Oxford Plant

Systematics With news from University Herbaria (OXF and FHO), Department of Plant Sciences, Oxford

OPS 15 May 2008

Details from watercolours and sketches made by Ferdinand Bauer for Sibthorp and Smith’s Flora Graeca . Ophrys tenthredinifera (top) and Hypericum olympicum (bottom).

Foreword

Contents This year OPS once again reflects the breadth of systematic activity being carried out within the Department of Plant Sciences. Foreword From a Darwin project in Bolivia, student reports covering plants in Argentina, Robert Scotland …………………………………………………………. 2 Bolivia, Peru, Brazil, a published book on the history of Flora Graeca , and a short News items ...... 2 piece covering Rosemary and Serena's visit to Gotland in Sweden following in the New Darwin Project in Bolivia footsteps of Linnaeus. In addition, Mike John Wood ……………………………………………………………… 3 Hopkins provides an interesting insight into databasing Brazilian herbaria, which Publications 2007 …………………………………………………………. 4 includes using a simple and inexpensive system in conjunction with BRAHMS 6 to collect specimen data from the herbaria in Student reports Oxford that can be utilized in Brazil. Denis Systematics of Amicia (Leguminosae, Papilionoideae) and plant Filer also provides a useful update on recent diversification and endemism in Andean seasonally dry tropical developments in BRAHMS. I hope you forests enjoy reading OPS 15. Tiina Sarkinen ………………………………………….….…………… 4

Robert Scotland Ecological setting and the evolution of Neotropical plants: origins and University Reader in Systematic Botany diversification of the Cerrado flora

Marcelo Simon …………………………………….……….…………... 5

News items Sibthorp, Bauer and the Flora Graeca Stephen A. Harris ………………………………………….…………… 7

We are pleased to report that Rosemary Flora Graeca online Wise , Departmental Botanical Artist, has not yet retired and is still illustrating plants Roger Mills …………………………………………………..…………. 8 for papers and books for the systematics research group in the Department. Following Linnaeus's journey through Gotland Rosemary reported on her 42 year career in Rosemary Wise ……………………………………………….………… 8 the Department in the last volume of OPS which was then thought to be on the eve of News from the Herbaria her retirement, but fortunately for the group Fielding-Druce (OXF) and Daubeny (FHO) she is still here! Serena Marner ………………………………………………...………. 10

A book written by Dr Stephen Harris , Curator of Oxford University Herbaria, Databasing collections and repatriating information – a digital shortcut entitled 'The Magnificent Flora Graeca: Mike Hopkins ……………………………………………….…..……... 11 How the Mediterranean came to the English Garden' was published towards the end of Publishing online from BRAHMS 2007. The book, published under the Denis Filer ………………………………………………….……..….. 12 Bodleian Library programme with reference to collections of books and manuscripts found in the Bodleian Library and other OULS libraries, was launched on 13 December 2007 by Dr Shirley Sherwood as special guest speaker. The book proved to Department of Plant Sciences be very popular and sold over half the print run in the first five weeks of sales. Stephen South Parks Road Harris was awarded the Sibthorp Medal for Oxford his work by the new Head of the OX1 3RB Department of Plant Sciences, Professor U.K. Jane Langdale. See article on page 7. Tel. +44 (0) 1865 275000

We would like to offer our congratulations Oxford Plant Systematics Research Group website: http://herbaria.plants.ox.ac.uk to Professor David Mabberley , a former staff member of Plant Sciences and Acting OPS is a yearly publication. Back issues of OPS can be viewed on the website at Curator of the Oxford University Herbaria http://herbaria.plants.ox.ac.uk/OPS.html (1992-4), on his recent appointment as Keeper of the Herbarium at the Royal Botanic Gardens Kew. From 1 st May he Typesetting and layout of this issue by Serena Marner will also be Visiting Professor of Plant Oxford Plant Systematics printed by Parchments of Oxford Sciences, University of Oxford.

2 Oxford Plant Systematics OPS 15 May 2008

endemic to the biome. New Darwin Project in Levels of endemism vary Bolivia throughout the region and across the different physio- A new project financed by the Darwin gnomies with highest rates Initiative to identify conservation priorities of endemism noted in in the Bolivian cerrados was approved in campo rupeste (Filgueiras April 2007. This project is headed in the UK 2002: 123), fortunately the by the Department of Plant Sciences and we least threatened vegetation have partners in Bolivia, Brazil and in the subtype because of its rocky United Kingdom. Our principal partner is nature. Within the biome the Museum of Natural History in Santa individual mountain ranges Cruz which houses the main herbarium in (known as chapadas in eastern Bolivia. Also involved are three Brazil and serranias in institutions with which we have long- Bolivia) often have species standing links going back many years: The narrowly endemic to the National Herbarium of Bolivia in La Paz, particular range. In Bolivia the Universidade de Brasília and the Royal two such ranges are Botanic Gardens at Kew. Each of these Recently burnt cerrado (campo limpo) with immediately apparent, the Serrania de campo rupestre behind Huanchaca situated in the Noel Kempff institutions can offer specific skills and Photo J.R.I. Wood expertise in particular plant groups or National Park and the long serrania that stretches from Chochis to the east of methods to help move the project forward. Cerrado is now one of the most threatened Santiago. Even at this early stage of the For the use of project results and their biomes in the world. According to project it is clear that these two serranias are implementation after the project’s end, we Cavalcanti and Joly (2002), at least 67% of outstanding in their plant diversity and in depend mainly on the Fundación para la the Cerrado region of Brazil had been Conservación del Bosque Chiquitano, a converted to intensive human use by the the presence of species endemic to the particular serrania as well as those with a Bolivian NGO set up in 1999 specifically to early 1990s and doubtless this percentage very restricted distribution within the promote the conservation in this region. has risen since. The same process has begun Cerrado biome as a whole. The Cerrados constitute a large biome in eastern Bolivia and is likely to accelerate One of the most diverse groups with centred on Brazil but extending into Bolivia, rapidly. The main threat at the moment is particularly high rates of endemism in the belatedly recognised as one of the the conversion of natural cerrado grassland biodiversity hotspots of the world (Myers et to pasture seeded with Brachiaria and other cerrados is the Mimosoid legume genus Mimosa (Simon & Proença 2000). While the al. 2000). Much of it is on the pre-Cambrian introduced grass species. Unfortunately this centre of diversity in this genus lies clearly shield, which forms the bedrock at the heart is extremely easy to achieve given the flat in the cerrados of central Brazil, as with so of South America. There are reported to be topography of most cerrado and the ease by many Cerrado genera, the cerrados of more than 10,000 plant species which scrub and scattered trees can be characteristic of this biome and although it removed by bulldozers. Other threats which Bolivia also harbour a significant cohort of has been relatively well-studied in Brazil it are already a devastating reality in narrowly restricted endemics and there is every prospect that the project will discover is poorly known and under-collected in neighbouring Brazil include the conversion others. Bolivian Mimosa have recently been Bolivia. Cerrado develops on poor soil with of cerrado to arable for the cultivation of examined by Margoth Atahuachi high levels of aluminium in areas with a soya, the use of lime and chemical fertilizers (Cochabamba) with Colin Hughes in Oxford pronounced dry season. A particular which change the composition of the soil and Mimosa is also the focus of new work characteristic is that it is subject to periodic and invasion by exotic species such as by Marcelo Simon, from Brazil as part of natural fire – something that is becoming Panicum maximum , which seems to be his D.Phil. in Oxford. The endemics from more frequent as a result of human activity - encouraged by excessive burning. and adaptation to fire is a key feature of the San Jose-Santiago serranias many cerrado plants. Cerrado does not include Mimosa jacobita develop on rich soils even when periodically Barneby, M. josephina Barneby, burnt. M. dalyi Barneby, M. Cerrado is a general term applied to a orbignyana Barneby and M. mosaic of vegetation types, broadly similar auriculata Benth. with two to the term “savanna” in popular usage. A other undescribed species bewildering range of terms have developed known from this area. From the in an attempt to classify numerous subtypes Sierra de Huanchaca, one new or “physiognomies” ranging from campo species, M. suberosa Atahuachi limpo , a type of treeless grassland, at one & Hughes was recently extreme to cerradão , a closed woodland at described, to add to M. the other. Amongst numerous additional huanchacae Barneby and terms, campo rupestre is used to describe another as yet undescribed the vegetation of the broken rock outcrops species thought to be endemic typical of the region and characterised by an there. It is notable that the abundance of species of Velloziaceae and majority of these species are Eriocaulaceae. Wetlands are also a feature A new Blepharodon (Asclepiadaceae) species described within the last 20 years, most of of the region and include flooded from the Cerrado them by Barneby (1991). It is also clear that grasslands, seasonal pools and areas of bog- Photo J.R.I. Wood the other serranias of eastern Bolivia remain like vegetation with Drosera , Xyris and extremely poorly known and under- Lycopodiella . In general, herbaceous Anyone who has seen a diverse cerrado collected botanically. Given the high levels members of the flora are very numerous and some weeks after it has been burnt will be of endemism in Mimosa and other Cerrado are reported to outnumber woody species by struck by the colourful diversity of flowers, genera, it is likely that new species will between 3 and 4.5 times (Filgueiras 2002: which recalls an alpine or Himalayan continue to be discovered as these areas are 122). meadow. The vast majority of these are explored during the project.

Department of Plant Sciences, University of Oxford 3

Trinick, M.J., Prescott, A.R., Sprent, J.I. & Bolivia, but also because it felt every bit as James, E.K. (2007). Burkholderia adventurous as Che’s early journey along phymatum is a highly effective nitrogen- the dusty roads of South America. fixing symbiont of Mimosa spp. and fixes The focus of my D.Phil. project is plant nitrogen ex planta . New Phytologist diversification in the fragmented seasonally 173 :168–180. dry tropical forest (SDTF) biome of the New World and during my fieldwork I Hawkins, J.A., Boutaoui, N., Cheung, K.Y., aimed to visit a set of these disjunct SDTF van Klinken, R. & Hughes, C.E. (2007). areas scattered across the Andes from Intercontinental dispersal prior to human northern Argentina to northern Peru. These translocations revealed in a cryptogenic SDTFs, although not as species-rich as wet

invasive tree. New Phytologist 175 : 575– forests, are especially rich in narrowly- Mimosa jacobita Barneby 587. restricted endemics, and the origins of this Photo Colin Hughes high endemicity remain poorly understood. Harris, Stephen, A. (2007). The The markedly disjunct distribution of SDTF References Magnificent Flora Graeca: how the with isolated patches of forest scattered Barneby, R.C. (1991). Sensitivae Censitae: a Mediterranean came to the English Garden. from northern Mexico to Argentina and description of the genus Mimosa Linnaeus Bodleian Library. 189pp. Paraguay suggests that either long-term (Mimosaceae) in the New World. Memoirs isolation (dispersal limitation) could have of the New York Botanical Garden 65 : 1– Hughes, C.E. , Rajanikanth, G., Robertson, driven species diversification and high 835. A., Filer, D.L ., Harris, S.A. & Bailey, C.D. endemicity, or that these forests were (2007). Serendipitous backyard previously more continuous with subsequent Cavalcanti, R.B. & Joly, C.A. (2002). hybridization and the origin of crops. vicariant fragmentation driving diversif- Biodiversity and conservation priorities in Proceedings of the National Academy of ication. Distinguishing between these the cerrado Region. In: Oliveira, P.S. & Sciences 104 : 14389–14394. hypotheses is the aim of my research. Marquis, R.J. (eds.) The Cerrados of Brazil : To tackle this question, the initial focus of 351–367. New York: Columbia University Stepkowski, T., Hughes, C.E. , Law, I.J., my D.Phil. project is to build densely Press. Markiewiez, L., Gurda, D., Chlebicka, A. & sampled and well-supported time-calibrated Moulin, L. (2007). Diversification of lupin species-level phylogenies for a number of Filgueiras, T.S. (2002). Herbaceous Plant Bradyrhizobium strains: evidence from SDTF plant genera as one way to test the Communities. In: Oliveira, P.S. & Marquis, nodulation gene trees. Applied & hypothesis of dispersal limitation. I have R.J. (eds.) The Cerrados of Brazil : 121–139. Environmental Microbiology 73 : 3254- selected three genera – Amicia (Legum- New York: Columbia University Press. 3264. inosae), Tecoma and Delostoma (both Bignoniaceae) with diverse endemics in the Myers, N., Mittermeir, R.A., Mittermeier, Wood, J.R.I. (2007). The Salvias highly disjunct and potentially isolated C.G., Fonseca, G.A.B. & Kent, J. (2000). (Lamiaceae) of Bolivia. Kew Bulletin 62 : SDTFs of the Andes but which are Biodiversity Hotspots for conservation 177–222. manageable in terms of total numbers of priorities. Nature 403 : 853–858. species. Such work demands densely Wortley, A.H., Harris, D.J. & Scotland, sampled phylogenies (all species and Simon, M.F. & Proença, C. (2000). R.W. (2007). On the taxonomy and multiple accessions of species) and hence Phytogeographic patterns of Mimosa phylogenetic position of Thomandersia . good taxonomic knowledge of species (Mimosoideae, Leguminosae) in the Systematic Botany 32: 415–444. limits. Taxonomic accounts are available for Cerrado biome of Brazil: an indicator genus Tecoma (recently revised by John Wood) of high-altitude centers of endemism? and Delostoma (revised by Alwyn Gentry). Biological Conservation 96 : 279–296. Student reports However, this is not the case for Amicia , with no recent revision, and I have

John Wood

Research Associate Systematics of Amicia (Leguminosae, Publications 2007 Papilionoideae) and plant diversification and

Bebber, D.P., Marriott, F.H.C., Gaston, K.J., endemism in Andean Harris S.A. & Scotland, R.W. (2007). seasonally dry tropical Predicting unknown species numbers using forests discovery curves. Proceedings of the Royal Society of London. Series B . 274 : 1651– During the past year of my D.Phil. studies, I 1658. have been extremely lucky to have had the opportunity to spend three months in the Bebber, D.P., Harris, S.A. , Gaston, K.J. & Andes doing field work. This required some Scotland, R.W. (2007). Ethnobotany, plant stamina but has been key to the success of discovery and the first written records of my first year studies. This short account of UK flowering plants. Global Ecology and my travels and field activities in Argentina, Biogeography 16 : 103–108. Bolivia and Peru might well be dubbed my Motorcycle Diary , not just because I Elliott, G.N., Chen, W.M., Chou, J.H., followed in the footsteps of Che Guevara Wang, H.C., Sheu, S.Y., Perin, L., Reis, from his birth place in the idyllic city of V.M., Moulin, L., Simon, M.F. , Bontemps, Rosario in Santa Fe, Argentina, to the area Amicia lobbiana Benth. ex Rusby C., Surtherland, J.M., Bessi, R., Faria, S.M., where he was killed near Vallegrande in Photo Colin Hughes

4 Oxford Plant Systematics OPS 15 May 2008 undertaken a new taxonomic account of the This shows a number of interesting features. genus as a prelude to building a phylogeny. First, species are generally deeply recip- By comparing phylogenetic patterns and rocally monophyletic but with very short estimating diversification times for these branch lengths between species. Second, three genera, and integrating these with there is a surprisingly large amount of what is known for other Andean SDTF plant genetic variation found within each species groups, I aim to assess the tempo and mode which often seems geographically of diversification in SDTF taxa. structured. I am currently running a large By early spring 2007, I was heading to legume-wide age estimate analysis with South America to start my field work to fossil calibrations to estimate divergence collect silica-dried leaf material for my times for Amicia and looking for ways to study taxa, and to study and collect Amicia quantify the geographic structure across the for the taxonomic revision. In late February, tree. I arrived in Buenos Aires, Argentina to visit For Tecoma , I have generated a similarly the herbarium at the Instituto de Botánica densely sampled phylogeny using four Darwinion. There I had the pleasure of chloroplast DNA markers and the ITS. meeting Dr Emilio Ulibarri who is currently Initial results suggest that although the story revising the large South American genus in Tecoma is more complicated, it resembles Adesmia , a close relative of Amicia . In Amicia . Rosario, I met up with Dr Darién Prado at This spring (April-May 2008) I will be the Universidad Nacional de Rosario on the heading to the field again – I have not been way to the northern parts of this very large scared off by my first visit to the Andes! country. Accompanied by Darien’s This time I will be focusing more on colleagues Luis Oakley and Olga Martínez Plant collecting in Tarija, Bolivia, with Alex ecological work, gathering plot-based from the Universidad Nacional de Salta, I Wortley, Pablo Duchen and Colin Hughes species abundance data for the second part pressing a species of Lupinus. travelled through the NW Argentinian of my D.Phil. project.

Andean Departments of Catamarca, Jujuy, In Peru we were joined by Aniceto Daza, Tucuman and Salta, collecting specimens of Tiina Sarkinen (D.Phil., 2 nd year) and for part of the time Jose Luis Marcelo, Amicia and Tecoma (and of course other from the Universidad Nacional Agraria La taxa on the Christmas wish list of people Supervised by Dr Colin Hughes (Oxford) Molina, Lima and Maria Baden from Royal back at FHO and E!). and Dr Toby Pennington (Royal Botanic Botanic Garden Edinburgh. We set out on The field collecting in Argentina was very Garden Edinburgh). Osk.Huttusen Säätiö the long haul north from Lima along the successful, and it was a pleasure to work and Helsingin Sanomain 100- coast of Peru to spend two weeks in the with such extremely enthusiastic and vuotisjuhlasäätiö scholarships. northern Departaments of Piura, La professional field botanists as Luis and Libertad, Cajamarca, and Amazonas. We Olga. After sad goodbyes, I travelled to the found and collected most of the SDTF Bolivian border, walked over the frontier valley-specific endemics of Tecoma , and it with a terrible headache caused by the was amazing to see how quickly the local Ecological setting and the 4000m altitude, and took the direct train flora changes over very short distances, new from Villazón to La Paz. In La Paz, I met up evolution of Neotropical endemics popping up over distances as short with Pablo Duchen from the Herbario plants: origins and as 50km. Up in the Mara ňon valley in Nacional de Bolivia and was joined a few northern Peru we were also collecting other diversification of the days later by Colin Hughes and Alex legumes and stopped to collect a white- Wortley (RBGE, Edinburgh). After a few Cerrado flora flowered Mimosa on a busy roadside that days of preparations, we headed off into the has proven to be a new species!, adding yet Yungas and started collecting. Alex had Introduction another species to the already long list of joined us to collect Calceolaria , a very In my research I’m interested in the patterns species of Mara ňon endemics. These species-rich and largely Andean genus of of plant diversity in the Neotropics, and exciting collections in this under-collected herbs and sub-shrubs in alpine meadows and particularly in the origins and evolution of part of Peru add valuable evidence to my cloud forests, locally known as Zapatitas , the Cerrado. The Cerrado is a floristically D.Phil. data, and looking back now it makes ‘Little Shoes’, due to their beautifully diverse and endemic-rich savannah biome me feel very lucky to have been there. shaped flowers. Colin as always was with more than 10,000 plant species The last spell of field work started once we 2 looking for more Lupins, and on many covering more than two million km of had returned to Lima from the North. I was occasions succeeded in finding them. Central Brazil and part of Bolivia. It has very keen to include the southern Peruvian Our journey in Bolivia took us to the been suggested that the Cerrado may have dry valleys in my sampling, especially of Yungas in La Paz Department, through originated as recently as four million years Tecoma and we headed to Cusco, and from Cochabamba to Chuquisaca, and finally ago, coinciding with the emergence of there to Arequipa, Moquegua and Tacna south to Tarija and back to the Argentinian inflammable C4 grasses as a dominant where three of the Tecoma endemics are border. We were lucky and had plenty to component of this ecosystem and the found. Returning to the UK at the end of collect, including a magnificent set of consequent onset of regular fires. One way May after three months of strenuous and multiple collections of all the species of of testing this hypothesis is to use dated hectic field work, I was happy to get stuck Amicia known from Bolivia. After getting molecular phylogenies of biome specific in the lab and settle in for a while. stuck in Tarija due to demonstrations and lineages to infer the age of a particular Since then, I have done a large chunk of road blocks surrounding the city, we finally biome. However, to date there are very few lab work and managed to generate some started the long two-day drive back to La phylogenies for Cerrado plant groups that respectable phylogenies for both Amicia and Paz across the bleak Bolivian Altiplano, can be used to test this hypothesis. I am Tecoma , with work on Delostoma planned stopping in the high cold city of Potosi. aiming to fill this gap by generating a for later this year. Using two chloroplast From La Paz, Colin and I continued to Peru, phylogeny for Mimosa , a species-rich sequence loci and the nuclear ribosomal Pablo stayed in La Paz, and Alex went back legume genus. Mimosa is an ideal study ITS, I have built a densely sampled and for a flying visit in the UK before leaving to group to test this hypothesis since it is robustly supported phylogeny for Amicia . Congo for another field trip! remarkably rich in fire-adapted narrow

Department of Plant Sciences, University of Oxford 5 endemics in the Cerrado, but at the same provided a reasonably good estimate of genetic analysis. With more than 500 time also highly diverse in other major around 18 million years for the genus species of Mimosa scattered across four Neotropical habitats such as seasonally dry Mimosa (crown node). Based on this continents, any attempts to build a tropical forest and rain forest, providing a information, I was able to estimate the ages phylogeny depend on effective collaboration good basis for comparisons. of many nodes of interest in a subsequent of this sort, and are especially effective with phylogenetic analysis of trnD-trnT someone as knowledgable and enthusiastic chloroplast sequences for 250 species of about Mimosa as Dr Grether. Mimosa , which included a substantial set of Collaboration has also continued with Dr Cerrado endemics. With around 50% taxon Euan James and colleagues at the sampling, this provides reasonable Universities of Dundee and York under the geographic, ecological and taxonomic project “In search of beta-rhizobia: coverage of this large, diverse and exploring the symbionts of Mimosa in widespread genus. Brazil”, which aims to gain a better Results from these analyses show that understanding of the symbionts associated Cerrado endemics originated independently with Mimosa and especially the unusual several times in Mimosa . However, the beta-rhizobia that have been found to fix majority of Cerrado endemics (75% of all nitrogen in root nodules of Mimosa . Again, Cerrado species of Mimosa ) fall into two this collaboration provided scope for joint large clades rich in fire-adapted species, fieldwork to assemble material of Mimosa suggesting two independent Cerrado species in Brazil in 2006. With parallel plant host radiations. It is notable that these two and symbiont phylogenies there is an independent species-rich lineages divers- opportunity to gain insights into the ified at the same time around 3.5 million evolution of nitrogen fixing symbioses that years ago. These dates also closely match involve beta-rhizobia. The project team the estimated age of another large endemic published a paper last year showing that Cerrado plant lineage in the tribe beta-rhizobia are indeed effective symbionts Microlicieae (Melastomataceae; Fritsch et of a broad range of Mimosa species, and a al. 2004). This relatively recent origin of suite of other published outputs from this Cerrado lineages of Mimosa and collaboration are expected during the Microlicieae contrasts with recent findings coming year. that Neotropical seasonally dry tropical forest lineages are generally older References (Pennington et al. 2004, Lavin 2006). I am Fritsch, P.W., Almeda, F., Renner, S.S., currently testing this further by dating Martins, A.B. & Cruz, B.C. (2004). Cerrado lineages in two other legume Phylogeny and circumscription of the near- genera, Andira and Lupinus . endemic Brazilian tribe Microlicieae My initial results suggest a scenario of (Melastomataceae). American Journal of recent diversification with rapid and Botany 91 : 1105–1114. successful adaptive radiation of an endemic- rich fire-tolerant flora. If recent divers- Lavin, M. (2006). Floristic and ification is confirmed for other Cerrado geographical stability of discontinuous plant groups, the question remains as to seasonally dry tropical forests explains what type of vegetation was in Central patterns of plant phylogeny and endemism. Brazil before the Cerrado, and what biomes In: Neotropical savannas and seasonally dry contributed with lineages that were able to forests: plant diversity, biogeography, and colonize the Cerrado. conservation. Pennington, R.T., Lewis, G.P. Fire adaptations in Cerrado Mimosa species. These initial findings were presented at the & Ratter, J.A. (eds.). Systematics Top : M. splendida , a pachycaul treelet. Biennial Conference of the Systematics Association, Special Volume Series 69 : Bottom: M. speciosissima growing from a woody Association in Edinburgh last August, 433–447. Boca Raton, CRC-Taylor and xylopodium. supported by a student award from this same Francis. organization. This is the last year of my Results D.Phil. and I am busy analysing data, Lavin, M., Herendeen, P. & Wojciechowski, One critical issue for this type of study is reading and writing manuscripts and my M.F. (2005). Evolutionary rates analysis of how to establish a time dimension for a thesis, which I hope to submit by the end of Leguminosae implicates a rapid phylogeny. A powerful technique to do this the year. diversification of lineages during the is to use the number of nucleotide Tertiary. Systematic Biology 54 : 530-549. substitutions on branches of a phylogenetic Collaboration tree derived from DNA sequence data During 2007, Dr Rosaura Grether from the Pennington, R.T., Lavin, M., Prado, D.E., constraining specific nodes on the tree with Universidad Autónoma Metripolitana Pendry, C.A., Pell, S.K. & Butterworth, an age, normally derived from fossil dates. (UAM) in Mexico came to visit our group C.A. (2004). Historical climate change and This allows ages for all nodes of a given for three months as part of her sabbatical speciation: neotropical seasonally dry forest phylogenetic tree to be estimated. To do this leave. Dr Grether has longstanding interests plants show patterns of both Tertiary and for Mimosa I have re-analysed Lavin et al.’s in legume taxonomy and specifically in the Quaternary diversification. Philosophical (2005) family-wide matK data set for genus Mimosa providing excellent scope for Transactions of the Royal Society of London legumes adding 130 sequences for collaboration on Mimosa phylogenetics. In B-Biological Sciences 359 : 515-537. Mimosoid genera using 12 fossils as 2006 I had the opportunity to carry out calibration points. The number and quality fieldwork with Rosaura’s research group in Marcelo Simon (D.Phil., 3 rd year) of the fossil calibrations that were used as Mexico and this recent visit to Oxford by Dr minimum age constraints for particular Grether focused on laboratory work to Supervised by Dr Colin Hughes and Dr nodes across the phylogeny, and the generate sequence data for a large cohort of Stephen Harris. Clarendon scholarship and improved taxon sampling in this analysis, Mexican species of Mimosa for phylo- Embrapa.

6 Oxford Plant Systematics OPS 15 May 2008

Professor in 1784 and promptly departed for hundreds of plants and animals and labelled Sibthorp, Bauer and the another protracted European tour; he them with numbers that referred to a Flora Graeca returned to Britain in late 1787. It was specific colour code. In Oxford, Bauer, during this tour that he conceived the idea of despite his strained relationship with the Flora Graeca and did the majority of the Sibthorp, used his sketches, together with associated fieldwork. the herbarium material and, perhaps, living At the end of the 18 th century, major material from the Botanic Garden, to botanical expeditions were being undertaken produce 966 magnificent, folio, life-sized by botanists from Continental Europe, botanical watercolours; on average, Bauer especially to the New World. John Sibthorp, produced one watercolour every one and concerned with his botanical reputation and one quarter days. These watercolours would legacy, decided to explore the relatively give the Flora Graeca its reputation for unknown western fringe of the Ottoman magnificence and contribute to its Empire. The Flora Graeca was to contain extraordinary cost. the botanical results of this exploration. He Sibthorp returned to the Ottoman Empire would be aided by the talents of the for a shorter, more traumatic fieldtrip, phenomenal botanical artist, Ferdinand without Bauer, in 1794. Sibthorp arrived Bauer (1760-1826), who was engaged for back in Britain, ‘more dead than alive’, in £80 per year (later rising to £100), and his late 1795 and died in Bath in February travelling companion John Hawkins (1761- 1796; he never returned to Oxford. In his 1841). Sibthorp and Bauer set off from will, Sibthorp provided the vision, in March 1786, travelled through incentives and financial resources to Italy and passed through the Straits of complete his Flora Graeca project. Messina to collect plants in Crete. During Through the perseverance of John Hawkins his Cretan trip, gems such as Teucrium and Thomas Platt, Sibthorp’s friends and Professor John Sibthorp (1758-1796) alpestre and Staehlina arborea were executors, the Flora Graeca was eventually collected. From Crete, Sibthorp and Bauer completed. James Smith, founder of the In Britain, the late-eighteenth century was travelled to the Turkish coast and then Linnean Society, was the first botanist to be an exciting time to be interested in plants. overland to Constantinople where they employed and was faced with unmounted Fortunes had been made exploiting plants, overwintered. Here they met Hawkins and, and unlabelled herbarium specimens, the great and the good were enthusiastic in early 1787, caught the Mediterranean Sibthorp’s notes written in an ‘execrable botanical patrons, new regions of the globe spring as they travelled to , where hand’ and Bauer’s largely unlabelled were being explored and their botanical they made the first botanical collections sketches and watercolours; Smith was paid treasures returned to Britain’s shores. known to western science and found the £75 per year for his work. Having brought Furthermore, new scientific methods were highly restricted Cypriot endemic sage, order to this chaos, Smith became engaged being used to investigate the botanical world Salvia veneris . From Cyprus they on the laborious task of ensuring that and Carl Linnaeus’ classification and meandered across the Aegean and arrived in Bauer’s watercolours were reproduced to naming systems were being widely adopted. Athens in June 1787. Excursions through the highest possible standard. Renowned In Oxford, John Sibthorp (1758-1796), followed until September 1787, botanical engravers, the Sowerby family, Third Sherardian Professor of Botany, when Sibthorp and Bauer returned to Britain were employed to produce the coloured extolled Linnaeus’ virtues in his and Oxford. Throughout his journey engravings. The first part of Volume One of undergraduate lectures: ‘the most interesting Sibthorp had collected vast quantities of the Flora Graeca was published in 1806. Period of the Progress of Botany, when the herbarium material, seeds, bulbs and animal When Smith died in 1828, his place was bold but systematic Genius of Linnaeus specimens, whilst Bauer had sketched briefly taken by Robert Brown, before John forged as it were a Chain, which Lindley completed the project in 1840, encompassed the whole of Nature’. when the final part of the Volume Ten was However, lecturing was only a minor part of published. The total project had cost John Sibthorp’s short career as an Oxford £15,572 6s 10d, and the income had been Professor; he was also responsible for the £15,581 12s. After 54 years the project was Botanic Garden and the Herbarium. complete, the balance, £9 5s 2d, was John Sibthorp was the only son of transferred to the University, and the income Humphrey Sibthorp, Second Sherardian from Sibthorp’s estate used to establish the Professor, by his second marriage. Sibthorpian Chair in Rural Economy (now Humphrey Sibthorp is reputed to have Plant Sciences). Each of the 25 subscribers published nothing and given but one lecture had paid £254 for their copy of the Flora in his 38 years as occuptant of the Graeca . Sherardian Chair. John Sibthorp was Sibthorp’s legacy was the discovery of educated as a physician in Edinburgh and hundreds of plants new to science and the came under the intellectual influence of the publication of one of the most costly, rare first teacher of Linnaeus’ classification in and magnificent botanical books ever Britain, John Hope. Sibthorp was to become written. Some of the species that Sibthorp very wealthy with the premature death of his introduced to Oxford are still known in mother and through the actions of his father British gardens, e.g., Crocus flavus , but the to convince the University that his son fate of the majority of the seeds and living should benefit from a ten-year Radcliffe material he returned to Britain is unknown, Travelling scholarship (worth £300 per as the records have been lost. However, the year). John promptly set-off to some of the Botanic Garden may contain a living most prominent botanical institutions in remnant of the Sibthorp’s botanical efforts Europe and returned to Oxford in 1783, in the shape of the iconic black pine tree. when he convinced his father to resign the Staehlina arborea from the Flora Graeca . Today, Sibthorp’s herbarium specimens are Sherardian Chair in his favour. John became housed in Oxford University Herbaria,

Department of Plant Sciences, University of Oxford 7 whilst his manuscripts, together with provide more sophisticated searching and plants, but geological features, bird colonies Bauer’s original water colours and sketches, links to the digital images of the and ancient buildings. There were too many are in the Department of Plant Sciences corresponding herbarium specimens, where locations and plants seen for me to mention Library. available. Meanwhile we hope the them all. browsable collection will whet your appetite Stephen A. Harris and make these wonderful drawings even Curator of Oxford University Herbaria more widely appreciated. For publication, high-resolution digital images can be ordered from OULS Imaging Services on http://www.ouls.ox.ac.uk/services/imaging_ services . Appropriate reproduction fees are payable. Orders should include the image number which is displayed as a ‘tooltip’ if you hover over an image with the mouse. For research and private study, the medium- resolution images on the web can be saved to disk as required, provided credit is given to Oxford University Library Services whenever reference is made to them.

Any queries or comments should be sent to [email protected] or [email protected] , who will be pleased to assist.

Roger Mills

Acting Keeper of Scientific Books, Map of Gotland showing places visited by Oxford University Library Services Linnaeus in 1741 and the Linnean Society party in 2007

Following Linnaeus's Our base was in the city of Visby, possibly the best-preserved medieval city in journey through Scandinavia and now a UNESCO World Fritillaria sibthorpiana from the Flora Graeca . Heritage Site. Excavations show that there

Gotland was a settlement here as early as the Stone

Age, benefited by the natural harbour and Flora Graeca online sources of fresh water. The 3.4 km While here in the United Kingdom it was defensive stone wall around the town was one of the coldest and wettest summers on th originally built in the 12 century with record, Serena Marner and I enjoyed Work has now been completed on digitising towers added at various times during the continuous hot sunshine in Sweden. As part the second phase of Sibthorp’s Flora next three centuries. Midsummer light in of the celebrations for the tercentenary of Graeca materials. For some time the initial Scandinavia is legendary. With such a busy the birth of Linnaeus, in June we attended pilot project, including selected scans from schedule, I spent a couple of happy hours an excellent symposium in the University of the Flora Graeca original drawings and drawing and painting before breakfast each Uppsala, entitled “ Unlocking the Past ” published engravings, together with the day! which covered many aspects of Linnaeus’s unpublished Fauna Graeca and Mediter- In June the plants were at their best. life and times. We visited his house in ranean Scenes , have been available on the Gotland conserves many extremely lush and Uppsala, the botanic gardens, his country Oxford Digital Library web site at species-rich parish meadows and we saw estate at Hammarby and paid homage at his http://www.odl.ox.ac.uk/odlcollections several, including those of Anga and Fide. tomb in the cathedral, for the occasion (select Flora Graeca for the 21 st century Linnaeus said that these meadows looked beautifully decorated with a border of living from the menu). Now all the original more like gardens than meadows, so plants of Linnaea borealis . Following this, drawings from these series, together with all colourful were the flowers, and it was very 32 of us set off on an expedition to the the plates and accompanying text from the difficult to walk, and more especially to take Baltic island of Gotland, organised and led published Flora Graeca are available for photographs, without flattening any. Many with perfect precision by Bengt Jonsell, browsing on species of orchids thrive in such areas, Roland Moberg and Eva Willén of the http://www.ouls.ox.ac.uk/isbes/taxonomic_c including Common spotted ( Dactylorhiza Swedish Linnaeus Society. ollections (select Flora Graeca in the 21 st sambucina ), Early purple ( Orchis mascula ), Linnaeus visited the islands of Gotland and Century from the left-hand menu). All 3278 Military ( O. militaris ), Burnt ( O. ustulata ) Öland in 1741, having been commissioned images, plus captions, can be seen here in Twayblades ( Listera ovata ), Frog by the Swedish government to survey the medium resolution. On this site there is (Coeloglossum viride ) and Lesser butterfly natural resources including medicinal and currently no search function, but the files (Platanthera bifolia ). Also abundant were economic plants, soils, and the commercial will be added to the Oxford Digital Library plants such as Meadowsweet ( Filipendula possibilities of various minerals and ores. site in due course where they can be ulmaria ), Yellow rattle ( Rhinanthus He kept a detailed diary, noting not only searched (by name of a species, for serotinus ), Common rock rose ( Helian- strictly scientific items but everything from example) and also zoomed to a higher themum nummularium ), Bloody cranesbill the construction of gates, local cures for magnification. (Geranium sanguineum ), Great burnet common illnesses, dialects, runic stones, We also plan to provide hyperlinks to the (Sanguisorba officinalis ), Herb-Paris ( Paris folklore and what the inhabitants wore. We images from the corresponding entries in the quadrifolia ) and the leaves of Lily-of-the- visited many of the places that he had found Herbarium BRAHMS database, which will valley – and many, many others. particularly interesting, not just to look at

8 Oxford Plant Systematics OPS 15 May 2008

By contrast, the rocky inland cliffs at offering contained fossils Torsburgen, which Linnaeus called a of many trilobites. We mountain, reach the height of 68m above found, as Linnaeus had sea level and presents a far dryer area, prone recorded in his diary, to fire. There must have been one some “Each stone was nothing years before Linnaeus’s visit as he notes that but a coral” and added the top of the “mountain” was bare of trees that there was enough "for and that he was able to count 30 church every man in the world to towers. The regeneration of fire-resistant get a cartload each". He plants such as Scorpion senna ( Hippocrepis saw the same series of emerus ) were in evidence following the fire retreating beach ridges, of 1992 but we were not able to see the noticed how the fossils churches for trees. A few kilometres away is further inland were more the inland cliff of Östergarnsberget. The weathered and was sure typical rock vegetation here includes that living material would Juniper, Pine, Herb Robert ( Geranium be evident out in the bay robertianum ) and Ciliate melick ( Melica but didn’t have time to ciliata) the only species colonising the stony look. Vincetoxicum hir- ground on the plateau. undinaria is one of the At Ulla Hau, north east Fårö (a small only plants to have island at the northernmost tip of Gotland) colonised these nutrient- another completely different vegetation type deficient beach ridges in is found which Linnaeus said reminded him any number. Its leathery of Holland. An enormous horseshoe-shaped th leaves are well adapted to th The Parish Church and 12 century defence tower at sand dune started to form in the 18 century Gammelgarn. The tower now houses a Linnaean exhibition. cope with the wind and following a period of deforestation and over salt. Occasional patches grazing. This was eventually stabilized by One calm, hot day we went by ferry to the were seen of Sea campion ( Silene uniflora ), the planting of Marram grass ( Ammophila island of Stora Karlsö. Geologically, the Lady’s bedstraw ( Galium verum) , Thyme arenaria ) on the seaward side and pine, island is composed of a core of reef (Thymus serpyllum ), Biting stonecrop alder and birch at the western extent. On the limestone and the power of the sea has (Sedum acre ) and Carline thistle ( Carlina dunes, apart from the Marram, the gouged out some spectacular caves. One vulgaris ). vegetation consists almost solely of the excavated in the nineteenth century revealed From 1361 until 1645 Gotland was under Grey-hair grass ( Corynephorus canescens ). continuous evidence of Stone Age Danish rule. At this time a great number of Searching for the dips in the sand inhabited occupation from 3000 to 2000 BC. The churches were built over the island in the by ant lions and teasing them amused many bones found in the systematic removal of Danish architectural style and looking very of us! The composition of the sand in this the 4m layer of soil indicate the changes in different from those on mainland Sweden. area is unique to Gotland, completely diet from the end of the Stone Age when Many have been restored to their full glory. lacking in calcite and only here, in the sheep, cattle and goats were domesticated We visited several. Gothem parish church wooded area, do we find species common to back to the earliest inhabitants who lived on is home to murals painted around 1300 mainland Sweden. We delighted in our first birds and seals. (Split human bones suggest depicting the farming year, month by sighting of Twin-flower, Linnaeus’s “own” there could have been cannibalistic month. The German artist also combined flower ( Linnaea borealis ), growing in the tendencies too). Stora Karlsö is except- dragons, knights, eagles and centaurs carpet of Bilberry ( Vaccinium myrtillus ), ionally interesting to botanists as 560 alongside the religious themes. Gammel- Cowberry ( V. vitis-idaea ), Crowberry species of vascular plants, 126 mosses and garn church lies below Östergarnsberget (Empetrum nigrum ) and various mosses. some 300 species of lichens have been inland cliffs and the defensive tower, built Linnaeus visited the nearby Ava farm, recorded. Linnaeus wrote that the only tree in the twelfth century, is one of the best which he said was “neat and tidy” and he saw on the island was the ash growing on preserved on the island. In Fårö church, an rested under an enormous oak which he the Bronze Age burial cairn of Röjsu, the oil painting and a poem commemorate the noted had a trunk diameter of 4.5m and a highest point of the island, which was used amazing story of the 15 local seal-hunters crown diameter of 28m. The tree is as a landmark by sailors. On such a hot day who in 1603 drifted in the Baltic for two considerably larger now and the owners take we were grateful for the shade provided by weeks when the ice floe they were on broke very great care of its preservation. the same healthy tree at lunchtime. In this loose. All survived (on seal meat) and the exposed area the predominant plants painting depicts them with hunting growing on the limestone shingle were equipment and dressed in their best clothes! Lesser meadow-rue ( Thalictrum minus) , Another curious painting in the same church Thyme ( Thymus serpyllum) and Sheep's shows a vicar of the seventeenth century fescue ( Festuca ovina ). Late afternoon we with his many offspring. Not all had perhaps lingered too long looking at the survived childhood but all were included, flora of the meadows and gazing down on the deceased painted at the age they had the colonies of guillemots and razorbills on reached, differentiated by the crosses above the towering sea cliffs. This necessitated a their heads. Of the four babies in the cradle, mad rush down a steep, rocky path through the only one to survive is shown with open an area of deciduous woodland to catch the eyes. ferry. Strandridargården was the home of Grodde nature reserve, north-west Gotland, generations of customs officers and is now was a highlight for many of us. A local preserved as a museum. Adjacent is the geologist suggested we took a ten minute mysterious and extensive sea-stack field of walk on the beach and for each of us to Kyllaj. Reminiscent of Easter Island, these

bring her two “pebbles” for identification. oddly-shaped monoliths are incongruously The trunk of the oak at Ava farm My two were identified as a rugose coral of sitting in grassland way above the sea. No the Palaeocyclys genus and my second two look alike and Linnaeus wrote that they

Department of Plant Sciences, University of Oxford 9 were “Arranged in a row like the ruins of Linnaeus' visit to Oxford'. This display was Islands by Mr Walter Scott (co-author of the churches and castles” and “From a distance prepared for a special meeting of the Shetland Flora 1987) and miscellaneous they looked like statues, horses, torsos and I Linnaean Society of London held in the species of British grasses collected by Mr do not know what kind of ghost”. Oxford Museum of Natural History on 29 John Killick (co-author of the Flora of September. The exhibit included two plant Oxfordshire 1998). 71 lichen specimens specimens of Pedicularis (Scrophulariaceae) collected in the UK, Poland, Russia, Finland collected by Linnaeus from Lapland, which and the USA were donated by Professor are now held in OXF. The specimens were Mark Seaward, University of Bradford, originally sent to Professor Johann Jacob which came from miscellaneous collectors. Dillenius by Linnaeus himself. A modified In addition type material of a new British version of the display is currently on show species of Taraxacum (Asteraceae), T. in the Department of Plant Sciences Library. ronae , was donated to the Druce Herbarium Another highlight of the year was meeting by Mr L.J. Margetts.

A rugose coral from Grodde, north-west Gotland and hosting a visit to the herbaria of Mrs The Fielding Herbarium has recently been Yasuko Ikenobo, Japanese Senior Vice providing a resource for researchers in the On our last day we visited the extreme Minister of Education, Culture, Sports, University's Research Laboratory for south of the island and one of Gotland’s Science and Technology and five of her Archaeology and the History of Art, on a main tourist attractions, the highly- colleagues from the Ministry of Education, project studying Egyptian historic photogenic stack rocks at Holmhällar. With Higher Education Bureau, the School chronology being undertaken by Mike Dee the extensive sedge fen of Muskmyr and the Curriculum Division of Japan and the and Dr Fiona Brock. sandstone grindery at Kettelviken (which Education Attaché from the Embassy of One of our historic collections has received provided the stone for the building of the Japan. This was part of a whole day visit on special attention this year and is now royal palace in Stockholm) yet to explore, 9 August to the University by the Japanese databased and with every sheet digitally time was limited here unfortunately. Just to guests who had also met with the Vice- photographed – this is the collection of see these stacks again and to spend a whole Chancellor, Dr Hood, and visited the Oxford 'lower plants' gathered by Johann Jacob day exploring the area would certainly Botanic Garden and the Oxford Museum of Dillenius to illustrate his Historia tempt me to return! Natural History. Muscorum published in 1741. There are Muskmyr is one of the remaining areas of We were delighted to have with us in the 176 sheets in the collection and these are fen land, most of the former wetlands herbaria, from September to December associated with 901 taxa. Each entry in the having been drained for agricultural use. 2007, Dr Rosaura Grether from the database is linked with an image of the The sedge Cladium mariscus which fringes Universidad Autónoma Metripolitana relevant herbarium sheet and related text the lake was once widely grown over the (UAM) Mexico who came to work with and plate from the Historia . Approximately island for bedding and thatching purposes Marcelo Simon (see page 6). 16% of the specimens have been annotated and was one of its most valuable crops. We were also very pleased to welcome as types. The material can be viewed at During the 3 km walk we again saw many back Professor Dr Walter Lack for a day in http://dps.plants.ox.ac.uk/bol/?historiamusc species of orchids, and a host of other September and to meet Dr Kräftner, orum specialities, including Gotland rock rose Director of the Liechtenstein Museum in (Fumana procumbens ). We were lucky to Vienna who came to look at the Flora and Daubeny (FHO) be able to see and photograph the Apollo Fauna Graeca and associated materials. butterfly ( Parnassius apollo ) feeding on Many other visitors came to look at a New accessions to FHO during 2007 sedums. variety of different collections or to meet included specimens collected in Sierra My main memories will remain associated with staff. Leone by William Hawthorne. This is a with the intense colour of Gotland. The substantial collection and the material is still roadside verges were unbelievably bright Fielding-Druce (OXF) in the process of being mounted and with red poppies, and the intense blue of databased. A collection of Mimosa (Fab- Viper’s bugloss ( Echium vulgare ), Loans sent out from OXF during 2007 aceae) made by Marcelo Simon in Brazil thousands of plants en masse in so many included another consignment of grasses has been accessed and mounted for FHO. places, creates a dominant feature of the which were sent to K at the request of Dr 182 duplicates of Mimosa sent as a gift from Gotlandic landscape. Bruno Ryves who has kindly offered to HUEFS Brazil, for identification by name them. Professor Dr Maas, who visited Marcelo, were also received. In addition Rosemary Wise the herbaria in August, also kindly offered miscellaneous species collected by Tiina Botanical Artist to name plants collected by Richard Sarkinen in Bolivia and Peru were accessed Schomburgk in our indet. folders and a loan and mounted. Other smaller donations of these was then despatched to Utrecht. included a few duplicates of Meliaceae from Increasing numbers of images of specimens S.E. Asia, China, Vietnam and Sarawak, News from the Herbaria were sent as 'internet loans', 16 transactions plus a few duplicates of Lupinus (Fabaceae) of this nature were sent during 2007 with collected in Brazil and a number of sheets determinations for material received in Leucochloron (Fabaceae) sent for the Work continued steadily in all areas of exchange to enable us to update our attention of Colin Hughes and John Wood. herbarium activity, the databasing and database records. Due to the successful field excursions in digitisation of the collections, the sending A large amount of material was returned to South America undertaken by various and receiving of loans as requested by OXF during the year, over 1400 specimens. members of the systematics research group various institutes/individuals and for the A large quantity of British Mentha (see articles on pages 4-6), we were able to research activities of the systematics group (Lamiaceae) was returned from Edinburgh send many duplicates from FHO to other in the Department. One of the highlights of which had been sent to E the year before for herbaria. Marcelo Simon's collections of the year for me was celebrating the a project involved with the 'Hybrid Flora of Brazilian Fabaceae were sent to HUEFS tercentenary of the birth of Carl Linnaeus the British Isles'. Brazil and many of Tiina Sarkinen's with members of the Linnaean Society of New accessions to OXF included duplicates collected in South America were London (in London, Oxford, Uppsala and in specimens of Pilosella and Heiracium distributed to BM, E, HUEFS, K, L, MO, Gotland, see page 8) and preparing an (Asteraceae) collected in the Shetland MONT, NY, RNG and US. exhibit on the theme of 'Celebrating

10 Oxford Plant Systematics OPS 15 May 2008

Incoming loan material during 2007 material in FHO to be duplicated in one or The process becomes even more efficient comprised mostly specimens of Amicia more of the Amazonian herbaria. Thus, when you have access to data from (Fabaceae) sent for study by Tiina Sarkinen, when staff in Oxford type in data from their databases with duplicates of the newly with some material of Tecoma Amazonian collections, they are likely to photographed specimens. For example, (Bignoniaceae) also being requested and repeat work already done in Amazonia. On working on an RDE file with the INPA received. Other loans received included the other hand, specimens present in database open, means that by just typing in material of Mimosa for study by Marcelo Amazonian herbaria may not have the same the collector name and number the INPA Simon, Lupinus for Colin Hughes and a quality of identification as their duplicates database is automatically searched for small quantity of Aglaia (Meliaceae) for in Oxford; Amazonian herbaria are visited duplicates. If duplicates are found then the Caroline Pannell. A number of loans were by specialists much less frequently than all the fields are automatically filled-in. returned after completion of studies and foreign herbaria. Tiny Amazonian herbaria, Most of the Brazilian specimens those consisted of material of Tecoma e.g., those in Macapá or Jari, receive very photographed in FHO do indeed have (Bignoniaceae), Strobilanthes (Acanth- few visitors. As a result, many specimens duplicates in the Amazonia herbaria. The aceae), Hippeastrum (Amaryllidaceae), remain with their original field material in OXF is another story, since Acacia (Fabaceae) and Faurea (Proteaceae). determinations, often only best guesses at much of this older material is not present in Loan material sent from FHO to other family or generic level, despite duplicates Amazonia, or indeed Brazil. In the long- herbaria included specimens of Podocarp- having been identified by renowned run, high resolution images of these aceae, Annonaceae and Thymeleaceae. specialists elsewhere. I estimate that about specimens should be made but there is little Work on the Darwin Trinidad and Tobago 40% of the determinations at INPA (located point in photographing material already project continued and in connection with in Manaus, the main herbarium for Western duplicated in Amazonia and available as this a number of colleagues from Trinidad Amazonian Brazil) would change with high resolution images. and Tobago spent three weeks in FHO in efficient interchange of information between In Amazonia, we have found images are July working on databasing specimens. herbaria. essential for data verification. Typing errors Step 1: Using medium-priced digital are ubiquitous in all databases. One error in Serena Marner cameras (we use FujiFilm FinePix S5800), a collection number causes incorrect Herbarium Manager medium resolution pictures are taken of specimen matching. If you can consult an each specimen. The label and whole image of the label data, such errors can be specimen are photographed. Separate photo- corrected immediately. The Embrapa graphs of det-slips and other annotations are Amazonia Oriental (IAN) herbarium has Databasing collections taken as necessary. Using a hand held, auto- been 100% imaged and databased this way and repatriating focus camera, it’s possible to take at least (ca. 332,000 images) and about half the 1200 photographs per day which, at 2 INPA herbarium has been photographed (ca. information – a digital images per specimen, covers 600 specimens. 260,000 images; ca. 40,000 per month). shortcut Step 2: The camera images are collectively Especially interesting for botanists in transferred to disk and appended into a Amazonia are the identifications of material The world’s herbaria are under pressure to BRAHMS image Rapid Data Entry (RDE) duplicated in other herbaria. On the trip to database their collections. With ever file. By viewing the images, the barcode (or Oxford I concentrated on those families improving access to the internet, access to accession numbers) are added to the file which had recently been determined by the contents of a herbarium need no longer together with information on the photo specialists: Chrysobalanaceae, Ebenaceae, be available only to visitors or by loaning category. An ‘e’ is added to a field Meliaceae and Sapotaceae. Theoretically, specimens. Label data can be recorded in FOTOSUFFIX for labels, ‘e1’, ‘e2’ for any these families should have ‘better’ databases and specimen images taken with further label images (e = ‘etiqueta’ in determinations than those available in digital cameras, and the whole can be made Portuguese). Within the image RDE file, Amazonia. When there is an apparent available via the web. The dynamics of using a BRAHMS processing function, the conflict between identifications between consulting herbaria is changed rapidly. image files are then physically renamed duplicates, one needs to make an educated However, the processes of preparing the using a combination of the herbarium name, decision about which is more authoritative, data and images can be anything but rapid barcode and the FOTOSUFFIX. Thus and especially check if the duplicates really and can be costly. At the end of January ‘FHO_00012345’ is an image of the whole are from the same collection; images help 2008, I visited the Oxford University exsiccata with barcode 00012345 and enormously with this task. Herbaria (OXF & FHO) for six days to test ‘FHO_00012345_e’ is its label. In the brief visit made to Oxford, some a methodology that could accelerate the Step 3: A further BRAHMS routine is used 9,000 images were taken which will result databasing process and, at the same time, to process the image RDE file directly into a in c. 4,000 specimens being databased in repatriate important information to the collection RDE file, gathering all the images Oxford, and the data for some 7,000 countries where the plants were originally associated with each specimen and linking specimens improved in the Amazonian collected. them to a single record for each barcode. herbaria. It is unlikely that there will be Amazonian Brazil has three medium-sized This produces an RDE without any data, but large numbers of changes in identifications, herbaria (180,000 – 220,000 specimens) and with images. Thus, if a specimen has three since Iain Prance and Terry Pennington about seven smaller ones. All the herbaria images, the collection RDE file will have have been active visitors to Amazonia have been databased using BRAHMS. This one record with three images added to the herbaria. If this methodology was applied at is an unusual situation - a region the size of RDEIMAGES memo field. New York or Kew it would make a huge Amazonian Brazil has all its local herbarium Step 4: In the collection RDE file, one can contribution to the quality of the Amazonian data in the same system. BRAHMS is also view the label image(s) and type in the data herbaria. used widely in other regions of Brazil and in directly. This has advantages. The typing Many thanks to Colin, Denis, Serena and several other South American countries – a can be done anywhere, not just in the Stephen and for making my visit enjoyable. significant achievement for the Department herbarium, and you are no longer obliged to This work was undertaken as part of the of Plant Sciences in Oxford. enter the data in taxonomic order. If Brazilian Government’s Biodiversity Res- As botanists often collect plants in collector name and number information is earch Program (PPBio) which has, as one of duplicates, there can be considerable over- entered first, you can order the records in its objectives, the dissemination of lap of content between herbaria. One would the manner they were originally collected, collection data on the web. expect to find most of the Amazonian which can save hours of repetitive typing of geographical data and reduce typing errors. Mike Hopkins, INPA, Manaus, Brazil

Department of Plant Sciences, University of Oxford 11

The data are then connected to GBIF via the A given BOL homepage can be associated Publishing online from BioCASe provider package which must be with multiple databases, each maintaining a BRAHMS installed locally. separate identity on the server but searched http://www.gbif.org/dataproviders/howto . as a whole. In this way, regional curation networks and other logical groups of Herbaria are under increasing pressure to databases can be established. publish their specimen data and images online and research projects often have this as a funding prerequisite. One of the priorities with BRAHMS 6, published January 2008, has been to simplify online publishing, enabling those with and without internet skills and/or server resources to present their data online with minimum fuss via a project-tailored interface. BRAHMS facilitates publishing online in three ways: • Projects can contribute data to global Details describing your institution, IPR, legal and on-line networks including the Global data restrictions, and acknowledgements are The BOL update service opened in BRAHMS 6 Biodiversity Information Facility registered within BRAHMS and submitted as part showing the homepage text editing section. HTML (GBIF) via a locally installed provider of the ABCD output. tag help is provided for text editing. From this service. Explicit links are provided dialog, homepages are designed, data uploaded using BioCASE Provider Software and homepage-database links edited. BRAHMS online (BOL) (BPS). Herbaria and projects often need to publish • Institutions using BRAHMS for data online with a more personal research projects, but with their own institutional or project specific interface. For in-house online service, can upload example, many herbaria wish to disseminate relevant data to this service using one and promote their holdings and provide of the BRAHMS transfer mechanisms. web-based curation services; a cluster of • Projects can publish directly to the herbaria may wish to have their data SQL Server BRAHMS online (BOL) accessed via a common portal thus service. providing a regional online service; or a Some institutions or projects may have good research project may wish to publish data reason to opt for all three of these options for a taxon group or geographic area. taking the view that the wider the access to To facilitate this, an online publishing their data, the better. service has been fully integrated with BRAHMS 6 with options to develop the Publishing to GBIF online interface ( i.e. homepage), upload data The Global Biodiversity Information as required and configure this to provide the An example BOL service for Gabon developed at Facility ( www.gbif.org ) is a mega-science required service. Some BRAHMS online Wageningen. The website links to 90% of all project to make the world's primary data on facts: plants known to have been collected from Gabon. biodiversity freely and universally available BOL is an SQL Server system by default http://dps.plants.ox.ac.uk/bol/?gabon. via the Internet . Selected data can be running on the BRAHMS server. BOL can uploaded from BRAHMS to GBIF (and be installed on your own Windows server. other portals such the Biological Collection It includes an ‘update service’, operated Access Service for Europe - BioCASe) by from within BRAHMS 6 that links to the installing a provider service on your local designated server. server. BRAHMS 6 has an export function The BRAHMS 6 update service is used to to generate tables conforming to ABCD configure a website and the attached data. (Access to Biological Collections Data) Homepage text and design images can be v.2.06. ABCD is a common data uploaded and configured with HTML tags. specification for biological collection units Style sheets can be applied. Search based on the GBIF infrastructure. parameters and result presentation can be (http://www.bgbm.org/tdwg/codata/schema/ ). selected for the data categories taxa; specimens/observations; seed records; living collections; DNA samples; and literature. Sites can be published in English, French, Spanish, German and Chinese. Data uploading restrictions may be A multiple database example. The Brazilian Government's Biodiversity Research Program applied, for example, excluding or reducing (PPBio) has established a BOL service providing data for protected taxa. When uploading, access to specimen data and specimen/label data are compressed locally, streamed to the images from nearly all herbaria of the Amazonian server then installed and stamped with the region of Brazil, initially some 270,000 specimens provider database identity. Data already on with 300,000 associated images. When made the server with the same database identity public shortly, the service will provide un- are replaced. Online data can be updated on precedented access to botanical data from the any selected time basis. Databases can also region, a boost to research and regional curation and should attract other herbaria with Amazonian be removed from the server or temporarily holdings to contribute further data and images. hidden. Denis Filer, Research Associate

12 Oxford Plant Systematics OPS 15 May 2008