J Hallori Bot. Lab. No. 97: 309- 3J6 (Jan. 2(05)

MODELLING THE DISTRIBUTION OF MlRABILIS MALMB. (, HEPATICOPSIDA) IN THE IBERIAN PENINSULA

CECIUA SERGIO', DAVID DRAPER2 AND CE:SAR GARCIA'

ABSTRACT . Cryptothallus mirabilis Malmb. exhibits a specialised ecology and is a non-photosyn­ thetic Iiverwort. This species was first described for Scandinavia and till now was considered a north oceanic species. The discoveries of C. mirabilis in different localities in Portugal are the first reliable records of the species in southern Europe and represent an extension of its geographical range. [n fact, in Portugal C. mirabilis does not seem to be rare in areas with oceanic influence, mainly in wet forestry habitats. It is possibly widespread in the Iberian Peninsula because it is easily overlooked owing to its subterranean growth. The objective of this work is to give new localities which con­ tribute to an evaluation of its ecological requirements, to the definition of its biogeographical distrib­ ution in Portugal, and to present a predictive map for the Iberian Peninsula. The aim of the model is to identify the distribution pattern and also some potential places of occurrence in Spain and provide information about the environmental range of the species in to improve actions for conserva­ tion. K EY WORDS: Cryptothallus mirabilis, ecology, predictive map, conservation, Iberian Peninsula

INTRODUCTION The discoveries of Cryptothallus mirabilis Malmb. in Portugal (Sergio & Seneca 1997, Sergio & Garcia 1999) are the first reliable records of the species in southern Europe and represent an important extension of its geographic range. At that time, owing to our limited experience of the ecology of this species, the dis­ covery of C. mirabilis was by chance. In the last few years we have again found C. mirabilis in 8 new localities during a survey of material of Hypnum cupressiforme Hedw. used for heavy metal biomonitorization. So we became convinced that it was indeed not a rare species in Portugal and may be present in other localities as well as in the northern part of the Iberian Peninsula. In fact C. mirabilis was considered to be a Boreal species in Eu­ rope, relatively rare but with a wide and patchy distribution: Greenland (Petersen 1972), Scandinavia (Dickson et al. 1975), Germany (Wiehle et al. 1989) in Britain (Crundwell in Hill et al. 1991, Paton 1999) and more recently France (Bates & Hodgetts 1995 and Boudi­ er et al. 1999). The specialised ecology exhibited by C. mirabilis is exceptional among liverworts. It is known that the thalli of this non-photosynthetic liverwort are colonized by fungal hyphae (Ligrone et al. 1993). It was recently demonstrated by Bidartondo et al. (2003) that Cryp-

I Universidade de Lisboa. Museu de Histria Natural, lardim Botnico/Centro de Ecologia e Biolo­ gia Vegetal, Rua da Escola Politecnica 58,1250- 102, Lisboa, Portugal. E-mail: [email protected] 2 Universidade de Lisboa. Museu de Histria Natural, lardim Botanico, Rua da Escola Politecnica 58, 1250- 102, Lisboa, Portugal. E-mail: [email protected] 310 1. Hattori Bot. Lab. No. 97 2 0 0 5

tothallus associates with fungi of the Tulasnella and the same fungi have the ability to form ectomycorrhizas on different trees species, for example in Pin us and Betula. Bidar­ tondo et a!. (2003) for the first time confirmed that the source of carbon, which sustains the liverwort, is the mycelium that provides a pathway for transferring it into the thalli. The objective of this work is to give the new localities where it was found so as to increase the definition of its biogeographical distribution in Portugal, as well as to present a first ap­ proach to a predictive map for the Iberian Peninsula. The aim of the model is to identify possible areas of occurrence in Portugal and Spain and to provide information about macro-environmental requirements of a species with a significant nature conservation im­ portance.

Why is it important to know the distribution and ecological requirements of Cryptothallus mirabilis? Recent research on the evolution of land and advances in DNA studies have led to the realisation of the importance of . One crucial species is Cly ptothallus mirabilis because of its ecology, its primitive characters as well as its mode of nutrition and the phylogeny process. Also it is important through biochemical aspects (Rycroft & Cole 1998). In Europe C. mirabilis was considered to be a Boreal species, but in north and cen­ tral Europe it is apparently confined to mires and bogs, under , and associated with Molinia litter, and is easily overlooked because of its subterranean growth (Schuster, 1992). In Portugal it seems more usual that C. mirabilis can develop below mats of Hyp­ num cupressiforme. The absence of Sphagnum species in most of its Portuguese localities may reflect some important differences between its ecology in the northern and the south­ ern areas where it occurs. Also some questions can be asked about the origin of this species in these meridional areas, which no doubt have certain Mediterranean characteristics. In boreal areas it is apparently confined to mires and bogs in vegetation related to the associations Salicion cinereae, Va ccinio-Betulefum , and sometimes Piceion (Dierf3en 200 I). In France is reported for the A ln etea glutinosae and more particularly in the alliance Alnion glutinosae under colonies of Sphagnum gr. recurvum P. Beauv., associated with S. angustifolium (Russow) C. o. E. Jensen and S. fallax (H. Klinggr.) H. Klinggr. (Boudier et al. 1999). As we have mentioned, in Portugal C. mirabilis seems not to be rare in areas with an oceanic influence mainly in wet woodland habitats, and can be widespread there. In fact, with our present knowledge, Portugal is evidently the southern end of its range but perhaps the species can occur in other intermediate areas.

Ecological and chorological data in Portugal Some important differences in the ecology of the species, between the northern and the southern areas of Europe, are evident. The ecological differences of the two areas are obvious but it is important to take into account the differences of climate and vegetation configuration. In the Portuguese territory two main ecological conditions can be described. I. Littoral areas in Quaternary deposits - Holocene (Beira Litoral: Aveiro - the five first localities in the list). C. SERGIO ET AL. : Cryplolhallus mirabilis in the Iberian Peninsula 311

The majority of localities are in an area of the terminal part of Vouga and Mondego rivers, with a fluvio-Iagunar Holocenic origin and a strong oceanic influence, and the plants produce abundant . These localities have different ecological conditions but in general correspond to Pin us pinaster Aiton plantations with very interesting semi-natural scrub vegetation always with Corema album (L.) D. Don and Calluna vulgaris (L.) Hull and sometimes Myrica gale L. and Stauracanthus lusitanicus (L.) Cubas. The ground of the forested areas is very wet, with Scirpoides holoschoenus (L.) Sojak, Erica ciliaris Loefl. ex L. and is more or less completely covered (80- 95%) with bryophytes and lichens (Hypnum cupressiforme ± 80% of total cover associated with Di­ cranum scoparium Hedw. , Campylopus flexuosus (Hedw.) Brid., Polytrichum juniperinum Hedw. and Cladonia rangiformis Hoffm.). 2. Northern and Eastern areas in complex schist-gaywackes - Silurian and Ordovi­ cian (Minho, Beira Alta, Beira Baixa and Alto Alentejo - the four last localities in the list). As in the 5 first localities C. mirabilis grows on acid soil in Pinus pinaster forest in heath land humid formations with high cover (60- 90%): Hypnum cupressiforme, Sphagnum denticulatum Brid., Dicranum scoparium, Scleropodium purum (Hedw.) Limpr., Polytrichumjuniperinum. The shrub vegetation comprises Pteridium aquilinum (L.) Kuhn, Erica ciliaris L., Erica tetralix L. fil. , Ulex minor Roth, Calluna vulgaris (L.) Hull, Pterospartium tridentatum (L.) Willk., or in drier conditions with Erica umbellata L. , Hal­ imium ocymoides (Lam.) Willk. and, in some localities, different species of Cistus.

Fertility In the majority of Portuguese localities the species is fertile, with male and female plants (Fig. 1) and the sporophytes and spores were observed in 5 localities. Some pub­ lished data indicate the presence of sporophytes in some areas (Pocock & Duckett 1984) and these seem to be common (Crundwell in Hill et al. 1991 , Paton 1999, Boudier et al. 1999). The mature spores appear in Portugal in January to March but seem to be earlier in France (March); but both are different from northern sites (in summer, Hallingback pers.com.). Unfortunately, we had no opportunity to examine fertile material from Euro­ pean herbaria (H, PC, BM, UPS and C). However, the spores represented by Schuster (1992) are apparently not different from Portuguese specimens. The capsule has a very par­ ticular dehiscence by 4 lines, little twisted, and the valves are coherent distally. So this opening structure is very interesting and the tendency to be spiralled presents some similar­ ity to that in Takakia (Higuchi & Zhang 1998).

Conservation approaches The conservation of a particular species begins with recording its distribution and studying its bioclimatic affinities and environmental requirements. On the other hand the establishment of correlations between distribution patterns and environmental factors could be an appropriate tool for management and conservation programmes (Begon et al. 1990). In a further step, if we know the relationships between the organism and the environmental variable we can predict its distribution (Johnston 1993). 312 1. Hattori Bot. Lab. No. 97 2 0 0 5

Fig. 1. Fertile pl ants of Cfyptothallus mirabilis Malmb. A. Female plants with peri­ anths and immature sporophytes; B. Male plants; C. Mature sporophytes growing up from a carpet of Hypnum cupressiforme Hedw.; D. First stage of a capsul e opening by 4 twisted lines with dehiscent spores. Material from: Aveiro, Ga fanha da EncarnaGao, Sergio, 2003 (L1SU) . Scale: A =8 mm; B=8 mm; C=6 mm; D= 0. 8 mm. C. SERGIO ET AL.: Oyptothallus mirabilis in the Iberi an Peninsul a 313

·0.3 ·OA ·0.5 - 0.6 ·0.7 ·0.8 ·0.9 ·1

Fig. 2. Distribution map of Cryptothallus mirabilis Malmb. in the Iberian Peninsula (Red dots of occurrence in UTM 10 X 10 Km). Predictive map obtained by a logistic regres­ sion (LR) method. The darkest areas have the highest values of occurrence probability.

Our knowledge of the species' distribution is in some cases extremely poor, particu­ larly in the Iberian Peninsula, and the need for the implementation of a new advance has become urgent. Predictive maps present a new approach in defining the distribution of taxa and they circumvent the lack of distribution data. These predictive maps are commonly ap­ plied to other species groups such as vascular plants (Guisan & Theurillat 2000) or animals (Pereira & ltami 1987) as well as to other Iberian bryophytes (Sergio & Draper 2002, Drap­ er et al. 2003). A particular application of predictive maps in conservation action on endan­ gered bryophytes is the selection of micro-reserves (Draper et al. 2003) or landscape sur­ veys (Vanderpoorten et al. 2005).

Modelling distribution in the Iberian Peninsula The predictive method was logistic regression (LR) and after that a logistic transforma­ tion (Hill & Dominguez 1994; Draper et al. 2003). The variables used were: Altitude; Slope; Aspect; Latitude; Longitude; Radiation; Annual and monthly precipitation; Mean monthly temperature; Maximum and minimum temperature; Thermal amplitude; Angstrom Index; Angot Index; Dantin Index; Emberger Index; Giacobbe Index; Gorezynski Index; Lang Index; Martonne Index. 3 14 1. Hattori Bot. Lab. No. 97 2 0 0 5

Climatic data were calculated according to Sanchez Palomares et al. (1999); biocli­ matic indexes were based on Tuhkanen (I 980). The predictive maps of selected bryophytes in the Iberian Peninsula, based on ecoclimatic variables, were made using a Geographic In­ formation System (GIS). All the records known at present were used to calculate the eco­ logical models. From the initial set of 44 variables only 3 of them were significant at 95%: Gorezynski index (index of continentality); March precipitation; Mean April Temperature. The significance of the model was p=O.OOOOO I. The resulting model was: y =5.1831 -0.030767* Gorezynski -0.032 168 * March Rainfall + 0.035708* April Temperature Transformed: p=expY/(expY+ I) The geographic interpretati on of these results (Fig. 2) allows us to consider C. mirabilis as an oceanic species but occurs in areas with dry and softly warm springs. The area in Iberian Peninsul a that has these characteristics is the coastal north part of Portugal and the Galicia County. The next step could be the implementation of specific fieldwork in order to certify the predictive distribution of this li verwort and to try to find C. mirabilis in suitable areas in NW Spain . In conclusion, the applied methodology for predicting bryophyte distribution is a first step and can give suggestions for the identification of new places where it could be found. It also gives the possibility of analysing the relationships between levels of different envi­ ronmental variables and the occurrence of a species to determine the main ecological re­ quirements. If we can discover these relationships it will be much easier to find new popu­ lations. It could be a useful increment in our knowledge and in future may contribute in knowing the more important traits of habitats where the species could occur. If new locali ­ ti es were obtained, powerful analyses and methods could be applied and re-evaluated. No geological variables were used in thi s survey owing to the heterogeneity of classi­ fication between Portuguese and Spanish data. But considering that C. mirabilis is an aci­ dophytic species it could be important to include lithology or geological data in future sur­ veys. However, the predictive map (Fig. 2), the knowledge of the ecological conditions of the sampled area, the field experience and the features of the pl aces where the species was found (showing a high bryophytes cover of 60- 95%), can contribute to finding new popula­ tions of thi s liverwort in the Iberian Peninsula.

Specimens and localities (F) = fertile I. Beira Litoral: Ave iro, Gafanha da Encarnayao, ±20 rn , 29TNE2395, 1997, 1999, 2002, 2003, C. Sergio 10628, 12474 etc. (U SU). (F) 2. Beira Litoral. En tre Furadouro e Cortegaya, estrada para Furadouro, ±20 rn , 29T E2928, 2002 C. Sergio 12425 (USU). 3. Beira Litoral. Torrao, a Sui de Furadouro, ± 20 rn , 29TNE275 1, 2003 R. Figueira & J. Fer­ reira (LJ SU). (F) 4. Beira Litoral. Vague ira, pr. Deposito das Aguas, ± 20 rn , 29TNE2289, 2002, C. Sergio & J. Ferreira 12475 (LJSU). (F) 5. Beira Litoral. Quiaios, Lagoa das Brayas ±50 rn , 29TNE 1655, 2002, C. 5ergio & J. Fer­ reira 122 12 (LISU). (F) 6. Beira Alta: Viseu, S. l OaD da Lourosa, 400 rn, 29TNE9196, 1998, C. Sergio 11109 (U SU). 7. Beira Baixa: I km de Proenya a Nova, ± 500rn, 29TNE9 103 , 1997, C. Sergio & c. Gm'cia, C. SERG IO ET AL.: Cryptothallus mirabilis in the Iberian Peninsula 3 15

10930 (USU). (F) 8. Alto Alentejo: Portalegre, Monte Rei, 750m, 29SPD4150, 1995, C. Sergio & A. Seneca, 9727 (LJSU). 9. Minho. Ponte de Lima, Labrujo, 500 m, 29TNG3733, 2001, M. Sim -Sim & c. Garcia (USU).

ACKNOWLEDGEMENTS Funding to fieldworks was provided by the project of the F.c.r. 35618/99. We tank Alan Roy Perry for constructive comments on an earlier version of the manuscript.

REFERENCES Bates, J. W. & N. G. Hodgetts. 1995. New and interesting bryophyte records from Brittany including Cryptothallus mirabilis, Ulola calvescens and Weissia perssonii new to France. Cryptogamie, Bryo!. 16: 191 -211. Begon, M., 1. L. Harper & C. R. Townsend. 1990. Ecology. Individuals, populations and communi­ ties. Blackwell Scientific Publicatios, Boston. Bidartondo M., T. Bruns. M. Weiss, C. Sergio & D. Read. 2003. Specialized cheating of ectomycor­ rhizal symbiosis by an epiparasitic liverworts. Proc. Royal Soc., BioI. Sci. 270: 835- 842. Boudier, P. , 1. Bardat, & S. Perrera. 1999. Cryptothallus mirabilis v. Malmborg (Aneuraceae, Hepati­ copsida) dans le Perche d'Eure-et-Loir (France). Cryptogamie, Bryol. 20: 189- 196. Dickson, 1. H., T. Koponen & T. Ulvinen. 1975. Cryptothallus mirabilis Malmb. loydetty jiilleen Suomesta. Luonnon Tutkija 79: 53- 56. Dierf3en, K. 2001. Distribution, ecological amplitude and phytosociological characterization of Eu­ rope bryophytes. Bryophyt. Biblioth. 56: 1- 289. Berlin, Stuttgart. Draper, D. , A. RosseIl6-Graell, C. Garcia, C. Tauleigne Gomes & c. Sergio. 2003. Application of GIS in conservation programmes in Portugal. Biological Conservation 113: 337- 349. Guisan A. & J-P. Theurillat. 2000. Assessing alpine plant vulnerability to climate change: a modeling perspective. Integrated Assessment I: 307-320. Higuchi, M. & D.-C. Zhang. 1998. Sporophytes of Takakia ceratophylla found in China. J. Hattori Bot. Lab. 84: 57- 69. Hill, M . O. & F. Dominguez Lozano 1994. A numerical analysis of the distribution of liverworts in Great Britain. In: M. O. Hill, C. D. Preston & A. 1. E. Smith, Atlas of the bryophytes of Britain and Ireland. vol. 3, Mosses (Diplolepideae). Colchester. Hill, M. 0., C. D. Preston & A. 1. E. Smith. 1991. Atlas of the bryophytes of Britain and Ireland. vol. I Liverworts (Hepaticae and Anthocerotae). Colchester. Johnston, C. A. 1993. Introduction to quantitative methods and modelling in community, population, and landscape ecology. In : M. F. Goodchild, B. O. Parks & L. T. Steyaert (Eds.), Environmental modelling with GIS. Oxford Univ. Press, New York: pp. 276-283. Ligrone, R. , K. Pocock & 1. G. Duckett. 1993. A comparative ultrastructural study of endophytic ba­ sidiomycetes in the parasitic achlorophyllous hepatic Cryptothallus mirabilis and the closely al­ lied photosynthetic species pinguis (). Can. J. Bot. 71: 666-679. Paton,1. A. 1999. The liverwort flora of British Isles. Colchester. Pereira,1. M. C. & R. M. ltami. 1987. GIS-based habitat modeling using logistic multiple regression: a study of the Mount Graham red squirrel. Photogrammetric Engineering and Remote Sensing 57: 1475-1486. Petersen, P. M. 1972. Cryplolhallus mirabilis Malmb. found on Disko Island, West Greenland 316 J. Hattori Bot. Lab. No. 97 2 0 0 5

(69° 15'N .. 53 °34'W.). Lindbergia I: 189- 190. Pocock. K. & J. G. Duckett. 1984. A comparative ultrastructural analysis of the fungal endophytes in Crvptothal/lIs mirabilis Malmb. and other British thalloid hepatics. J. Bryol. 13: 227-233. Rycroft. D. S. & W. J. Cole. 1998. 15-Acetoxypinguisone and a cyclocuparanol form liverwort Cryp­ tothallus mirabilis Malmb. 1. Chemical Research, Synopses 9: 600- 60 I. Sanchez Palomares, 0., F. Sanchez Serrano & M. P. Carretero Carrero. 1999. Modelos y cartografia de estimaciones climaticas termopluviometricas para la Espaiia peninsular. Instituto Nacional de Investigacion y Tecnologia Agraria y Alimentaria. Ministerio de Agricultura, Pesca y Ali­ mentacion. Madrid. Schuster, R. M. 1992. The Hepaticae and Anthocerotae of North America. vol. 5. Chicago. Sergio, c. & A. Seneca. 1997. The first report of Cryptothallus mirabilis Malmborg (Hepaticae, Aneuraceae) in Southern Europe (Portugal). Cryptogamie, Bryol. 18: 49-51. Sergio, c. & c. Garcia. 1999. Cryptothallus mirabilis Malmb. (Aneuraceae. Hepaticae ) in Portugal. New ecological and phytogeographical data. Haussknechtia Beiheft 9 (Riclef-Grolle­ Festschrift): 343-346. Sergio. c. & D. Draper. 2002. How to evaluate species when distribution is poorly understood. The use of predictive studies for Iberian bryophytes. Portugaliae Acta BioI. 20: 37-48. Tuhkanen, S. 1980. Climatic parameters and indices in plant geography. Acta Phytogeogr. Suecica 67: 1- 109. Vanderpoorten, A .. A. Sotiaux & P. Engels. 2005. A GIS-based survey for the conservation of bryophytes at the landscape scale. Biological Conservation 121: 189-195. Wiehle. w. . C. Berg & R. Grolle. 1989. Cryptothallus mirabilis Malmborg neu in Mitteleuropa. Her­ zogia 8: 107- 124.