Rosa caesia subsp. caesia Sm.,
Hairy Dog-rose
Account Summary
Native, very rare, possibly under-recorded or mis-identified.
1949; MCM & D; glen to the west of Lough Fadd.
This subspecies of Dog-rose is apparently extremely rare throughout Ireland. It might well be so, but in common with all Irish roses, it is very possibly under-recorded too. The FNEI 3 suggests that subsp. caesia is or was recorded, albeit rarely, in all three VCs in NE Ireland (H38-H40). Some of the records listed in FNEI 3 date from 1899 and the most recent date listed is 1955.
The New Atlas map, on the other hand, features just four hectads in all Ireland (no Fermanagh records are included). Two of the four Irish stations that the hectad map features are in Co Down (H38) and one in Co Londonderry (H40); all three are pre-1970 in date. The remaining Irish station is located further SW in Co Sligo (H28) and it dates from the 1987-99 period.
The two Fermanagh records in the Fermanagh Flora Database were made by Meikle and co-workers over 70 years ago. Geographically they lie between the extremely few stations for this taxon mapped in the New Atlas. The Fermanagh sites are remote from one another in the west of the VC. The dates and locations are: a glen west of Lough Fadd, 1949, determined by N.Y. Sandwith; and Garrison, near Lough Melvin, 1950. Mr N.Y. Sandwith, as ever, followed Wolley-Dod's approach to rose identification and naming. He identified the plant as R. coriifolia Fr. var. typica W.-Dod. This taxon and name translates as the rarer of the two R. caesia subspecies, subsp. caesia, but ideally RHN and the current author (RSF) would like to see bushes of the plant in Fermanagh today to prove that these earlier botanists were correct in their findings. The most likely alternative would be that the plant in question is another R. caesia cross with R. canina, since the hybrid R. × dumalis appears to be much more common than either or indeed both forms of R. caesia. Another possibility is a pubescent form of R. canina.
The most recent treatment of R. caesia by Sell & Murrell (2014) recognises the extensive variation within the species by naming no less than 19 varieties of it.
Rosa caesia subsp. vosagiaca (N.H.F. Desp.) D.H. Kent (R. caesia subsp. glauca (Nyman) G.G. Graham & Primavesi, R. afzeliana Fries) Glaucous Dog-rose
Native, very rare but probably under-recorded.
1948; MCM & D; limestone pavement near Crevinish Castle, Gubbaroe Point, Lower Lough Erne.
July to September.
There are just three records for this rose subspecies in the Fermanagh Flora Database, but it is very possibly frequent, probably being ignored, overlooked and lumped into R. canina agg. or perhaps mistaken for R. × dumalis, although again there only are a handful of records of the latter in any case. Two of the probable R. caesia subsp. vosagiaca records are by Meikle and his co-workers, the first being listed above. The difficulty here is that this castle is several kms NE of Gubbaroe Point where the limestone pavement occurs. However, there is a second record for 'Crevinish Castle', dated 1950, so it appears safe to assume that these details refer to two separate stations. The first record is listed as having been determined by N.Y. Sandwith and a voucher very probably exists somewhere, most likely at K where Desmond Meikle worked.
Paul Hackney found the only other Fermanagh record of this taxon in September 1974, the station being given as, "below bridge on abandoned railway embankment near Lough Bresk, near Kesh". There are vouchers in BEL (Accession Nos H5546 (two sheets) and H5555). This plant was determined as R. afzeliana Fr. by R. Melville in 1980.
Glaucous Dog-rose can be distinguished from other Dog-roses by its caesious leaflets (ie of a lavender grey-green colour), being almost hairless and blue-green beneath – certainly far from tomentose. The leaflets are also often folded along the length of their midrib, which exposes the arching stems more fully to the sun, resulting in an even stronger wine-red stem coloration in comparison with subsp. caesia (Roses Handbook).
There are only five other Irish stations for R. caesia subsp. vosagiaca shown in the map of this subspecies in the BSBI Roses Handbook, two of which are pre-1950. The stations stretch from N Tyrone (H36) and Co Antrim (H39) in Ulster, southwards to Co Wicklow (H20). The New Atlas hectad map improves very slightly on this representation, with further stations in coastal Cos Sligo (H28) and Antrim (H39).
Growth form and preferred habitats
A small, much branched, usually pale greyish-green, summer annual therophyte up to 10 cm tall, with digitately or palmately lobed leaves, this sexually reproducing form of Aphanes is very much more local than the apomictic A. arvensis (Parsley-piert). It is greener and more slender than the latter, flowers from April to October and has even smaller fruits than A. arvensis s.s., showing no constriction between the upper and lower parts, and the sepals are convergent (New Flora of the BI 2019, p. 277, Figs 1, 2). It is less common than A. arvensis s.s. and appears to be confined to short turf, mossy areas on acidic sandy or gravelly soils, or dry rocky ground, eg on roadsides, along tracks and in quarries and sand-pits. It is not as confined to well-drained soils as A. arvensis, but is more definitely a plant of acidic conditions (Garrard & Streeter 1983). In Fermanagh, these conditions are also found locally on or near lakeshores and on river banks.
Fermanagh occurrence
The limited number of records that have accumulated in the Fermanagh Flora Database (18 finds in 14 tetrads) are mainly the work of RHN, facts that strongly suggest this rather insignificant-looking little species is under-recorded. Otherwise, as the tetrad map shows, it appears to be very local around the Tempo area, with very few (six or seven) records elsewhere in the county.
Additional to the first record are the following: Poll Beg District, NW of Boho, 11 June 1978, M.J.P. Scannell, DBN; all the remaining records involve RHN – Knockennis, 3 km NE of Brougher Mountain, 7 July 1988; Pubble Bridge, Tempo River, 1 October 1988; fen at Feddan Bog, 8 June 1992; fields at Largy Lough, 13 August 1992; sand pit at Pubble Bridge, 11 September 1994 & 20 August 1999, with RSF; Drumcreen, Ballinamallard River, 16 April 1995; roadside Ballyreagh, 5 km NW of Tempo, 31 December 1995, with HJN; roadside at Tempo, 13 April 1996; Tully, W of Edenmore, 21 June 1997, with RSF; N of Many Burns Bridge, Many Burns River, 3 May 1999; Pubble Forest, 1 December 2001; sandpit at Pubble Bridge, 28 August 2004, with RSF; Agnaglack, 20 April 2009, with HJN; Gublusk Bay, Lower Lough Erne, 28 February 2010; Killyreagh House near Tamlagh, 10 June 2010.
British and Irish occurrence
The New Atlas map shows the species widespread throughout both islands, but much more thinly scattered in Ireland, yet with a slightly greater presence in the south and the sunny south-east corner, both areas which attract more visitors and where the local recorders are more energetic than the norm (Preston et al. 2002). Thus the mapped distribution of A. australis across the whole of Ireland suggests probable under-recording in comparison with the situation in Britain.
European and world occurrence
It stretches northwards from a scattered presence in Spain and Portugal through W Europe to S Sweden and eastwards to NE Poland, the Carpathians and the Adriatic. Very local in the Balkans and present only on the W Mediterranean Isles (Minorca, Mallorca, Sardinia and Sicily), but also recorded on Madeira (Press & Short 1994; Sell & Murrell 2014). Beyond Europe, it occurs in Morocco and Algeria and is introduced in E & S parts of N America (Hultén & Fries 1986, Map 1156).
Threats
None.