure. LECA that had been left undisturbed for longer and allowed to accumulate more organic matter had highly structured assemblages and showed less enrichment than the other substrates. Nematode assemblages within the coir showed very little enrichment but had higher MI and were more structured than compost. Problematic plant-parasitic nematodes were not found in any of the samples. We present the ﬁrst glimpse into the role of nematodes in aquaponics systems and encourage future studies to explore how nematodes in these systems affect yield and plant health.

The biodiversity of soil ecosystems is integral to its functional roles in decomposition, nutrient cycling, maintenance of soil stability, primary production, and the prevention of plant disease (Barrios, 2007;Van Der Heijden et al., 2008;Ferris & Bongers, 2009;Jiang et al., 2018). The abundance, composition and diversity of the soil biota determine the rates at which these ecological services can take place (Barrios, 2007;Yang et al., 2018). It is thus unsurprising that the soil biota would be used as an indicator of ecological soil health.
Nematodes once renowned for their parasitism of both plants and animals are now also widely recognised for both their ecosystem services in soils and their suitability as indicators of soil health (Ferris et al., , 2001Yeates et al., 2009). Ecosystem services provided by nematodes include the distribution of bacteria and fungi throughout the soil, decomposition and nutrient cycling, nitrogen mineralisation and the maintenance of soil food webs (De Ruiter et al., 1993;Jiang et al., 2017Jiang et al., , 2018. In fact, bacterivorous nematodes can be responsible for up to 40% of nitrogen mineralisation by ingesting bacteria and making this available to surrounding plants via excretion (De Ruiter et al., 1993;Ferris et al., 1998). Nematode assemblages with generalist and specialist predatory nematodes can also have a suppressive effect on plant-parasitic nematodes (Steel & Ferris, 2016). Assemblages with a greater abundance of these predatory nematodes are associated with undisturbed soils and have been shown to have more suppressive effects than their more disturbed agricultural counterparts (Steel & Ferris, 2016).
Nematodes are numerous, diverse, and sensitive to disturbances and changes within the soil ecosystem Ferris et al., 2001;Vervoort et al., 2012).
The composition of nematode assemblages can be used to predict disturbances to the system, nutrient availability, structure of the soil ecosystem and whether bacteria or fungi are the predominant decomposition pathway (Ferris et al., 2001;Ferris, 2010).
The indicative potential of these nematode assemblages has led to their use in a multitude of agricultural studies, from the effects of tillage to the benefits of organic farming (Treonis et al., 2010;Steel et al., 2012;Herren et al., 2020;Song et al., 2020). Nematode studies make up an important part of our understanding of conventional soil-based farming, but a swathe of new food productions systems such as hydroponics, aeroponics and aquaponics are emerging that have varying levels of (or no) reliance on soils (Khan et al., 2020). These food production systems grow crops in a variety of substrates but there have been no studies of their resident nematode assemblages.
Aquaponics is an emerging industry (König et al., 2018) combining aquaculture with hydroponics. Water is cycled around the system in a single loop or two decoupled loops (Goddek et al., 2019). Nutrient-rich water from the fish tanks is used as a substitute for synthetic fertiliser for edible plant production (Rakocy et al., 2006). Plant growth and mineralisation of nutrients by the microbiome in the hydroponic beds allows for the recycling of the aquaculture system water. Aquaponics as a technology has the potential for sustainable food production, with reduced environmental impacts of fish and edible plant production when compared to conventional recirculating aquaculture and hydroponics, respectively (Greenfeld et al., 2022).
The typical substrate in an aquaponics system is lightweight expanded clay aggregate (LECA) as it is sterile, porous, chemically inert and does not influence pH. Nitrifying bacteria colonise the aggregate and serve to convert fish waste into plant fertiliser (Tyson et al., 2004). Consequently, the microbial diversity has been investigated within the components of aquaponics systems (Schmautz et al., 2017;Eck et al., 2019), but despite the involvement of nematodes in nitrogen mineralisation, their acknowledged importance in traditional farming systems and the siting of many aquaponics systems in outdoor locations, nematode assemblages have not been considered.
Typically, containerised seedlings and their container are transplanted to LECA media beds for on-growing. Containers have either drainage holes or are made up of compressed growing media that breaks down so over time the root system of established plants is supported by the LECA. The LECA can be allowed to build up with a soil-like detritus or alternatively is cleaned and replaced regularly. The abundance and diversity of nematodes that can reside on the aggregate, in the surrounding detritus, within a coir plug or be maintained in soil or compost added to the system, should be of key interest to any grower trying to maximise yields and minimise plant disease.
In this exploratory study we investigate the nematode assemblages that can be found within the hydroponic media beds of a commercial aquaponics farm in the southwest of England. We identified nematodes with traditional morphological methods to determine the assemblages present. The aims of this study were: i) to determine the abundance, feeding types and heterogeneity of the nematode assemblages across the hydroponic media beds sampled; ii) to identify empirically the features of the media beds that may affect the nematode assemblages; iii) to compare nematode assemblages in compost to the same compost within the media beds; and iv) to provide evidence-based suggestions for further research into the role of nematode assemblages in enhancing plant growth in aquaponics systems.

EXPERIMENTAL SYSTEM
Sampling was performed at the commercial aquaponics facility Bioaqua Farm, Blackford, Somerset, UK, on 28 August 2020. The facility is a small-scale commercial integrated producer of rainbow trout (Oncorhynchus mykiss) and vegetables in an outdoor aquaponics system. Water is circulated in a closed system from the ponds to a settling tank and filters then to the hydroponic component, before returning to the ponds via a sump. The hydroponic compartment is situated in a polytunnel with all crops produced under a natural photoperiod with no supplemental lighting. Four ebb and flow media beds filled with inert lightweight expanded clay aggregate (LECA) receive equal water inputs from the system immediately after the biofilter from inputs distributed evenly across the beds (i.e., equal nutrient profiles).
Coconut coir plugs were used to grow seedlings before on-growing in the hydroponic media beds. Two methods of transplanting are used on site. Coir plugs containing seedlings are either transplanted directly into the hydroponic beds just below the surface of the LECA or into plant pots filled with compost. The compost-filled pots are embedded into the media bed just below the surface 496 Nematology of the LECA and have drainage holes to allow water flow between the media bed and pot. The facility produces a mix of crops dependent on demand and the season. Some beds are used for annual, seasonal crops grown in monoculture. These beds are harvested once every 3-4 months and replanted. Other beds contain a mix of annuals and perennials grown in a polyculture. These beds are partially harvested each growing season, with some crops left in situ for multiple years and continually harvested during the growing season.

SUBSTRATE SAMPLING
Samples were taken from monoculture and polyculture hydroponic media beds and included three substrates: coconut coir plugs, compost and lightweight expanded clay aggregate (LECA) ( Table 1). The top 8-10 cm of substrate was discarded to avoid sampling the dry zone of the media bed. Substrate samples were collected using a 3 cm diam. corer, collecting 10-cm-deep samples from a single location and mixing to obtain a minimum of 30 cm 3 of substrate. All mixed samples consisted of a single substrate plus any organic build-up or root system present.
All substrates present in a bed were sampled and samples were taken both from within the root systems of the planted crops present (all three substrates) and from nonplanted spaces between individual plants (LECA only). A minimum of 100 cm 3 of material was collected for each substrate from composite samples from a minimum of three locations in a random sampling pattern.
The monoculture bed consisted of rocket (Eruca vesicaria) transplanted in coconut coir plugs 3-4 months before the sampling date. The coconut coir plugs were partially disintegrated but not decomposed upon sampling. Two samples were taken from the monoculture bed: planted coconut coir plug material from within the root system of the rocket plants (containing root material), and LECA from the non-planted spaces between the plants (containing organic build up but no root material).
The polyculture bed consisted of basil (Ocimum basilicum) transplanted in compost pots 3-4 months before the sampling date and sorrel (Rumex acetosa) transplanted in coconut coir plugs 5-6 years before the sampling date. The coconut coir plugs used to grow the sorrel had decomposed in the bed and were not evident in our sampling. Three samples were taken from the polyculture bed: planted compost from within the root system of the basil plants in pots (containing root material), planted LECA from within the root system of the sorrel plants (containing root material and a high build-up of organic matter), and LECA from the non-planted spaces between the plants (containing a medium level of organic buildup but no root material). The basil pots were filled with Melcourt brand all-purpose peat-free compost. Remaining bagged compost stored on site was sampled on the same date as the basil pots to provide a comparison between compost internal and external to the system. Over time solids accumulate in the LECA within the hydroponic media beds due to biofilm development, root and rhizome growth, accumulation of vegetative debris, the precipitation of chemicals and build-up of any remaining suspended solids from the input water. During harvesting the hydroponic media may be cleaned or rinsed to remove accumulated solids to avoid clogging and potentially harmful build-up of mineral salts. The monoculture beds undergo seasonal turnover (with all plants removed during harvest before transplanting of the following season's crop), while the polyculture beds contained both seasonal crops and mature plants left in situ for continual harvest over multiple seasons. The quantity of organic matter apparent on the LECA reflected the varying levels of disturbance between and within the beds according to crop turnover. The monoculture beds were disturbed and replanted with rocket 3-4 months prior to this study and had little visible solid accumulation. LECA from the spaces between the basil (transplanted 3-4 months prior to this study) and sorrel plants (left in situ for 5-6 years) underwent patchy disturbance and had a higher accumulation of solids with roots and vegetative debris visible between the aggregates. LECA from within the root system of the sorrel plants had the lowest disturbance and a high coverage of mature plants. These LECA were almost completely covered in organic matter with roots and vegetative debris surrounding the aggregates and macroinvertebrates (primarily oligochaetes) were evident in all collected samples.

NEMATODE EXTRACTION AND IDENTIFICATION
Free-living nematode taxa were extracted from a minimum of 100 cm 3 of each substrate using the tray method (Whitehead & Hemming, 1965), during a 48 h period. The resulting extract for each sample was passed through three 50 μm sieves and the solution condensed. Nematodes were quantified and identified to type using a Leica M50 stereomicroscope (Leica Microsystems). Individual nematodes were heat-relaxed in water at 65°C and immediately fixed in single strength TAF (Southey, 1986). Specimens were mounted in TAF on glass slides, sealed with paraffin wax and nail varnish and identified to genus where possible using a high-power compound microscope (Zeiss Axio Imager 2) utilising differential interference phase contrast (DIC). The assignment of feeding guilds was made following Yeates et al. (1992).

IDENTIFICATION OF OTHER LIFE
Following the extraction and identification of nematodes from the sample, the other organisms present in the solution were registered and quantified as abundant, poor abundant, or rare, using a combination of simple identification guides available on the internet (Microscopy-uk. org.uk, undated;Alejandro Garcia Z, undated;Math Science Nucleus, undated;Moreira & Diego, 2010). These data are available in Supplementary Table S1.

ABUNDANCE AND DIVERSITY
The highest mean abundance of individual nematodes was in the compost control (10 952 individuals (100 cm 3 substrate) −1 ) ( Table 2). Of the samples from the hydroponic beds, the highest mean abundance was found when compost was the substrate, although at only 2% of the abundance of the control compost external to the system. The coir supported the next highest mean abundance. LECA supported the lowest mean abundance of the substrates, around 10-15% of the number of individuals found in the compost in the hydroponic beds and 20-30% of the coir.
We recorded a lower number of genera and family richness for compost both within and external to the system when compared with the other substrates; this was despite the higher abundances of individuals. LECA in the more disturbed monoculture beds recorded a similar number of individuals but fewer genera and families than the less disturbed LECA in the polyculture bed.

GENERA DOMINANCE AND PERMANENCE
The dominant genus in the compost control, Halicephalobus (31.4% of individuals), was not found in the hydroponic beds. The control compost and compost within the hydroponic bed shared only Pelodera as a common genus at a similar relative abundance (27.3 and 24.2%, respectively), but both were dominated by nematodes from the family Rhabditidae making up over half of individuals sampled (52.6 and 54.0%, respectively) ( Table 3).
The compost control was the most consistent in terms of permanence, with all four genera represented in all the samples taken. The highest recorded abundance in the compost from the hydroponic beds was Butlerius spp. (31.4%); however, almost all individuals were recorded 498 Nematology Mean numbers of nematodes (100 cm 3 substrate) −1 for each substrate and location ± standard error (n = 9 for planted LECA from the polyculture bed and n = 3 for all other locations).
from a single sample. Poikilolaimus was the only genus showing permanence across all three samples (26.8% relative abundance). The coir in the monoculture bed was dominated by Plectus spp. (53.5%) with individuals recorded in all samples, followed by Achromadora sp. (38.6%). Both the coir and the LECA utilised in the monoculture bed showed occurrences of Tridentulus spp. and Monhystrella spp., though at low relative abundances (3.5 and 0.9%, respectively, in the coir and 6.7% for both in the LECA). Pelodera spp. had a high relative abundance in the monoculture LECA (80.0%).
LECA samples from non-planted patches in the polyculture bed shared the same two dominant genera as the coir (relative abundance of 35.1% Achromadora and 34.0% Plectus). LECA from within the root system of sorrel in the polyculture bed had Basiria sp. present in all samples and at the highest overall abundance (25.6%). Pelodera spp. were present in both the compost and the LECA (both planted and non-planted areas) from the polyculture bed and was overall the dominant genus in the hydroponic beds (18.6% individuals across both hydroponic beds).
During the identification of the samples from the aquaponics systems, first recordings in the UK for five nematode genera were made: Monhystrella, Oscheius, Tridentulus, Poikilolaimus and Prodorylaimium. Newly recorded genera for the UK were present in both beds and all substrates sampled. There were two new UK recordings found in the compost control:

MATURITY
The compost control, compost from the polyculture bed and LECA in the monoculture bed were dominated by individuals with a cp value of 1 (83.9, 85.3 and 83.3% relative abundance, respectively) and had no recorded individuals above a cp value of 2 (Fig. 1). The relative abundance of individuals with a cp value of 1 in the monoculture bed was almost a hundred times lower in the coir (0.9%) than LECA (83.3%), while in the polyculture bed the relative abundance of individuals with a cp value of 1 was higher in the compost (83.3%) than the LECA (9.3 and 14.0%). LECA from the polyculture bed had nematodes from c-p 1 to 5 and the highest MI values (Table 4), followed by coir from the monoculture bed, which had nematodes up to a maximum cp of 4. More than half of individuals in the coir had a cp value of 2 (57.9%).
Stock compost, compost from the hydroponic media beds and non-planted LECA from the monoculture bed had relatively low MI values (1.16, 1.15 and 1.17, respectively). The coir and both non-planted and planted LECA Vol. 25(5), 2023 499 1 Mean numbers of nematodes of a given genera (100 cm 3 substrate) −1 for each substrate and location ± standard error (n = 9 for planted LECA from the polyculture bed and n = 3 for all other locations).
2 Guilds indicate trophic group and cp value. The assignment of feeding guild was made following Yeates et al. (1992). Feeding types are predators (Pr), eukaryotic (Eu), fungivores (Fu), omnivores (Om), herbivores (He) and bacterivores (Ba). 3 'Unidentified' indicates individuals were identified to family level only. 4 Total nematodes from all genera (100 cm 3 substrate) −1 for each substrate and location ± standard error (n = 9 for planted LECA from the polyculture bed and n = 3 for all other locations).
500 Nematology Fig. 1. The percentage of the nematodes within assemblages assigned to each cp-value by substrate (coir, lightweight expanded clay aggregate (LECA) or compost) and location. The bed from which the sample is taken is indicated with braces. from the polyculture bed were comparatively higher (2.43, 2.80 and 3.0, respectively). The only potential plant-parasitic nematode found was Basira spp. in the LECA from the polyculture bed, resulting in a MI slightly lower than the MI for this bed only.

FEEDING TYPES
The identification of nematodes based on their morphological features was used to compare the feeding types between the samples (Fig. 2). Bacterivores made up over half of individuals in the compost control, the compost in the polyculture bed and both substrates in the monoculture bed. Nematodes from the non-planted LECA from the monoculture bed were entirely bacterivores, whereas the non-planted LECA from the polyculture bed consisted of bacterivores, eukaryotic feeders and predators.
The LECA in the polyculture bed had a wider range of feeding types, but bacterivores were still the most abundant feeding type in all LECA except those taken from within the root system of the sorrel in the least disturbed area of the polyculture bed. The LECA planted with sorrel showed equal dominance of herbivores and omnivores (25.6%) and was the only sample to contain all six feeding groups recorded in this study. Herbivores were unique to the LECA in the polyculture bed where sorrel roots were present.
Vol. 25(5), 2023 501 Eukaryotic feeders were only present in the monoculture coir (38.6%) and polyculture LECA. In the polyculture LECA the relative abundance of eukaryotic feeders was five times higher in the non-planted (35.1%) than planted (7.0%) areas of the bed.
Samples from the polyculture bed had a higher recorded abundance of predators than the monoculture bed. The highest relative abundance of predators was the compost in the polyculture bed (31.4%), followed by areas of non-planted (19.6%) and planted (16.3%) LECA in the polyculture bed. Predators were absent from LECA from the monoculture bed and compost control and represented only 2.6% of feeding types in the coir in the monoculture bed.
The compost control contained 16.1% fungivores, a feeding type absent from compost transplanted in the polyculture bed. Fungivores were only present in planted LECA from the polyculture bed, at a relatively low abundance compared to the other feeding types (11.6%).

ENRICHMENT AND STRUCTURE INDICES
The enrichment and structure indices were calculated based on the morphological identification of the nematodes in the samples and the food web analysis can be seen in Figure 3. The compost control and all samples from the hydroponic beds excluding the coir scored high on the enrichment index. Compost both external to and within the system scored low on the structure index and high on the enrichment index. LECA from the monoculture bed scored low on the structure index, while LECA from the polyculture bed scored high on both enrichment and structure.
The channel index could only be calculated for the planted areas of LECA in the polyculture bed and the stock compost. Fungivores were absent from all other samples from the hydroponic beds. The CI values for both were low (<50%), suggesting bacterial decomposition channels were predominant in the stock compost and all sampled areas of the hydroponic beds.

Discussion
Our findings present the first snapshot in time for nematode assemblages in a commercial aquaponics system and provide a platform on which to improve understanding of the role of nematodes in aquaponics. The relative abundance of bacterivorous nematodes found throughout the substrates (LECA, coir plugs and compost) within the hydroponic media beds reflects the nutrient-rich environ-502 Nematology  Fig. 3. Food Web Analysis for aquaponics substrate samples. Structure Index (SI) and Enrichment Index (EI) are shown. The left-hand quadrants can be considered immature and the right-hand quadrants mature. The top two quadrants can be considered enriched, and the bottom two quadrants depleted. The shape of the icons represents the bed from which the samples were taken (monoculture, polyculture or stock compost) and the colour represents the substrate type (coir, lightweight expanded clay aggregate (LECA), or compost) and location. ment and suggests nematodes may play a role in nutrient cycling within aquaponics. Nematode-driven nutrient cycling in aquaponics may mirror nitrogen mineralisation in traditional soil systems (De Ruiter et al., 1993;Ferris et al., 1998) with nematodes increasing the availability of inorganic N to the surrounding plants. Compost in the hydroponic media beds provided a suitable habitat for nematodes, predominantly bacterivores, but contained predatory nematodes and no fungivores compared with compost external to the system (comprised mostly of bacterivores and some fungivores). The compost from within the hydroponic beds was free of parasitic nematodes but did not reflect a structured community (SI score of 0). The lack of community structure is unsurprising, given the short period of time the compost is present (as it is added and removed with the seasonal crops). The presence of the basil plant within the compost in the hydroponic media bed may have played a role in the differences between the assemblages of the compost samples within and external to the hydroponics media bed and future replication should include a control of basil grown in compost external to the system. The cultivation of nematode assem-blages high in bacterivores and free of plant-parasitic nematodes could be of interest to commercial growers, but further studies would be required to determine if there are any meaningful effects on crop yields. The LECA from the polyculture beds, with higher build-up of organic matter, show mature (MI) nematode assemblages with a wide range of feeding types present. By contrast, LECA with low levels of organic matter build-up, such as those surrounding the rocket from the monoculture bed, had assemblages with a low maturity index and harboured only bacterivores. The greater detritus build-up, whilst also increasing the available habitat for nematodes, is reflective of the level of disturbance. LECA that are removed and cleaned regularly will have low levels of detritus and this process is a considerable disruption to the system. By contrast, LECA left in the hydroponic media bed from season to season are relatively undisturbed and will build up considerably more detritus. Seasonal harvesting events in the monoculture bed create a fluctuating pattern of plant density and a lower maturity of plants and root systems present compared to the more stable, on-grown polyculture bed. The Vol. 25(5), 2023 503 food web analysis further strengthens these finding with the LECA with lower organic build up and seasonal crop from the monoculture media bed showing high levels of enrichment (high EI value) and no structure (zero SI), whereas the polyculture LECA whilst still showing high levels of enrichment have high levels of structure. If we compare between the two polyculture LECA, we find that the planted LECA with the highest level of organic matter build up has higher structure than the non-planted ECP from the same bed. This may reflect disturbance, though the proximity of the planted LECA to the sorrel and the inclusion of sorrel root material may be responsible for the higher MI and SI values. These findings are comparable with traditional soil systems where less disturbance allows for ecological succession. If LECA are new or regularly cleaned there will not be an opportunity to produce structured nematode assemblages in the aquaponics system. Whilst it is possible that some of these differences (particularly at the individual species level) reflect the differing number and type of crops between the hydroponic media beds (rocket monoculture vs basil and sorrel polyculture), in the context of the bed layout and our wider results it seems likely to play only a small role in determining differences in either feeding type or maturity indices. Interestingly, the LECA from the polyculture bed showed less enrichment than the monoculture bed, perhaps due to the increased structure of the nematode assemblage. The nonplanted LECA from the polyculture bed showed much less enrichment than the planted LECA from the same bed, though this may be due to the absence of an embedded plant in turn reducing possible exploiter species compared with the planted LECA. Some growers and LECA sellers recommend against allowing the build-up of organic matter on LECA as it might lead to salt accumulation, but our findings suggest that detritus build-up may be creating a soillike environment for structured nematode assemblages. Perhaps this study may provide evidence in the battle towards gaining organic status for aquaponics in the EU, with these soil-like substrates showing properties of healthy soils, sustaining diverse nematode and nonnematode communities. In fact, the food web analysis of LECA subject to lower disturbance and with a high buildup of organic matter reflected a much more structured and less enriched habitat than the compost used external to and within the system. Although we note the compost used in this research was younger than the organic matter buildup in the sorrel bed, it likely reflects the age and maturity of compost used by many growers.
The coir plug (and associated rocket root material) was also a suitable substrate for nematodes but was unique in that it showed no signs of enrichment. Given that nutrient availability in the root zone is key to crop quality and yield this could have significant implications. The very low enrichment score is particularly surprising given the continually circulating nutrients. Reports on the effectiveness of coir as a growing medium varies according to plant species, source of the coir and the growing conditions (Konduru et al., 1999;Abad et al., 2002;Linderman & Davis, 2003;Mariotti et al., 2020) but it may have a relatively low cation exchange capacity and therefore ability to retain nutrients in comparison to alternative substrates such as peat (Abad et al., 2002). According to sources such as Handreck (1993), deficiencies in the chemical properties and buffering ability of coir may be negated by fertilisation, but despite high nutrient levels in the water supplied to the coir in the studied system the nematode assemblages present failed to support enrichment. The assemblages found in the coir plug were more mature (MI of 2.43 compared with 1.15-1.16) than the compost samples but less than the LECA with high organic matter build-up (MI 2.80-3.00). The nematodes reflected an assemblage with low structure with nematodes predominantly having cp values of 2 and 3 but very few cp 1 values. Perhaps, the coir is not a suitable habitat for the classic cp-1 exploiters and, instead, nematodes feeding on eukaryotic organisms in the circulating fluid become more established and so the key nematode indicators of enrichment are not present. Given the increasing use of coir as a substrate for planting, understanding the ability of coir to retain nutrients will be of interest for future studies.
The nematode taxa found in the hydroponic media beds are predominantly those associated with terrestrial habitats, in particular, moist soil systems, but there were very low levels of plant-parasitic nematodes. The only potential plant parasite identified was Basira found in the LECA from the polyculture bed, though this is argued to be a 'plant feeder' rather than a 'plant parasite' and is thought to cause little damage (Munawar et al., 2021). It will be important to investigate whether the low levels of plant-parasitic nematodes found in this study are typical of aquaponic facilities and, if so, whether any conditions of the system responsible can be replicated in conventional agricultural soils. Plant-parasitic nematodes are estimated to reduce global crop production by around 10% (Whitehead, 1998) and so determining methods and conditions for their control is economically important. 504 Nematology This study made the first recordings in the UK of five nematode genera: Monhystrella, Oscheius, Tridentulus, Poikilolaimus and Prodorylaimium. These newly recorded genera for the UK were present in both beds (for all substrates sampled) and have all been recorded in terrestrial soils elsewhere in the world, often in moist or very moist soils. Achromadora, Actionalamidae, Plectus, Coomansus and Mylonchus, all frequently found in this study, are also commonly found in moist terrestrial samples. Ironus is commonly found in freshwater habitats and has been reported in soil. Pristmatolaimus is typically considered to be a freshwater taxon but has been found in moist soils. It is perhaps unsurprising given the combination of substrates, including soil and organic matter, with the aquatic environment of an aquaponics system that we would find nematodes that thrive in moist conditions, but it is noteworthy that many of these are found in terrestrial soils and, although the species composition is different to a typical UK soil, they are functionally similar to the nematode assemblages found in terrestrial soils. It is possible that some nematodes identified may be just passing through and not resident to a given substrate, in turn altering the cp scores. However, given the distribution of nematodes to certain locations in this study, this would likely only be a small factor in the overall composition of the assemblages. It appears the moist conditions of the aquaponics systems are establishing assemblages that could perform similar roles to functional soils. Despite the functional similarities of the newly identified taxa for the UK, the low levels of plant parasites and the abundance of nematodes favouring moist conditions all indicate nematodes assemblages quite different to those of conventional UK agricultural soils and composts. However, it is also noteworthy that there were two new UK recordings found in the stock compost: Metarhabditis and Halicephalobus. The recording of new genera in the compost likely reflects the under-researched nature of free-living nematodes. There are few skilled nematologists (Coomans, 2002) and many are focused predominantly on identifying terrestrial and, more specifically, plant-parasitic nematodes. The reliance on comparisons to previously published literature rather than use of type specimens for descriptions of new nematode species further hinders identification of nematodes from underresearched populations such as those in this study (Abebe et al., 2014).
This first exploration of nematode assemblages in aquaponics systems tells a story of ecological succession where LECA left undisturbed can build up organic mat-ter that harbours structured communities of nematodes comprised of taxa that can be found in terrestrial soils. It suggests that substrate type and growing methods influence the nematode assemblages present in aquaponics and that these may impact on plant health and yields. Our findings shine a light on nematode assemblages in aquaponics but are limited in scope. Given that a functional commercial system uses many different crops and composts over time, the exact reason for any nematode assemblage is tricky to ascertain. The aquaponic farm produces a diverse range of seasonal and year-round crops, meaning it was impossible directly to compare substrates or level of disturbance using a single plant species. Furthermore, we only sampled at one time point and so it is not clear given the controlled conditions whether seasonal fluctuations of nematode populations will take place. Our data suggest there is an increase in the diversity and structure of nematode assemblages within the hydroponic media beds over time (when disturbance is low) and samples from more time points may have helped support these claims. There is also a small risk that some small early-stage juveniles from certain free-living nematode taxa may not have been recovered using the three 50 μm sieving steps, and the use of a 45 μm sieve in future work may enhance the recovery of these nematodes. Despite the limitations of sampling from an operating commercial facility, our findings provide a useful snapshot of the nematode assemblages within an aquaponics system. This research sets the scene for a range of possible experiments that could benefit aquaponics farmers. It will be important to determine which substrates are likely to be best for crop growth by measuring yields under controlled conditions. This should be performed on a range of different crops and in both commercial systems and laboratory settings. Determining nutrient retention and availability in the coir will further our understanding of the best substrates for crop growth and these findings will be widely applicable outside the field of aquaponics. The greater number of nematode families, higher MI and SI scores of the LECA with organic matter buildup from the polyculture bed compared with the LECA with much less organic matter build-up suggest that experiments looking at succession of nematodes over time would yield interesting results. The aquaponics system studied had very low levels of plant parasitism, and determining whether any inherent properties of the system help to supress harmful nematode populations would be of interest to both aquaponic and conventional growers. In essence, we hope this research will provide Vol. 25(5), 2023 505 a platform for future studies seeking to improve the viability, productivity and profitability of aquaponics.