Nasturtium officinale W.T. Aiton (= Rorippa
nasturtium-aquaticum s.s. (L.) Hayek), Water-cress

Account Summary

Growth form, identification and preferred habitats

This white-flowered, creeping, emergent or floating perennial water-cress species s.s. can only be distinguished from the species in the broad sense (s.l. or agg.), when the plant has well-formed fruits and seeds. The seed pods of N. officinale s.s. are short and round, and the brown seeds themselves are in two rows per locule (Howard & Lyon 1952a). While these are good, reliable identification characters, Rich (1987b, 1991) warns that confusion can still occur with N. microphyllum (Narrow-fruited Water-cress), as its seeds, although typically forming a single row, may also occur in two rows, but they only do so at the base of the fruit pod. The most reliable distinguishing characteristic of this section of the genus is the sculpturing of the mature seed coat, and since a microscope is required to examine the seed, this is not a field character.

N. officinale is typically found in clean, flowing, often shallow waters, and experiencing fairly well-illuminated growing conditions. These conditions are met in lakes, streams, springs and ditches, and on marshy ground and flushes beside such waters, including seasonally wet grazing meadows and quarry drains.

Associated species include the very much more commonly recorded and ecologically similar Apium nodiflorum (Fool's-water-cress), Myosotis scorpioides (Water Forget-me-not), Veronica beccabunga (Brooklime), Mentha aquatica (Water Mint) and Filipendula ulmaria (Meadowsweet).

When considering altitude as a delimiting ecological factor, there is a tendency for N. officinale to occupy more or less exclusively lowland sites, being replaced at higher altitude by either the closely related N. microphyllum, or by their hybrid, N. × sterilis (Howard & Lyon 1952; Grime et al. 1988; Preston & Croft 1997).

Variation

N. officinale s.s. is vegetatively extremely plastic in response to varying environmental growing conditions. Such variation extends from the overall scale of the plant (ranging from dwarf to giant (6-200 cm)), leaf shape, size and colour, growth-form, life-form which can be emergent or floating, and phenology being perennial or annual (the latter generally appearing in drier sites) (Rich 1991).

In the past, a number of varieties were distinguished, some of which had a genetic basis, others merely growth forms, so that confusion reigned (see Howard & Lyon 1952, p. 228). Both Manton (1935) and Howard & Manton (1946) recognised that Nasturtium officinale s.s., is a fertile tetraploid species (2n=32).

Flowering reproduction

Flowering and fruiting occurs freely under both dry and moist growing conditions, stretching from the end of May through to September, or even November in a mild year. However, the reproductive process is restricted or prevented in very wet or totally submerged situations. The terminal racemes of small white flowers are self-compatible and seed set is normally abundant. No doubt at least some seeds result from crossing achieved with the assistance of insects visitors, since numerous species of flies and a few bees and beetles are attracted by a plentiful supply of nectar and pollen (Howard & Lyon 1952). Seed production is very variable, but normally it is abundant. Dispersal is usually achieved by floatation in water, effected by an air bubble formed around the seed coat. Seed may also be dispersed over long distances and between water bodies embedded in mud, which is then transported by animals including birds and man. There is no dormancy requirement for germination, but buried seed can remain viable in the soil for between one and five years (Howard & Lyon 1952; Thompson et al. 1997).

Disturbance and vegetative reproduction

Water-cress plants are frequently grazed by a wide range of invertebrates, and in some of its seasonally drier, grassier sites, plants are also very obviously subject to grazing and trampling by cattle which must limit vegetative growth and frequently prevent flowering and fruiting. However, all water-cresses (ie both species plus their hybrid) have great powers of vegetative reproduction and patch formation through their stems creeping and rooting profusely at their nodes. As with associated plants of similar vegetative properties, detached portions of plant occasionally disperse by floatation, particularly after spates or floods, compensating to some degree for the loss of seed production under such circumstances (Ridley 1930, pp. 182 & 546).

Fermanagh occurrence

In Fermanagh, N. officinale s.s. is slightly more frequently recorded than N. microphyllum, ie appearing in 61 tetrads (11.6%), compared with 44 (8.3%) for the latter, but this is hardly statistically significant. N. officinale s.s. is thinly but widely scattered throughout the VC, but is a little more frequently found on the Western Plateau.

Irish occurrence

The New Atlas map shows N. officinale s.s. as scattered but locally frequent in Ireland, having a minimum of one hectad with recent records in every Irish VC except NE Galway (H17). At the same time, several Midland and western Irish VCs appear to have little more than this level of presence recorded. It is possible that this taxon's apparent preference for, or in some parts of England, restriction to, eutrophic or mesotrophic calcareous waters and soils, and its almost total avoidance of peat (both acidic and alkaline) may, at least in part, explain this distribution pattern. Airy Shaw (1948) first observed this ecological tendency distinguishing N. officinale s.s. from N. microphyllum, but he cautioned that, "it can in no way be expected to hold good over the whole country".

British occurrence

In Britain, the New Atlas map shows N. officinale s.s. as being common and widespread in lowland areas, apparently becoming more coastal in occurrence further north, particularly in NE England and E Scotland. This distribution might well reflect the species' known sensitivity to frost, but since it is recognised that there is an unknown degree of under-recording of this and the related taxa, there is a danger, however slight, that it could simply represent the extent of the local recording effort.

It must also be remembered when attempting to interpret the mapped distribution that water-cress has been cultivated for thousands of years (see also our account of the species aggregate for historical uses), the plants commercially involved being this species, and to an even greater extent, the hybrid, N. × sterilis, and the major region of large-scale cultivation in the last 150 years has been in the S and E of England (Howard & Lyon 1952; Preston & Croft 1997). We might therefore expect a greater possibility of human dissemination of these plants in and around the areas of cultivation and their markets.

European and world occurrence

The European distribution illustrated in Jalas & Suominen (1994, map 2327), shows this species widespread and native in S & W Europe, southern parts of C Europe, but absent from Norway, Sweden and Finland except as an introduction in the extreme south of Sweden. In contrast to the situation in Britain, N. officinale s.s. is the more generally cultivated water-cress in European countries and, in particular, this form of the plant has been widely introduced, cultivated and naturalised elsewhere in the world, including in N & S America, many parts of Africa, the Far East and Australasia (Hultén 1971, Map 195 and p. 204). The species was first recorded in New Zealand, for example, in 1852 and it is now common in streams, ditches and drains throughout (Webb et al. 1988).

Names

The name 'Nasturtium', or rather the Latin 'Nasturcium', was given by the Classical Roman author, Pliny, to a plant with a distinctly pungent flavour or odour. It is derived from two Latin words 'nasus' nose and ' torquere' meaning 'to twist' or 'torture' (Gilbert-Carter 1964; Grigson 1974), descriptive of the reaction the original plant invoked, and one that some palates might still present to the peppery bite of water-cress. The name 'Nasturtium' became linked with the S American genus Tropaeolum, a flower in a completely different family, simply on account of the pungent smell and flavour of its leaves. The first Tropaeolum species was introduced to Europe through Spain from Peru in the late 16th century and it was originally given the name Nasturtium indicum, Indian Cress (Gerard 1633, p. 251-2; Grigson 1974).

Threats

None apart from the possibility of excessive pollution or eutrophication.

HAS CHANGED NAME SINCE THE FLORA WAS PUBLISHED

Nasturtium microphyllum (Boenn) Rchb. (= Rorippa microphylla (Boenn.) Hyl. ex Á. Löve & D. Löve), Narrow-fruited Water-cress

Native, occasional. Native world distribution uncertain, but probably so in Europe and in C Asia: very widely introduced and naturalised beyond this range.

1860; Smith, T.O.; Ardunshin.

June to November.

Growth form, identification and preferred habitats

Like the two other forms of water-cress, N. microphyllum is a patch-forming perennial found floating in clean running, productive rivers, streams, springs, ditches and in lakes and ponds subject to fluctuating levels, plus as an emergent on marshy ground beside such waters. Unlike the other forms, it is not commercially cultivated and has not been spread in that manner.

Vegetatively, the two water-cress species (N. officinale s.s. and N. microphyllum) are too similar and too variable to distinguish from one another. Although this plant bears the specific epithet 'microphyllum' (Greek, meaning 'small-leaved'), the name is a very misleading one, since apart perhaps from the terminal leaflet, its leaves and leaflets are not significantly smaller or narrower than those of its close relative N. officinale s.s. In reality, leaf shape and size are absolutely valueless for distinguishing the two species and their hybrid. In comparison with N. officinale s.s., however, the seed pods of N. microphyllum are long, thin and curved and they contain a single row of seeds − except sometimes near the base of the fruit valve. The seedcoat also has a much finer grained texture or network of reticulations than N. officinale s.s., having 12-20 depressions or cell alveoli across their greatest width (Rich 1991, pp. 152-3; Stace 1997, photomicrograph Fig. 254, 11).

N. microphyllum occupies very similar moderately nutrient-rich, productive, somewhat disturbed, shallow flowing water or waterside marsh habitats to those of N. officinale s.s. (Water-cress), but while the two may co-exist in more definitely calcareous situations, this species has a rather wider ecological range, being tolerant of moderately acidic and less-calcareous environments than the latter. All forms of water-cress are absent from unproductive, strongly acidic, very shaded or highly disturbed or exposed habitats (Grime et al. 1988). N. microphyllum is a little more shade tolerant and is slightly the more frost resistant of the two species, and therefore can occasionally be found at higher altitudes − ie the current altitude record in B & I stands at an exceptional 550 m near Moor House in Westmorland & Durham (VCs 66 & 69) (Preston & Croft 1997; Preston et al. 2002).

Seed dispersal and vegetative reproduction

It should be borne in mind, that since it is not cultivated, N. microphyllum is less likely to be introduced or spread by man than the two cultivated forms, N. officinale s.s. and the hybrid, N. × sterilis (Howard & Lyon 1952). On the other hand, since all forms of water-cress grow on muddy stream banks, seed of the two species may be relatively easily transported over long distances embedded in mud adhering to birds' or other animals' feet, or on their external surfaces.

Broken stem portions of all three forms of water-cress, including the sterile hybrid, have an amazing ability to form adventitious roots within a matter of hours, even on portions of the inflorescence, allowing the plants to reproduce very readily.

Fermanagh occurrence

As is the case in nearly every part of B & I, Robert Northridge and the current author are sure that N. microphyllum is under-recorded in Fermanagh. It really should be more frequent than the occasional, limited records in the Fermanagh database spread across 44 tetrads (8.3%) suggest. As the tetrad distribution map indicates, it is thinly and widely scattered throughout the VC. At the same time, it is certain that it is nothing like as common as N. officinale s.l. (Water-cress), a not terribly surprising conclusion, since this entity represents the vegetative forms of three closely related taxa, which unfortunately cannot be properly distinguished without ripe fruits.

Unusual ecological behaviour of N. microphyllum in New Zealand: While the vast majority of N. microphyllum populations are recorded in or beside shallow water less than 75 cm deep, in New Zealand Michaelis (1976) found the species growing in and around very clear cold springs, not only as an emergent and as a mat-forming, floating plant near the shoreline but, most unusually, a somewhat dwarfed growth form was also growing completely submerged and rooted in gravel. This had translucent, dark green, rolled up leaflets, and occurred at depths ranging down to 6.5 m. At depths below 100 cm it was, of course, non-flowering. It also showed little seasonal change in growth form, other than a slight elongation of the internodes in summer (Michaelis 1976).

A brown winter pigment form

Plants of N. microphyllum have a strong tendency to turn a deep purple brown anthocyanin colour during the colder winter months, a feature which when combined with lesser vegetative vigour than 'Green' Water-cress, N. officinale s.s., or the even more hardy 'Brown' or 'Winter' Water-cress (the hybrid N. × sterilis), has excluded its use in commercial cultivation (Howard & Lyon 1952, p. 242).

Genetic makeup and species origin

N. microphyllum was originally thought to be an autotetraploid, ie the result of spontaneous chromosome doubling within a species, N. officinale being put forward as the putative ancestor (Manton 1935). However, N. microphyllum is now recognised to be an allopolyploid octoploid species (2n=64), which has arisen through hybridization between N. officinale s.s. and an unknown species. The unknown parent was presumed to be a member of the closely related genera Rorippa or Cardamine (Howard & Manton 1946). On account of the seeds being arranged in the pod in a single row and of its morphological resemblance to N. microphyllum, Oefelein (1958) suggested Cardamine amara (Large Bitter-cress) might be the unknown parent. A recent large-scale isozyme analysis of Nasturtium and Rorippa species has proven that the unknown parent is a Rorippa species. R. sylvestris (Creeping Yellow-cress), R. amphibia (Great Yellow-cress) and R. palustris (Marsh Yellow-cress) were considered likely candidates, and all of them carried the marker alleles of N. microphyllum, while Cardamine amara did not and is definitely not involved in the parentage (Bleeker et al. 1999).

Being octoploid, and thus having paired chromosomes, N. microphyllum is perfectly fertile; like N. officinale s.s., it is self-compatible and sets copious seed. Also, like N. officinale s.s., it has a very long flowering and fruiting season, but it comes into flower about two weeks later than the former. The larger white flowers of N. microphyllum are probably even more attractive to insect pollinators than those of N. officinale s.s. The first flowers of water-cress plants that open set only a few seed and, especially in N. microphyllum, plants set very little seed if they are crowded or growing in less than favourable conditions (Green 1962). Several properties, including vegetative reproductive ability, seed dispersal, germination and longevity, all appear almost identical in the two water-cress species (Howard & Lyon 1952).

British and Irish occurrence

While there certainly still is an unquantified degree of under-recording of the two white-flowered water-cresses and their hybrid, both species are undoubtedly common and widespread throughout B & I (Preston et al. 2002). In Britain, N. microphyllum is the less frequent and abundant species in the southern half of the country, but on account of its wider ecological range, it probably is the more common and abundant of the two further north, and especially so in E Scotland.

In Ireland, N. microphyllum appears to be slightly less common overall than N. officinale s.s., but it is probably the more frequent of the two species in the central region of the island and on higher ground throughout. In these two situations, winters are noticeably colder, and peat soils and their acidic influence on drainage water are very common and widespread (Preston et al. 2002).

European and world occurrence

Beyond B & I, N. microphyllum is very probably under-recorded in Europe, where it is mainly noted in the west from N France, Belgium, the Netherlands, Denmark, NW Germany, Switzerland and N Italy and very thinly scattered further south and east (Jalas & Suominen 1994, Map 2328). The native range is said to extend to C Asia and E Africa and it has been introduced to S Africa, N America and Australasia (Rich 1991).

Threats

As we have an imperfect record of the presence of Narrow-fruited Water-cress throughout B & I, we cannot readily ascertain what population changes are due to the widespread environmental pressures on wetlands, such as local drainage operations and any form of pollution (including nutrient enrichment). However, we can predict that either or both these factors would quickly result in serious losses of Water-cress through suitable habitat destruction and increased competition from taller wetland species (Grime et al. 1988).

Rorippa islandica s.s. (Oeder ex Gunnerus) Borbás, Northern Yellow-cress

Native, very rare. Eurosiberian boreal-montane.

28 September 1995; McNeill, I.; on dumped material, roadside waste ground, Crocksoult Td, SE of Glen Lodge.

June to September.

Taxonomy and growth form

R. islandica s.s. and R. palustris (Marsh Yellow-cress) are two morphologically and ecologically very similar pale yellow flowered taxa which originally were combined, and from the point of view of the field botanist and the ecologist, they very conveniently remained so until the late 1960s. At that date, the Scandinavian taxonomist Bengt Jonsell revised the NW European species within the genus, and he split the previous R. islandica s.l. group into two species of different ploidy level. The tetraploid form, now recognised as R. palustris, has become recognised as by far the commoner and more widespread taxon in comparison with the less weedy diploid plant, R. islandica s.s. (Jonsell 1968). Plants of both species are habitually inbreeding and generally perform as annual species, although they can sometimes also behave as biennials, or even as short-lived perennials. Jonsell (1968) found that individuals could survive in cultivated plots for up to five years.

Identification

R. islandica s.s. is quite difficult to distinguish from R. palustris and, rather rarely, the two taxa can be found growing together (Goodwillie 1995; Chater & Rich 1995). Microscopic examination of the coat of ripe seed is used to confirm their identification, the seed coat of R. islandica s.s. being more finely colliculate, ie having smaller indentations than those of R. palustris, and thus forming a finer, less-distinct, net-like pattern (Jonsell 1968; Rich 1991). Those familiar with the plant find that it can be distinguished from R. palustris by its fruit characters, even at a distance of several metres. Fruits of R. islandica s.s. are sometimes secund and swept downwards (ie the whole fruiting inflorescence appears one-sided), and since the stalks of the fruits are relatively short, the arrangement looks rather dense. In contrast, the fruits of R. palustris are arranged evenly around the stem and they are not so densely crowded. The fruit shapes are also distinctly different to the practised eye or hand, those of R. islandica s.s. having a more rectangular, square-ended appearance in section – as can be ascertained by trying to roll them between finger and thumb (Rich 1991; Chater & Rich 1995).

Irish occurrence and preferred habitats

R. islandica s.s. is regarded as being fairly rare in B & I, although not quite as rare as was previously supposed (Rich 1999; New Atlas; BSBI Database accessed 21 June 2020). Survey work associated with the BSBI Atlas 2000 project found that since its recognition as a separate species, R. islandica s.s. has been quite frequently discovered in Ireland, having previously been overlooked, a situation undoubtedly also the case elsewhere in these islands.

The habitat of the first Irish station in Connemara which Jonsell himself discovered in 1964 (or rather rediscovered, since herbarium specimens in BM and DBN indicated that R.S. Shuttleworth of Berne had collected the plant in the same area as early as 1831 and 1832 (Scannell 1973)), was bare (probably muddy) ground near the Atlantic coast at Renvyle, W Galway (H16). Here, the plant grew as a small, prostrate or decumbent annual in association with Potentilla anserina (Silverweed), Cardamine pratensis (Cuckooflower), Ranunculus repens (Creeping Buttercup) and Nasturtium officinale s.s. (Water-cress) (Scannell 1973).

In 1971, Maura Scannell herself found the second Irish station for the plant in a somewhat more inland site, where it occupied bare ground towards the centre of a dried out turlough (ie a vanishing lake − a grazed grassy hollow which regularly floods during wet periods), in the Burren limestone region of Co Clare (H9) (Scannell 1973). A total of 15 additional turlough records soon cropped up in W Ireland, all of similar habitat, ie seasonally inundated wet meadows, in cattle- or sheep-grazed, trampled and often manure-enriched, muddy ground.

By 1995 the Irish distribution of R. islandica s.s. included a further four west coast Irish VCs (SE Galway (H15), NE Galway (H17), Roscommon (H25) and E Mayo (H26)), plus another in Co Kilkenny (H11) further to the SE of the island (Goodwillie 1995). Over the space of several visits to the Kilkenny station, Goodwillie noticed small specimens which he reckoned probably germinated in mid-August, reaching the fruiting stage at the end of October. This observation provided evidence of phenological plasticity, which fitted this annual species well for existence in ground subject to intermittent flooding and the other forms of habitat disturbance already mentioned. Most Irish records of R. islandica s.s. have been associated with seasonally or periodically inundated, cattle- or sheep-grazed, trampled and often manure-enriched, sometimes muddy pastures around lakes and turloughs in limestone areas like the Burren, Co Clare (Scannell 1973; Goodwillie 1995).

Recently, this taxon has begun to turn up at numerous locations in the north of Ireland, thanks almost exclusively to the plant hunting efforts of Ian & David McNeill together with John Harron and John Faulkner (McNeill & Hackney 1996; McNeill 1998). The habitats of these widely scattered NI sites include some very heavily disturbed amenity areas in public parks, including well-trodden paths and verges, cracks in concrete caravan park hard standings, beside a sheep dipping enclosure and in dumped material on waste ground.

Fermanagh occurrence

The first Fermanagh county record listed above is supported by a verified voucher specimen. Three earlier records by an EHS Habitat Survey Team that locate the species in muddy situations at Inishroosk, Upper Lough Erne in both a marsh and on the lake shoreline on 21 June 1985, and subsequently from Trannish Island, Upper Lough Erne on 16 June 1987 might well have been genuine discoveries. However, without herbarium vouchers they were not fully acceptable first county records, and originally Robert Northridge and the current author as BSBI Recorders for Co Fermanagh, noted them with caution. However another survey of selected sites in Upper Lough Erne commissioned by NIEA and carried out in 2006 by ENSIS (English Nature Site Information System) recorded R. islandica at four sites, including Trannish Island, 2 July 2006, which of course supports the 1987 record at that site made by the EHS Habitat Survey Team. The details of three other ENSIS sites are: Belleisle ASSI, 5 July 2006; Castle Lough, 22 July 2006; and Derrykerrib Lough, 22 July 2006.

The outcome of this is that we now have a total of ten records from five tetrads as shown in the Flora of Co Fermanagh map: the two McNeill records to the north of the VC are quite remote from the remainder around Upper Lough Erne. Examination of the BSBI database shows just one additional site at Portora Boathouse made by RHN and Tim Rich in 2013.

British occurrence

The New Atlas hectad map showed R. islandica s.s. as rare, scattered and decidedly northern and western in its distribution in B & I. In the last 20 years, it has been much more widely recorded throughout the whole latitudinal range of Britain from Plymouth to Orkney though it remains most concentrated in the same northern and western range although thinly scattered elsewhere. In 1991, R. islandica s.s. was recorded growing in rock crevices beside a river in SW Wales and, subsequently, over the next few years nine additional stations were found in the same Welsh region, occupying a surprisingly wide range of habitats. These included damp depressions along a river flood plain, around muddy pools in either nutrient-poor or rich conditions, ditches and hedgerows and, much more unexpectedly, on somewhat drier but still disturbed conditions on waste ground and rubbish tips. The only obvious shared factor in these situations is their ecologically open, fairly damp character, conditions maintained by periodic disturbance which most often involves flooding, grazing and trampling, but which occasionally arises as a consequence of human activities as varied as dumping, trampling and fishing (Chater & Rich 1995; Rich 1999). The Welsh representation is still very noticeable and even more widespread today (BSBI Database Map accessed 21 April 2020).

European and world occurrence

Beyond B & I, R. islandica s.s. is similarly confined to the NW extremities of Europe (including Nordland fylke in N Norway, Iceland and S & W Greenland). However, it does also crop up as very disjunct populations around alpine lakes in the mountain ranges of S Europe, stretching W to E from the Cantabrian Mountains to the Northern Pindhos (Jonsell 1968, Fig. 7; Jalas & Suominen 1994, Map 2323).

Jonsell also recorded an area of R. islandica occurrence in Russian Siberia, but since he distinguished and named this particular geographical form subsp. dogadovae (Tzvelev) Jonsell, it must not be quite the same plant that is met in B & I, which is referable to subsp. islandica (Jonsell 1968).

Seed dispersal

Jonsell (1968) remarked that R. islandica s.s. is especially common in Iceland and is particularly abundant around Lake Myvatn, a noted wildfowl site. This fact, together with the then known distribution of the plant, led him to suggest that geese might well be transporting agents for seed of the species during their annual migration, either internally in their gut, or else embedded in mud on their feet. Unlike many other Brassicaceae, the seed coat epidermis of R. islandica, and of R. palustris for that matter, does not produce sticky mucilage when wetted, but rather the cells are hollow, which may help confer the prolonged buoyancy they demonstrate in water (Jonsell 1968; Chater & Rich 1995).

The probable means of seed transport in this species, both local and long-distance (eg water flow, mud and birds) are entirely conjectural at present. In the nature of stochastic properties, these things always are. However, the relevant arguments have been discussed very clearly by Chater and Rich (1995).

• Find out more about the format of the species accounts
• View abbreviations
• View references for all species accounts
• View a gazetteer of all the sites