Oecologia (Berlin) (1988) 75:459464 Oecologia 9 Springer-Verlag1988 Correlations between genet architecture and some life history features in three species of Solidago B. Schmid*, G.M. Puttick**, K.H. Burgess, and F.A. Bazzaz Department of Organismic and Evolutionary Biology, Harvard University, Cambridge, MA 02138, USA Summary. Members of the genus Solidago are among the most widely studied model systems in plant population biol- ogy. A comparative study of Solidago canadensis, S. alt- issima, and S. gigantea in an experimental garden showed that the three species had different patterns of shoot growth and development, leaf morphology and physiology, and biomass allocation at harvest. These differences were also found in the field. Contrary to some current taxonomic usage, our results show that S. canadensis should ecologi- cally be treated as a separate taxon distinct from S. alt- issima, and that the latter may be grouped together with S. gigantea. Many of the biological differences between S. canadensis and the other two taxa, such as differential in- vestment into sexual reproduction versus clonal growth, may be explained by differences in genet architecture. These architectures concern high compared to lower within-genet shoot density resulting from differences in rhizome lengths among the taxa (shorter in S. canadensis than in S. altissima and S. gigantea). Key words: Clonal growth - Genet architecture - Leaf de- mography - Sexual reproduction - Solidago Species of the genus Solidago (Asteraceae) are among the most ecologically studied wild plants in North America (e.g. Werner et al. 1980; Hartnett and Bazzaz 1983, 1985a-c; Abrahamson and McCrea 1986; Schmid and Bazzaz 1987; and references therein). Nevertheless, little is known about their comparative biology, partly because of difficulties in these species' taxonomic affiliation, and because investiga- tors have usually worked with a single taxon. This is the case for Solidago canadensis L., S. altissima L. (= S. eana- densis var. seabra Torr. & Gray), and S. gigantea Ait., which are morphologically similar and taxonomically close- ly related. However, the rhizome systems of these three species have different architectures. Solidago canadensis has short rhizomes (rarely longer than 5 cm) and forms mono- clonal patches with high shoot density within genets, while S. altissima and S. gigantea have longer rhizomes (up to 20 cm or more), lower within-genet shoot densities, and * Present addresses: Botanisches Institut der Universitgt Basel, He- belstrasse 1, CH-4056 Basel, Switzerland ** Chemistry Department, Northeastern University, Boston, MA 02115, USA Offprint requests to: B. Schmid different genets may intermingle (G.D. Maddox et al., in preparation). Unfortunately, because of above-ground morphological similarities, many ecologists in the past have followed the traditional treatment and, in their work, may have used the name S. canadensis to mean S. altissima (Abrahamson and Gadgil 1973; Smith and Palmer 1976; Werner and Platt 1976; Stinner and Abrahamson 1979; Werner 1979, 1985; Hartnett and Bazzaz 1983, 1984, 1985a-c). In this paper we investigate phenologicai and physiolog- ical variation between and within these species in a common garden and in the field. We were particularly interested in: 1) whether relatedness among the three taxa as defined by their current taxonomic treatment (see Croat 1972; Abrahamson and McCrea 1986, and references therein) is matched by ecological similarity; 2) whether differences in genet architecture are correlated with and therefore may explain differences in life history characteristics among the three species; and 3) comparing patterns of ecological varia- tion in these species in the homogeneous environment of the common garden with patterns found in the more hetero- geneous environment of the field. Materials and methods On 21-22 May 1985 we collected connected systems of at least eight shoots with rhizomes and roots from similar- sized clones of S. canadensis and S. altissima at Broadmoor Wildlife Sanctuary and of S. gigantea at Pine Banks Park (Middlesex Co., Massachusetts). The multi-shoot systems were first separated into several connected pairs of nearly equal sizes by severing old rhizomes. Each pair consisted of the old stem base of the parent shoot and two ot its daughter shoots produced by current rhizomes. In addition, sister shoots in half of these pairs were severed from each other by cutting through the middle of the stem base re- maining from the parent shoot. All pairs, both connected and severed, were immediately transplanted to the experi- mental garden on Harvard University campus. Shoots of the same clones were put together in 20-cm diameter clumps of four or eight, giving low or high density treatments. We collected experimental material relatively late (21-22 May), although plants had resumed growth by the end of April at the field sites. Selection of more homogeneous shoots is easier at this later stage, and previous experience with Solidago has shown that both simultaneous budbreak and bolting are difficult to achieve after earlier transplanta-