DISTRIBUTION, ECOLOGY AND CONSERVATION STATUS OF Trichonzaizes
sl2eciosum WILLD. (PTERIDOPHYTA) IN THE AZOREAN ARCHIPELAGO
FREDERICK J. RUMSEY, JOHANNES C. VOGEL & MARY GIBBY
RUMSEY,F. J., J. C. VOCEL, & M. GIBBY2000. Distribution, ecology and
Willd. (Pteridophyta) in the
conservation status of Trichomanes specios~~m
Life and Marine Sciences. Supplement 2(Part
Azorean archipelago. Arq~~il2e'lugo.
A): 1 1- 18. Ponta Delgada. ISSN 0873-4704.
Trichor71unes speciosurii Willd., a globally rare and threatened pteridophytc, is most
abundant in the Azorean archipelago. Even here it has undoubledly declined through
habitat destruction. Thc species is unique amongst European-Macaronesian l'erns in that
over much of its range the life-cycle is not completed. Whereas the species relies almost
exclusively on asexual means of propagation throughout northern and central Europe, by
gametophytic gemmac, a norn~al sexual alternation of generations is the predominant
pattern of its reproductive biology in thc Azores. Gametophyte populations existing in the
absence of the sporophyte have been detected in the Azores but do not show the relative
abundance witnessed in northern and central Europe. The distribution of both generations
is given and their ecology discussed. Populations on man-made. i.e. recent, habitats on the
island of Flores indicate colonisation, hence dispersal, at a local level. Although the species
must bc considered not threatened in the Azores as a wholc, Azorean populations merit
conservation as they provide thc key to understanding thc factors controlling the species
reproduction. This information is vital for successful conservation elsewhere.
Frederick J. Rumsey (e-mail: fir-@r~/~~n.ac.uk),
.lohannrs C. Vogel & Mary Gibby,
Corzsr~rvufionBiology Laborator?, Cryptogarrlic Plant Rcsecirch Divisiorl, Department of
Botany, Nc~turalHistory Muse~or~.
Cromwell Roatl, Lonrlor~,SW7 5BD, U.K.
INTRODUCTION
1961). reported from Asia, the Caribbean and
South America. Morphologically similar taxa in
the section Lacosteopsis Prantl (= Vandenboschin
Trichornnnes
specios~~nl Willd.
(syn. Copel.) also occur in tropical Africa, North
Vnndenboschia specios~z (Willd.) Kunkel). the America and Australasia.
'The Killarney fern has been considered one of
Killarney fern, is the sole native European
representative of the genus Trichomnrles (TUTIN Europe's most vulnerable species threatened by
et al. 1993). the largest of the genera of the filtny- habitat destruction and a victim of over-collection
ferns (Hytnenophyllaceae). This tnorphologically (WALTER& GILLET1998). It has been accorded
reduced and very distinctive fatnily achieves legal protection throughout its range under the
greatest abundance in the cloud forests of the Bern Convention (ANON.1979) and the Habitats
tropics and sub-tropics. more rarely extending to Directive (ANON.1992).
Trichomanes speciosurn is found on all of the
constantly humid areas in the temperate zones.
Azorean islands with the exception of Santa
Trichonlunes
sl~eciosurn is
tetraploid
1993). First reported
(MANTON1950), its closest affinities and possible Maria (HANSEN& SUNDING
diploid progenitors have still to be resolved. It is by SEUBERT& HOCHSTETTER(1843) and later
a Macaronesian - European endemic (TUTINet al. given by DROUET(1866) as occurring on Faial
1993), that was formerly confused with T. and Flores, WATSON(1 870) recorded its presence
radicarls Sw. (see for example DANSEREAU on five of the islands, those previously listed and
SZo Miguel, Terceira and Pico, with TRELEASE moderated climate, such as the Azores.
(1897) further adding Corvo. The presence of the Accordingly, field surveys were performed in
species on SZO Jorge was noted by PALHINHA April 1994 (Flores. Terceira, Faial, Pico),
(1966) in his catalogue of the vascular plants of October 1996 (Flores, Terceira, Pico) and
the Azores. Discovery on Graciosa, the least September 1998 (Corvo, Flores, Terceira, SZo
elevated and driest of the islands was Miguel).
comparatively recent and supported by a
specimen at AZU (Poqo do Ratinho, Guadalupe,
DISTRIBUTION
Graciosa, alt. 60 m. 31.08.1995. F. Pereira 125!).
SJOGREN
(1973) was the first to present maps
of the species distribution within the Aqores, The distribution as currently known on all of the
plotting both his field records and those from islands for which this species is recorded, with the
the exception of Graciosa, is documented in Figs. 1-6.
localised literature and herbarium sources f o ~
islands of S. Miguel, Pico, Terceira and Faial. Distribution has been mapped on a 2 km grid
and RASBACH(1973) level for all islands, based on the UTM grid.
Independently WILMANNS
produced a map of the species occurrence on the These maps have been derived from literature
first of these.
sources (SJOGREN1973, 1978, 1979, 1997;
Trichomrrnes speciosu~n is unique amongst WARD 1970; WILMANS& RASBACH 1973),
European ferns in that its gametophytic herbarium material (BM, K, AZU) and our own
generation, the sexual or gamete-bearing phase of field surveys (1 994- 1998).
the life-cycle, is perennial, gemmiferous and
No comprehensive systematic account of the
capable of persisting and dispersing in the distribution of the pteridophyte tlora of the
absence of its sporophyte. This latter ability, Azorean archipelago has been published. It is thus
'gametophytic independence', is apparently a very difficult to assess the true abundance of species,
rare condition amongst the homosporous ferns. It the likely factors affecting their distribution and
has been reported from three families, all whether they are declining, or increasing in
widespread as epiphytes in tropical regions, the numbers and range. Those distribution maps
Grammitidaceae,
Vittariaceae
and
the previously given for T. speciosum do not cover
Hymenophyllaceae, but is only well developed in the generally under-recorded western islands
those rare species found at the temperate extremes where the species is most abundant. Across the
of these families ranges (RUMSEY& SHEFFIELD archipelago the percentage of grid cells per island
1996). The discovery of the distinctive but easily
in which the fern occurs rises from 0-5.7 for the
overlooked gametophyte generation of T. eastern group, to 8.2- 10. I for the central to 42.9specios~l~n
in northern Europe (RUMSEYet al. 63.6 for the western group. SJOGREN
(1973) lists
1990) post-dates the few floristic accounts only four localities from Flores, WARD (1970)
concerned with the distribution and ecology of mentions a further two but states that the species
this (and other pteridophyte) species in the is "common in damp shaded places". This is
Azorean archipelago, (e.g. SJOGREN1973, 1979; probably somewhat of an exaggeration, although
& RASBACH1973). In
W A R D 1970; WILMANNS
it is certainly widespread (Fig.6) and is locally
view of the relative abundance of the gametophyte very abundant. Continuous sporophyte colonies of
generation over much of north-western and greater than 100 m in area have been seen in
central Europe (see RUMSEY, JERMY & several locations, eg. Ribeira da Silva (fig. 7). On
SHEFFIELD 1998), where climate has been Flores it occurs from near sea level (15m Sta.
conjectured to be the prime factor effecting Cruz, Gonqalves. LISI) to c. 820 m on the
western slopes of the highest peak, Morro Alto
'gametophytic independence', it was clearly
desirable to establish the extent to which (914 m). For those islands for which maps andlor
gametophytes occurred, and where, in areas with descriptive accounts (e.g. Wilmanns and Rasbach,
abundant sporophytic growth and a more
1973) have been published our brief surveys have
'
Fig. 1. Distribution of Trichomatzes sprcioswn Willd. on Siio Miguel. Earlier records that we have confirmed are
given as e, earlier unconfirmed reports as*. Where gametophytes alone occur in a tetrad they are represented by an
open symbol, o.
Fig. 2. Distribution of Trichornatzes speciosunz Willd. on Terceira.
increased the known distribution. The species
detection from a range of apparently previously
unrecorded sites is in part due to the hitherto
overlooked gametophyte generation but the
majority relate to new records of the sporophyte.
We conclude that detailed survey would reveal
the species in many more localities.
Gametophytes are not always apparent in close
proximity to sporophyte colonies but in many
such cases are present elsewhere within the
locality. The very extensive mats of gametophyte
found in sheltered caverns in northern and central
Europe have only been seen in a few lowland
gulley sites on Flores, Corvo and S. Miguel. In
most sites gametophytes are present in small
quantities (patches with an area of < 10 cm2),
although juvenile sporophytes are often produced.
Sporophytic recruitment has been noted in 19
sites distributed throughout the archipelago.
Gametophytes and juvenile sporophytes are
always associated with rocky substrates.
Sporophyte colonies present on tree boles. humic
banks, etc. we consider to have secondarily
colonised such habitats via the growth of the long
creeping rhizome.
Levels of sporophyte fertility fluctuate
annually, presumably linked to climatic variables.
The frequency with which colonies become
fertile, the proportion of fertile fronds per colony
and numbers of sporangia per frond varies greatly
throughout the species range. At its northern
limits in Great Britain the majority of colonies
have never been recorded to produce spores
(RATCLIFFE et a]. 1993). Even within the Azores
Fig. 3.
Fig. 4. Distribution of TricJ~nmnr~es
speciosirrn Willd. on Faial.
there exists a marked difference in fertility
between the usually highly fertile western island
populations and those in the central and eastern
groups. The precise factors affecting fertility
remain to be resolved.
ECOLOGICAL
AND
PHYTOSOC1OLOGICAL
OBSERVATIONS
SJOGREN (1973) considered
Trichomnr7e.s
specios~lmone of the best ecological indicator
species for habitats with the highest relative
humidity and regarded it phytosociologicall y as a
differential of his Festucetutn jubatae association,
a community rich in endemic taxa characterised
by their low drought tolerance and need for
shelter. The species occurred less regularly in the
Erico-Myrsirretum, the most widespread natural
vegetation of the islands. These associations
together comprise the natural component of the
Juniperion brevifolii alliance, the evergreen
(scrub-) woodland vegetation of middle to upper
altitudinal range (>500 m), i.e. the cloud-zone.
Below
this
altitude
occur
the
most
anthropogenically affected woodlands, now often
dominated by the alien Pittosporum und~llarum
Vent. These are predominantly found
in the altitude range 300-600 m. and have
effectively replaced the natural transitions
between the coastal Myricn dominated scrub
( S J ~ R E N '(1s 973) Festuciotl petroene alliance)
with the species-rich mixed laurus-silva of
moderate altitudes. SJOGREN(1973) conceded
that transitional zones between these communities
are particularly prevalent on the damper western
Fig. 5. Distribution of Trichotnanes speciosurn Willd. on Pico
Fig. 6. Distribution of Trichott~arlesspeciosum Willd.
on Flores and Corvo.
islands, where the altitudinal zonation is in any
case more compressed. He further suggested that
by virtue of its dense canopy the spread of
Pittosporurn woodland has favoured the
colonisation of a range of species associated with
the Juniperiorz brevifolii below their original
lower altitude limit throughout the archipelago.
SJOGREN(1984) described T. speciosum as
generally occurring above 500 rn, although fairly
frequent in the western group between 200 and
500 m, i.e. essentially restricted to the cloud zone
region.
Is this categorisation of the species as nonlowland accurate and if T. speciosum occurs in
these areas is it the result of a recent extension
downwards in altitudinal range? Our survey
revealed T. speciosum to be present in shaded,
deep streambeds with waterfalls, in the lower
level woodlands on all of the island groups. We
have found both generations as low as 190 m on
S. Miguel and at 90 m on Flores, although within
the central group few localities have been found
to occur below 400 m. The species and especially
its gametophyte are undoubtedly under-recorded
in these streambed gulley sites, which in most
cases are difficult to survey. Was T. speciosum
present at these altitudes prior to the arrival of
Pittosporum undulatum, or is it a comparatively
recent colonist? Regular associates in these
situations include Diplaziilrn cautintun~ (Cav.)
Jermy and Pteris incompleta Cav. both of which
occur as scattered individuals in sheltered cloudzone sites but which never occur with the same
abundance or luxuriance as at these lower levels.
It is also noteworthy that this assemblage is found
between 250 and 500 m in the Macaronesian
vegetational enclave in the drier Algeciras area of
southern Spain (RUMSEY& VOGEL 1998). Other
pteridophytes showing habitat preferences for
lowland woodland sites include Asplenium
hemiotiitis L. and A. aeoricum Lovis, Rasbach
Trichomarzes specios~~rn
is adapted to
life in a specialised habitat. This
narrow ecological amplitude closely
limits the number of potential sites,
especially for the species sporophyte
but such suitable sites occur very
patchily over a wide geographical
area. As an attractive plant, present in
small numbers i n many west
European states, it was a prime
candidate for legislative protection.
Indeed, this species has taken on an
almost
iconic
status
amongst
conservationists (PAGE1997) and yet,
following field and laboratory
organism
investigation
of
the
throughout its life-cycle, the necessity
for active conservation is being reassessed. Until 1990 and the first
report of the garnetophyte generation
in natural situations (RUMSEYet al.
1990) it was widely considered that
this species was among the most
vulnerable and threatened of Europe's
plant taxa. Populations over much of
its range were not only small but
static, with no observed regeneration,
or colonisation of novel areas
(RATCLIFFEet al. 1993). This was
coupled with a perceived threat
through
- collection, a major cause of
Fig. 7 . Vc1-y cxlensivc colony of Tric1loii1ciric.sspeciosu~rl,gulley ;It c. decline in Ireland in the last century.
330 m alt., Ribeira d a Silva, Flores
While the situation was believed to be
less bleak in Macaronesia, where
and Reichstein. While the presence of these fern large populations were known to exist in northern
asseniblages provide support for the view that T. Madeira and in the Azores, the continuing
speciosum could be of natural occurrence at low destruction of the "Lausus-silva" and the
altitudes, the species has certainly also colonised alteration of habitats by invasive aliens gave
areas within the cloud-zone during this time-span.
considerable cause for concern. The discovery of
This is evidenced by its presence on lava fields of a widespread gametophyte distribution, akin to a
known age, e.g. the Misterios Negros of Terceira vast viable seed-bank in angiosperm terms, has
and various of the lava fields on Pico. Similarly, somewhat diminished the claim for the species to
T. speciosun~ is widespread and perhaps most priority conservation attention (see RUMSEY
et al.,
frequently encountered on Flores as a plant of in press, for discussion). Within Macaronesia the
walls, usually where shaded by the alien gametophyte is not an obvious component of the
H)ufrarzgea ~nacroplzylla(Thunb.) Ser.
vegetation when compared with the sporophyte, in
contrast to the situation in northern and central
Europe. It does, however, occur in sites from
which the sporophyte is currently absent,
particularly those at lower altitudes. T h e
gametophyte thus extends the species distribution
and may harbour genetic variation not currently
present within extant sporophyte populations, i f
genetic diversity is maintained between, not
within populations, as in northern and central
Europe ( R U M S E Y et al. 1998). T h e high
proportion o f sites present on man-made
structures within the western group in the Azores
indicates the species ability to colonise novel
habitats when they become available. Man-made
sites are, however, also vulnerable to destruction
and are less likely t o support populations over the
long term than larger natural rock features.
Conservation actions that seek to protect a
mixture o f "natural" source populations and more
transient but perhaps more genetically diverse
man-made sites would be o f benefit to the survival
o f the species. T h e ability to persist as the
gametophyte generation in degraded habitats
renders it less vulnerable than most species in this
hygrophilous community to habitat destruction.
A n ability to inbreed, to produce potentially
widely dispersible propagules and to propagate
vegetatively may also overcome problems o f
habitat fragmentation ( R U M S E Yet al. in press).
These aspects o f the species biology, together
with the abundance o f sites on the western islands
o f Flores and Corvo, suggest that the species must
be considered Not Threatened within the
archipelago as a whole. However, i f considered
separately the islands would individually warrant
a higher risk status. Further work is still necessary
to establish extent and status on several o f the
islands.
For conservation action to succeed it must be
grounded on a rigourous scientific basis, with a
firm understanding o f the biology o f the species
involved. T h e work presented here suggests that
the Azores does maintain globally significant
populations o f T. speciosum. However, i f
conservation action is promoted on a "species"
not "habitat" basis then other Azorean species
must be considered more threatened and requiring
more immediate action. T h e maintenance o f the
unique vegetation within which Trichoinarles
speciosum occurs is unquestionably o f paramount
importance as it supports many endemic taxa
whose continued survival is now o f very real
concern. T h e species itself acts as the preferred
substrate for a range o f epiphyllous hepatics many
o f which are almost exclusively confined to this
1997). T h e presence
habitat in Europe ( S J O G R E N
o f this species can thus be used as an indicator o f
a particular cryptogainic coinmunity o f
considerable conservation importance. T h e
continued legal protection afforded to this fern
thus by default acts t o protect a wide range o f less
obvious but equally important taxa.
ACKNOWLEDGEMENTS
W e wish to thank the Natural Environment
Research Council who funded our research into
this species under grant no. GR3109451 and the
Direc~iioRegional de Ambiente for permission t o
collect material. W e also wish to thank S e r v i ~ o s
Florestais, Eduardo Dias and Helen Martins for
logistic and other help, Alison Paul, Bob Press
and Clive Jermy for assistance in the field and t o
the latter for provision o f his earlier unpublished
records. W e also wish to thank Malcolm Penn for
his great assistance in the preparation o f the
distribution maps.
REFERENCES
ANON. 1979. The convention on the conservalion o f
Europcan wildlife and nalural habitats. Strasbourg,
Council o f Europe.
A N O N 1992.
.
Council directive 92/43/EEC o f 21 May
I992 on the conservation o f nalural habitats and o f
wild faunas and floras. OfSicinl Journal of the
Europearl Comrnu~~ities
206: 7-49.
DANSEREAU,
P. 1961. ~ t u d e s macaronCsiennes. 1 .
GCographie des
cryptogames
vasculaires.
Agronomia Lusitarzica 23: 15 1 - 18 1.
DROUET,H . 1866. Catalogue dc la flore des iles
A~ores,prCcCdC de 1' itinCraire d'un voyage dans
cet archipel. Mkrnoires rle In Socie't; ncnrlen~iq~~e
rl'agricult~ire, Science, Arts et belles lettres
rlkpnrtement tle I'A~lbe3: 8 1-233.
~/LL
HANSEN,
A. & P. SUNDING
1993. Flora of Macaronesia.
Checklist of vascular plants (4th edition).
Sotnnzerfeltin 17: 1-295.
MANTON,
I. 1950. Problenls in cytology arzd evolrltiorz
irz the Ptericlophyta. Cambridge University Press,
Cambridge. 3 16 pp.
PAGE,C. N. 1997. The ferrzs of Brirair7 arzd Irelarzrl
(2nd edition). Cambridge University Press.
Cambridge. 540 pp.
PALHINHA.R. T. 1966. Cc~tcilogo rlas Plar~tas
Vasculares (10s Acores. Lisboa. 186 pp.
RATCLIFFE,
D. A,, H. J. B. BIRKS,& H. H. BIRKS1993.
The ecology and conservation of the Killarney fern
Trichornnnes s/7eciosurn Willd. in Britain and
Ireland. Biologicnl Conservatiorz 66: 23 1-247.
& E. SHEFFIELD
1998. The
RUMSEY.
F. J., A. C. JERMY
independent gametophytic stage of Trichomanes
sl,eciosurn
Willd. (Hymenophyllaceae), the
Killarney Fern and its distribution in the British
Isles. Wnrsor~ia22: 1-19.
& D. R. FARRAR1990.
RUMSEY,F. J., E. SHEFFIELD
British filmy-fern garnetophytes. Pteridologist 2:
40-42.
RUMSEY.
F. J. & E. SHEFFIELD
1996. Inter-generational
ecological niche separation and the "independent
gametophyte" phenomenon. Pp. 563-570 in:
CAMUS,J . M., M. GIBBY& R. J. JOHNS(Eds)
Pterirlology irz perspecrive. Royal Botanic Gardens,
Kew. 700 pp.
RUMSEY,F. J. & J. C. VOGEL 1998, Trichonzar~es
specios~imWilld. (Hymenophyllaceae) in southern
Spain. Fern Gazette 15: 197-203.
RUMSEY: F. J., J . C. VOGEL,S. J. RUSSELL.J. A.
BARRETr & M. GIBBY1998. Climate, Colonisation
and Celibacy: Populalion Structure in Central
European Trichomaiz~sspeciosum (Pteridophyta).
Botnr?icn Acta 1 I I : 48 1-489.
RUMSEY.F. J., J. C. VOGEL,S. J. RUSSELL,J. A.
BARRETT& M. GIBBY (in press). Population
structure and conservation biology of the
cndangered fern Trichornanes speciosuni Willd.
(Hymenophyllaceae) at its northern distributional
limit. Biological Jo~rrrzalof the Lirzrzearz Society 66.
1843. Uebersicht der
SEUBERT,
M. & C. HOCHSTETTER
Flora der Azorischen Inseln. Archiv fiir
Naturgeschiclrte 9: 1-24.
SJOGREN,
E. 1973. Recent changes in the vascular flora
and vegetation of the Azores Islands. Merndrias (la
Sociedarle Broteriarzn 22: 1-453.
SJOGREN,
E. 1978. Bryophyte vegetation in the Azores
Islands. Mer~drinsrln Socierlarle Broterinrzn 26: 5283.
SJOGREN, E. 1979. Contributions to the vascular flora
and vegetation of the island of Corvo (Azores).
Boletin1 rlo M ~ i s e ~Murzicipal
i
rlo Furzchnl 32: 1987.
SJOGREN,
E. 1984. Acores Flores. Direc~HoRegional
de Turismo, Horta.
SJOGREN,E. 1997. Epiphyllous bryophytes in the
Azores Islands. ArquipPlago. Life and Marine
Sciences 15A: 1-49.
TRELEASE,
W. 1897. Botanical observations on the
Azores. Arzrzunl Re110r.r of the Misso~iriBotarlical
Garderl 8: 77-220.
TUTIN:T. G., N. A. BURGES,A. 0. CHATER.J. R.
D. M. MOORE,D.
EDMONDSON.
V. H. HEYWOOD.
H. VALENTINE,S. M. WALTERS& D. A. WEBB
1993. Flora E~lrol~aeaVol. 1 , (2nd. Ed.)
Cambridge University Press. Cambridge. 58 1 pp.
WALTER:K. S. & H. J. GILLETT1998. IUCN Red List
of threatened plnrrts of the ~ ~ o r l r lNew
.
York
Botanical Garden, New York. 928 pp.
WARD, C. M. 1970. The Pteridophytes of Flores
(Azores): a survey with bibliography. British Fern
Gazette 10: 1 19- 126.
WATSON,H. C. 1870. Botany of the Azores. Pp. 1 13288 in GODMAN.
F. DU C. (Ed.) Nat~cralHistory of
rile Azores or Westerrz Islnrzrls.Van Voorst,
London.
0. & H. RASBACH
1973. Observations on
WILMANNS,
the pteridophytes of SHo Miguel, Azores. British
Fern Gazerre 10: 3 15-329.