HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

Electronic Letters to:

Systematics: Susanne S. Renner, Li-Bing Zhang, and Jin Murata A chloroplast phylogeny of Arisaema (Araceae) illustrates Tertiary floristic links between Asia, North America, and East Africa Am. J. Bot. 2004; 91: 881-888 [Abstract] [Full text] [PDF]

eLetters: Submit a response to this article

Electronic letters published:

Re: Fossil identifications used in molecular dating studies should be justified

Susanne S. Renner, Susanne S. Renner, Li-Bing Zhang, and Jin Murata   (23 September 2004)

Fossil identifications used in molecular dating studies should be justified

Elizabeth J. Hermsen, Maria A. Gandolfo   (23 September 2004)

Re: Fossil identifications used in molecular dating studies should be justified 23 September 2004
Susanne S. Renner, Professor University of Missouri - St. Louis, Susanne S. Renner, Li-Bing Zhang, and Jin Murata Send letter to journal: Re: Re: Fossil identifications used in molecular dating studies should be justified renner{at}umsl.edu Susanne S. Renner, et al.	The fossil (USNM 36885), whose identity is questioned by Hermsen and Gandolfo (2004) in their Letter to the Editor, is pictured in Fig. 1. Frank H. Knowlton (1860-1926), a paleobotanist with the U.S. Geological Survey who collected mainly in Colorado and Wyoming in 1896, described it in a book-length publication on the flora of the Middle Miocene Latah Formation near Spokane (Knowlton, 1926). Between June 2002 and June 2003, we discussed images of this fossil with aroid specialists and paleobotanists, and prompted by Hermsen and Gandolfo’s Letter, we did so again over the past weeks. Our views about this fossil obviously are our own. Hermsen and Gandolfo contend that this fossil is a Liquidambar (Altingiaceae), as did Brown (1946), and we therefore sent the images reproduced in Fig. 1 to K. Pigg (Schools of Life Sciences, Arizona State University, Tempe, Arizona) who just concluded a study of permineralized Liquidambar from Yakima Canyon, Washington (Pigg, Ickert-Bond, and Wen, 2004). The Yakima Canyon Liquidambar fossils are essentially the same age as the disputed taxon from the Latah Formation, and Pigg, Ickert-Bond, and Wen (2004) discuss the difficulties of interpreting spherical infructescences preserved as compression-impression remains. They show how flattening and weathering prior to fossilization can alter the appearance of extant Liquidambar fruits (l.c.: Figs. 6 and 10), causing loss of persistent styles and sometimes loss of the clear demarcation of the characteristic biloculate condition in the Altingiaceae. The identity of the fossil described by Knowlton (1926) appears equivocal to K. Pigg (email communication, 24 July 2004) because it lacks the persistent stigmas one would expect to find in Liquidambar (but not Altingia) as well as any detail of the biloculate fruits one would expect in Altingiaceae. As mentioned by Hermsen and Gandolfo, the same plate that shows Arisaema hesperia also illustrates an infructescence identified by Knowlton as Liquidambar (l.c.: pl. X, Fig. 10); this infructescence has beautiful persistent stigmas. Liquidambar leaves and infructescences are also well known at the Clarkia and Emerald Creek floras of northern Idaho (Smiley and Rember 1985; see Rember's website http://www.mines.uidaho.edu/~tertiary/images/dirfour/Liquidambar.jpg), and the Liquidambar infructescence illustrated on Rember's website likewise clearly shows the characteristic persistent styles expected in Liquidambar. Two other paleobotanists whom we consulted, S. Manchester (Curator of Paleobotany, Florida Museum of Natural History, USA) and V. Wilde (Forschungsinstitut Senckenberg, Frankfurt, Germany), after studying the images reproduced in Fig. 1 also arrived at the view that Knowlton’s Arisaema hesperia cannot reliably be assigned to Liquidambar. What about Knowlton’s assigment of this fossil to Araceae? Familial assignment of a fossil depends on a researcher’s accumulated expertise and image databanks. By the time he was working on the Latah fossils, Knowlton had the appropriate broad experience. With electronic communication, however, it is now possible for several researchers with different morphological and taxonomic expertise to consult about images and thus perhaps bring to bear more diverse background knowledge than may previously have been possible. In the specific case of Knowlton’s ‘Arisaema hesperia’, none of the extant botanists and paleobotanists whom we consulted suggested a different familial fit for this fossil than Araceae. This does not mean that any of them spontaneously suggested that this fossil might represent Arisaema. However, as discussed by Hermsen and Gandolfo (2004), the Latah strata from which this fossil comes are between 12 and 20 Ma old (we used 16-18 Ma in our molecular clock calibration [Renner, Zhang, and Murata, 2004 and below]; Graham [1999: p. 262] gives an age of 15.8 Ma). Climate at the latitude of Washington at that time was already similar to present-day conditions, and the Latah flora represents a mixed mesophytic forest, with some 75% of (today) Asian species (Graham, 1999: p. 260). We therefore checked the inflorescence and infrutescence morphology of all Araceae that today occur in North America. These are Arisaema, Calla, Colocasia, Lysichiton, Orontium, Peltandra, Pistia, and Symplocarpus. Calla has 6-9 stamens, and their scars would be visible or at least there would be wider spaces between the individual gynoecia or fruitlets in the fossil (Fig. 1). In Colocasia and Peltandra, the lower part of the spatha would persist at the base of the inflorescence or infrutescence. Pistia has very different inflorescences than the fossil, and in Symplocarpus, the gynoecia and styles are strongly pointed, which likewise does not fit with ‘Arisaema hesperia.’ The inflorescence of Lysichiton, finally, is a 10 cm long cylindric rod, rather than a globose structure. This leaves Orontium, which has thin, cylindric inflorescences that in fruit can turn into globose structures (see the fossil Orontium described by Cockerell [1926] from the Eocene, originally thought Miocene, of Florissant). Orontium would fit Arisaema hesperia in possessing a small corolla around each flower, which might correspond to the dark annulus around some of the globose gynoecia/fruitlets visible in Fig. 1. On the other hand, the subtending peduncle in Orontium is as broad as the inflorescence/infructescence, which does not fit with Arisaema hesperia with its quite narrow stem. Knowlton (1926: p. 29) writes about the peduncle “Of the supporting organ or peduncle 3.5 centimeters is preserved, but it was longer than this, as it runs off the matrix without evidence of attachments. It is slightly more than 2 millimeters broad and is striate or wrinkled longitudinally, a feature which indicates that it was fleshy.” The striate lines are visible on his photo (Fig. 1, left image), less so in the differently lit modern image of the fossil (Fig. 1, middle image). Hermsen and Gandolfo’s (2004) statement that the extent of Knowlton’s morphological analysis comparing USNM 36885 to extant Arisaema consists of the few lines that they quote is incorrect. (They also inadvertently dropped one word from their second Knowlton quote.) That Latah is a mixed mesophytic forest flora (Graham, 1999: p. 260) may also argue in favor of Arisaema, which is a woodland genus, rather than Orontium, a semi-aquatic or swamp taxon. In terms of geographic ranges, Arisaema today extends slightly further north than does Orontium (Mayo, Bogner, and Boyce, 1997). Smith and Stockey (2003) have recently analyzed the interior structure of Eocene Araceae, and we therefore asked them whether it might be possible to discern the seed structure of Arisaema hesperia (if this fossil could be sectioned). Selena Smith (Dept. of Biological Sciences, University of Alberta, email of 1 Sep. 2004) answered that Arisaema hesperia “looks like a compression/impression [fossil] and therefore not likely to have internal anatomy preserved. If one were to collect more fossils […] it might be worthwhile sectioning a specimen to see if more 3-D structure could be teased out.” Given the available evidence, Knowlton’s assignment of this fossil to us remains convincing. Even so, and to avoid having to rely on a single calibration point, we conducted a parallel study that used enlarged taxon and genome sampling and a Bayesian approach (Thorne and Kishino, 2002), permitting multiple simultaneous calibration time windows, with minimal and maximal bounds set by fossils or geologic events (Renner and Zhang, 2004). With the additional Araceae genera included it became possible to use two other fossils as minimal bounds (not absolute constraints): 60-my-old Peltandreae leaves from Europe, Kazakhstan, North Dakota, and Tennessee, and 45-my-old leaves from central Germany that closely match modern Colocasieae (Wilde, Kvacek, and Bogner, in press; details in Renner and Zhang, 2004). Alternative analyses furthermore explored the effects of using a fourth and fifth constraint, viz. a maximal bound of 85 my for the age of Protarum sechellarum based on the age of the Seychelles archipelago and an maximal bound of 100 my for the age of Pistia clade, based on the oldest fossil record of Arales. The Arisaema hesperia fossil thus became one of three to five constraints. Prompted by Hermsen and Gandolfo’s Letter we have rerun the Bayesian analysis again, using the settings described in Renner and Zhang (2004), but excluding the Arisaema hesperia fossil. The results show, as before, that the disjunctions between the Asian and North American species of Arisaema date to the Oligocene and Miocene. Specifically, the three North American species of Arisaema, A. dracontium, A. macrospathum, and A. triphyllum, appear to stem from at least two independent colonizations, one around 35-40 my (95% credibility interval 19-55 my), the other possibly as young as 12 my (range 2-28 my). The 95% credibility intervals for all estimates were tabulated in Renner and Zhang (2004). Given the large credibility ranges and the fact that modern approaches to molecular clock dating permit sliding times (such that a node constrained to a minimal age often comes out as much older), Hermsen and Gandolfo’s argument about the unknown date of the Knowlton fossil is beside the point. Nobody knows the precise age of Arisaema hesperia, and the time window that we used falls nicely within the range of dates cited by Hermsen and Gandolfo (2004). We agree with Hermsen and Gandolfo that skepticism is appropriate when dealing with pre-1970s assignments of fossils to modern taxa, especially of very old fossils. A good example is Brown’s (1962) assignment of Paleocene leaves from the Ferris formation in Wyoming to Melastomataceae (details in Renner, Clausing, Meyer, 2001). (This earlier experience with an assignment by Brown may have biased us somewhat against Brown’s reassignment of Knowlton’s fossil to Liquidambar.) Arisaema provides an excellent example of why reliance on several fossils, wherever available, is so desirable. With the electronic facilitation of image exchange between paleobotanists and workers specializing in extant plants, the flow of information will continue to accelerate. The real bottleneck, of course, lies in the time-consuming study of morphology and preparation of the fossils. Where fossil calibration remains problematic (which will often be the case), the new Bayesian methods, because of their computational speed, allow the use of multiple sliding calibration windows, which should reduce the effect of any single constraint and which allows testing the effects of including/excluding different calibration fossils or geologic events. Acknowledgements: We thank Josef Bogner, Kathleen Pigg, Steve Manchester, Volker Wilde, Selena Smith, and Jun Wen for detailed comments on Arisaema hesperia and Liquidambar, Scott Wing, curator of paleobotany at the Smithsonian Institution, Washington, DC, for two color images of the type of Arisaema hesperia, and Trish Consiglio for producing Fig. 1. BROWN, R. W. 1946. Alterations in some fossil and living floras. Journal of Washington Academy of Sciences 30: 344-355. BROWN, R. W. 1962. Paleocene flora of the Rocky Mountains and Great Plains. United States Geological Survey Professional Paper 375: 1-119. COCKERELL, T. D. A. 1926. A Miocene Orontium (Araceae). Torreya 26: 69. GRAHAM, A. 1999. Late Cretaceous and Cenozoic history of North American vegetation. Oxford University Press, New York. PIGG, K. B., S. M. ICKERT-BOND, AND J. WEN. 2004. Anatomically preserved Liquidambar (Altingiaceae) from the middle Miocene of Yakima Canyon, Washington state, USA, and its biogeographic implications. American Journal of Botany 91: 499-509. http://www.amjbot.org/cgi/reprint/91/3/499. HERMSEN, E., AND A. GANDOLFO. 2004. Letter to the Editor of American Journal of Botany. KNOWLTON, F. H. 1926. Flora of the Latah Formation of Spokane, Washington, and Coeur d’Alene, Idaho. United States Geological Survey Professional Paper 140A: 17-81, plates 8-31. MAYO, S. J., J. BOGNER, AND P. C. BOYCE. 1997. The genera of Araceae. Royal Botanic Gardens, Kew, London. RENNER, S. S., G. CLAUSING AND K. MEYER. 2001. Historical biogeography of Melastomataceae: the roles of Tertiary migration and long-distance dispersal. American Journal of Botany 88: 1290-1300. RENNER, S. S., AND L. ZHANG. 2004. Biogeography of the Pistia clade (Araceae): based on chloroplast and mitochondrial DNA sequences and Bayesian divergence time inference. Systematic Biology 53: 422-432. RENNER, S. S., L. ZHANG, AND J. MURATA. 2004. A chloroplast phylogeny of Arisaema (Araceae) illustrates Tertiary floristic links between Asia, North America, and East Africa. American Journal of Botany 91: 881-888. REMBER, W. C. Website: Tertiary Research Center. The Clarkia flora of northern Idaho. http://www.mines.uidaho.edu/~tertiary/images/dirfour/Liquidambar.jpg. SMILEY, C. J., AND W. C. REMBER. 1985. Physical setting of the Miocene Clarkia fossil beds, northern Idaho. Pp. 11-31 in C. J. Smiley, ed., Late Cenozoic History of the Pacific Northwest—Interdisciplinary Studies on the Clarkia Fossil Beds of Northern Idaho, Pacific Division of the American Association for the Advancement of Science, San Francisco. SMITH, S. Y., AND R. A. STOCKEY. 2003. Aroid seeds from the Middle Eocene Princeton chert (Keratosperma allenbyense, Araceae): Comparisons with extant Lasioideae. International Journal of Plant Sciences 164: 239-250. THORNE, J. L., AND H. KISHINO. 2002. Divergence time estimation and rate evolution with multilocus data sets. Systematic Biology 51:689-702. WILDE, V., Z. KVACEK, AND J. BOGNER. In press. Fossil leaves of Araceae from the European Eocene and notes on other aroid fossils. Int. J. Plant Sci. Figure 1. Left image: Knowlton’s (1926: pl. X, Fig. 1) photograph of Arisaema hesperica; middle image: Image of USNM 36885, the type specimen of Arisaema hesperica; right image: Infructescence of extant North American Arisaema triphyllum (photo G. Gusman).
Fossil identifications used in molecular dating studies should be justified 23 September 2004
Elizabeth J. Hermsen, Graduate Student Cornell University, Maria A. Gandolfo Send letter to journal: Re: Fossil identifications used in molecular dating studies should be justified ejh23{at}cornell.edu Elizabeth J. Hermsen, et al.	In their paper on Tertiary floristic links in the genus Arisaema, Renner et al. (2004) cite a putative Arisaema fossil from the Miocene Latah Formation, Spokane, WA, USA, published by Knowlton (1926) to calibrate their tree for molecular clock dating. The fossil cited by the authors of this paper is a single infructescence, which Knowlton originally described as “a globose head of fruits or ‘berries’ borne on a long, naked peduncle” (Knowlton, 1926, p. 29). Knowlton interpreted the fossil, which he figured as the sole specimen of the new species A. hesperia (Knowlton, 1926, pl. 10, fig. 1), as “closely similar to the fruit of the living Arisaema triphyllum (Linné) Torrey, the Indian turnip of the eastern United States” (Knowlton, 1926, p. 29). For calibration of one of 26 equally most parsimonious trees calculated using combined sequences of the trnL intron, trnL-F spacer, and rpl20-rps12 spacer, the node subtending A. triphyllum and A. amurense was fixed at 18 and 20 million years old (Renner et al., 2004, p. 884 and fig. 4, placement of the fossil indicated by an arrow), the older end of the absolute age range given by Renner et al. for the Latah flora. No citation was provided for the absolute age range given. The placement of A. hesperia on the phylogenetic tree is evidently in recognition of the close resemblance suggested by Knowlton between A. triphyllum and the fossil A. hesperia of the Latah Formation, as quoted above. The same fossil was used to provide a lower bound of 18 Ma for the node subtending A. triphyllum and A. amurense in Renner and Zhang (2004, p. 428 and fig. 3 node 3). The fossil taxon A. hesperia, typified by the specimen USNM 36885 (Knowlton, 1926, pl. 10, fig. 1), the only specimen of A. hesperia published, was removed to the genus Liquidambar in 1946 by Brown, who synonymized A. hesperia with Liquidambar pachyphyllum Knowlton, stating: “Arisaema hesperia Knowlton (34), p. 29, pl. 10, fig. 1 = Liquidambar pachyphyllum Knowlton. This specimen is a much worn sweetgum fruit as can be seen by comparison with the better preserved specimen (Fig. 10) shown on the same plate but identified by Knowlton as Liquidambar” (Brown, 1946, p. 352). Brown’s synonymy of A. hesperia with L. pachyphyllum was also catalogued by LaMotte (1952, p. 75, 205). Renner et al. (2004) do not address this reassignment in their paper, nor provide justification for preferring the interpretation of the fossil as an Arisaema infructescence or for its placement on the node subtending A. triphyllum on their phylogenetic tree. Even disregarding the reassignment of the fossil specimen to Liquidambar, Renner et al. should provide a justification for the placement of USNM 36885 on the node subtending A. triphyllum, as the extent of Knowlton’s morphological analysis comparing USNM 36885 to extant Arisaema, other than the fragment quoted above, consists of the following: “The fossil form has a longer, more uniform-sized peduncle than the living species. the head is globose rather than ovoid, and the berries are apparently wrinkled instead of being smooth and shining. Whether these differences are sufficient to warrant a generic separation seems doubtful, and if the form is correctly interpreted, as it seems to be, the presence of the genus in this flora is placed on secure footing” (Knowlton, 1926, p. 29, incorrect punctuation maintained). As pointed out by Wolfe (1973) a common method of identifying angiosperm megafossils, even up until 1973 and in some cases beyond, was “picture-matching,” which was predicated on the assumption that “1) For every given fossil leaf, there is one extant species that has leaves more similar to the fossil than does any other extant species; and 2) the fossil must belong to a species that is congeneric with that one extant species” (Wolfe, 1973, p. 337). When such a technique is employed, as Wolfe and Schorn (1990) point out, “The characters that result in the ‘match’ are typically not stated, and, even if they are, no evaluation is made as to whether the characters are taxonomically critical . . . If foliar characters (or characters of other organs) are described and compared, the descriptions and comparisons are typically phrased in undefined terms or terms of uncertain meanings” (Wolfe and Schorn, 1990, p. 1). The result is that many fossil floras, especially those from the nineteenth and early twentieth centuries, are badly in need of systematic revision. For instance, when Dilcher (1971) revised the Eocene flora of southeastern North America originally published by Berry (1916, 1930), he found that ca. 60% of familial and generic determinations were incorrect. “Picture-matching” is the technique exemplified in the quotations from Knowlton (1926) given above. As the fossil taxa used in calibration represent the lynchpin of molecular dating analyses, care must be taken that fossils used in molecular dating studies have been classified using rigorous comparative morphological and/or anatomical analyses, and, preferably, placed in cladistic analyses to determine their phylogenetic positions. If a fossil has been identified through “picture-matching” and never revised using more rigorous techniques, the specimen should be reexamined before use in a molecular dating study. Furthermore, geochronologic and stratigraphic context are fundamental to the interpretation of paleobiological data, and, therefore, a primary source or sources should be cited for the age determination of the sediments in which fossils used in calibration occur, and, if possible, the method of determination of the age of the sediments (e.g., correlation to other rock units, radiometric dating, palynology, etc.) should be listed, as well as the error on the absolute age used in calibration, if one has been calculated. In the case of the Latah Formation, Knowlton (1926) did not provide the absolute age range of 18 to 16 Ma given by Renner et al. (2004) in their paper, nor could he have provided it. Evernden and James (1964) calculated a K/Ar date of 14.5 Ma for the Spokane florule of the Latah flora (Evernden and James, 1964, p. 958, 969), with a sample taken from locality UCMP 3940, the Spokane Brickyard locality (UCMP On-line Database, 2004, http://elib.cs.berkeley.edu/ucmp/). Gray and Kittleman (1967) later gave a K/Ar date of 13.0 + or - 0.5 Ma for basalt overlying leaf- and pollen-bearing beds at Deep Creek Canyon (Gray and Kittleman, 1967, p. 270, 286), the same general area from which Knowlton (1926) reported that USNM 36885 was collected (Knowlton, 1926, p. 30). Gray and Kittleman (1967, p. 281) go on to state that Latah Formation sediments in the Spokane Valley range anywhere from at least 20.6 to 12.1 Ma based on K/Ar dates obtained from the localities they sampled, and, thus, fossils from the “Latah flora” in the Spokane region may fall anywhere in that age range, depending on where they were collected geographically and stratigraphically. The most recently published absolute age for the Latah flora of Spokane, Washington, and Coeur d’Alene, Idaho (the flora described by Knowlton, 1926), we could locate was ca. 15.8 Ma (Graham, 1999, p. 260, 262), though the method by which this age was arrived at was not clear from the text. BERRY, E. W. 1916. The Lower Eocene floras of southeastern North America. United States Geological Survey Professional Paper 91, 481 p. BERRY, E. W. 1930. Revision of the Lower Eocene Wilcox flora of the southeastern states, with descriptions of new species, chiefly from Tennessee and Kentucky. United States Geological Survey Professional Paper 156, 96 p. BROWN, R. W. 1946. Alterations in some fossil and living floras. Journal of Washington Academy of Sciences 36: 344-355. DILCHER, D. L. 1971. A revision of the Eocene flora of southeastern North America. The Palaeobotanist 20: 7-18. EVERNDEN, J. F., AND G. T. JAMES. 1964. Potassium-argon dates and the Tertiary floras of North America. American Journal of Science 262: 945-974. GRAHAM, A. 1999. Late Cretaceous and Cenozoic History of North American Vegetation. Oxford University Press, New York, USA. GRAY, J., AND L. R. KITTLEMAN. 1967. Geochronometry of the Columbia River Basalt and associated floras of eastern Washington and western Idaho. American Journal of Science 265: 257-291. KNOWLTON, F. H. 1926. Flora of the Latah Formation of Spokane, Washington, and Coeur d’Alene, Idaho. United States Geological Survey Professional Paper 140-A: 17-50 and 57-79, plates 8-29. LAMOTTE, R. S. 1952. Catalogue of the Cenozoic plants of North America through 1950. The Geological Society of America Memoir 51, 381 p. RENNER, S. S., AND L. ZHANG. 2004. Biogeography of the Pistia clade (Araceae): based on chloroplast and mitochondrial DNA sequences and Bayesian divergence time inference. Systematic Biology 53: 422-432. RENNER, S. S., L. ZHANG, AND J. MURATA. 2004. A chloroplast phylogeny of Arisaema (Araceae) illustrates Tertiary floristic links between Asia, North America, and East Africa. American Journal of Botany 91: 881-888. WOLFE, J. A. 1973. Fossil forms of Amentiferae. Brittonia 25: 334 -355. WOLFE, J. A., AND H. E. SCHORN. 1990. Taxonomic revision of the Spermatopsida of the Oligocene Creede flora, southern Colorado. United States Geological Survey Bulletin 1923, 40 p., 13 plates.