|
|
||||||||
University of Alberta, Edmonton, Alberta, Canada T6G 2E9
Received for publication December 21, 1999. Accepted for publication May 30, 2000.
| ABSTRACT |
|---|
|
|
|---|
Key Words: Lake Athabasca Salix sand dune vessel element willow wood anatomy
| INTRODUCTION |
|---|
|
|
|---|
10 000 yr ago (Raup and Argus, 1982
|
There have been no detailed structural investigations of Lake Athabasca's endemic willows. Structural information may provide insight into functional strategies of these endemic sand dune species. The habitat of the Athabasca sand dunes can be described as desert-like, with endemic willows occurring on large expanses of open, actively blowing sand. Dune slacks are critical areas for establishment of endemic willows (Raup and Argus, 1982
). If a willow seedling survives sand accretion, it eventually becomes a woody shrub on the open sands. Endemic willows are rooted in the water table, which may be several metres below the sand surface (G. Argus, personal communication). Overall, the open sand habitat can be described as xerophytic, although these endemic shrubs have access to water from the water table and precipitation. Weather data extrapolated from Uranium City, Saskatchewan indicate that the Athabasca sand dune region receives
3640 cm of precipitation annually, of which
12.5 cm is rain during the growing season (Raup and Argus, 1982
).
Endemic willows (S. brachycarpa Nutt. var. psammophila Raup, S. planifolia ssp. tyrrellii, S. silicicola, and S. turnorii Raup) are confined to open sand areas on the dune interiors. However, S. planifolia ssp. tyrrellii is also found along the shoreline of Lake Athabasca, occurring sympatrically with its putative sister species, S. planifolia Pursh ssp. planifolia. Thomson Bay is a site on Lake Athabasca's shoreline where these two willows co-occur, and collections were taken from this site, while other endemics were collected from the Yakow Lake dune (see Fig. 1). The progenitor taxa are thought to be S. brachycarpa Nutt. var. brachycarpa, S. planifolia ssp. planifolia, S. alaxensis, and S. lutea Nutt., respectively.
The primary objectives of this study were: (1) to evaluate quantitatively and qualitatively vessel element (VE) characters in endemic willows, (2) to compare VE structure in endemic willows with that of their associated boreal or arctic progenitors, (3) to relate VE structural information in endemic willows to functional significance in this northern, open sand habitat, and (4) to address taxonomic implications of structural data in each endemicprogenitor pair.
This study is the first extensive investigation of internal structure of the Lake Athabasca sand dune endemic willows as compared with their closely related boreal or arctic putative sister species. The evaluation of adaptive VE characters in endemic willows will lead to a greater understanding of the role they play in survival in a northern, open sand habitat. The open sand habitat may be more suited to specialization in water conduction than the habitats occupied by progenitors. In particular, we chose to examine VE characters correlated with water conduction (Carlquist, 1988
). The Lake Athabasca sand dune habitat provides a unique opportunity to investigate structural adaptation in recently derived plant taxa.
| MATERIALS AND METHODS |
|---|
|
|
|---|
Ten or more stems from each species were softened in boiling water. Cross-sectional and radial-longitudinal sections were made using fresh razor blades. Sections were air-dried, coated with gold-palladium, and observed using a JEOL JSM-6301FXV scanning electron microscope.
Stems were also prepared for light microscopy (LM). After boiling, at least five stems per taxon were cut in cross section and stored in 70% ethanol. Sections (1524 µm in thickness) were prepared using a sliding microtome, mounted on glass slides, and stained with Safranin O/Light Green using a protocol modified from Johansen (1940)
.
General observations of VE characters were made. Fields of view (at 300x magnification) were examined from at least ten stems of each species. One to three fields of view (per stem) from electron micrographs were used to obtain quantitative VE data, as well as qualitative observations. Values for VE density (number of VE/mm2) and VE clustering (number of VE/group) were obtained from electron micrographs.
For measuring VE lumen diameter and VE length, LM was employed. Vessel element lumen diameter data were obtained from cross sections. Lumen diameter values were determined from early wood vessels in the most recent growth increment in at least five stems from each willow species. Measurements were taken at the widest region of the lumen.
Jeffrey's method (Johansen, 1940
) for wood macerations was used to obtain VE length measurements. Five stems from each taxon were macerated, and at least 25 vessel elements were measured per stem. Macerations were viewed using differential interference contrast (DIC) microscopy.
Statistical t tests and ANOVA were used to analyze quantitative VE data in endemicprogenitor pairs.
| RESULTS |
|---|
|
|
|---|
|
|
Lumen diameters for S. planifolia ssp. planifolia were significantly greater than the endemic ssp. tyrrellii, but there was no significant difference in lumen diameter between the endemic ssp. tyrrellii and the sympatric ssp. planifolia from Thomson Bay. In addition, the widespread ssp. planifolia was significantly greater in VE lumen diameter than the Thomson Bay ssp. planifolia (Figs. 46).
Salix alaxensis had greater VE lumen diameters than that of the derived endemic, S. silicicola, although this was not statistically significant. Lumen diameter, however, for VEs of S. lutea was significantly less than that of the closely related endemic, S. turnorii (Figs. 7, 8).
Vessel element clustering
Vessel element clustering varied among all willow taxa in this study. The S. brachycarpa endemicprogenitor pair had similar clustering values, although the endemic var. psammophila had more VEs per group than its associated putative sister species, var. brachycarpa.
The number of VEs per group was greater in the endemic, S. planifolia ssp. tyrrellii, than in the sympatric ssp. planifolia. Clustering values for the endemic ssp. tyrrellii and the sympatric progenitor ssp. planifolia were greater than that of the widespread ssp. planifolia, but no statistical significance was detected.
A significant statistical difference was observed between S. alaxensis and its derived endemic, S. silicicola, as the progenitor had a greater VE clustering value than the endemic (Figs. 9, 10). Salix lutea had a smaller VE clustering value than the endemic S. turnorii, but this difference was not statistically significant.
Vessel element length
Vessel element length varied within endemicprogenitor pairs, and some of these differences were significant. Salix brachycarpa var. brachycarpa had an average VE length value that was significantly less than that of the endemic var. psammophila (Figs. 11, 12). In the S. planifolia endemicprogenitor group, both the endemic ssp. tyrrellii and the widespread ssp. planifolia had similar values for VE length. The sympatric progenitor S. planifolia ssp. planifolia from Thomson Bay had the smallest average VE length. No significant difference was observed between the VE length values of S. silicicola and its progenitor, but the average value was greater in S. silicicola. Average VE length for the progenitor S. lutea was significantly less than that of its derived endemic, S. turnorii (Figs. 13, 14).
| DISCUSSION |
|---|
|
|
|---|
Correlations with habitat
In mesic habitats, tension is lower in conductive systems, so VEs tend to be longer and wider; in xeric environments, VEs are generally shorter and narrower (Carlquist, 1975
). In our study, results do not reveal distinct trends when comparing VEs from endemic willows with those of the respective boreal or arctic putative progenitor. Overall, endemics had longer VEs than their progenitors, with the exception of the allopatric widespread S. planifolia ssp. planifolia. The widespread progenitor ssp. planifolia had a similar VE length value compared to the endemic ssp. tyrrellii. Based solely on VE length data, the belowground environment of the open sands may be considered mesic as willows are rooted in the water table.
Endemics appear to have greater VE clustering than the progenitors, with the exception of Salix silicicola. This endemic had significantly fewer VEs per group than its widespread progenitor, S. alaxensis. As many of the progenitors were collected from peatlands and riparian sites, the significant difference in VE length between S. silicicola and S. alaxensis may be due to growth habit rather than habitat. Salix alaxensis has a rather prostrate growth form (at its collection site by a stream), while the habit of S. silicicola is upright.
Studies have shown that VE diameter can decrease due to water stress (Vitis; Lovisolo and Schubert, 1998
), as well as low-watering and low-nutrient regimes (Cereus; Arnold and Mauseth, 1999
). The sandy substrate does not hold a lot of water; the only standing water on the open sands is that of the exposed water table. Thus, one might expect the endemic willows to have narrower vessels than their putative progenitors due to this sandy substrate, which is also low in nutrients (S. E. Macdonald, personal communication, University of Alberta). This hypothesis can be supported by data from two of the four Lake Athabasca endemic willows. Both Salix planifolia ssp. tyrrellii and S. silicicola had narrower VE lumen diameters than their widespread sister species. In contrast, the endemic S. brachycarpa var. psammophila had a greater average VE lumen diameter than its sister taxon, but no statistical significance was detected. Salix turnorii had a significantly greater VE lumen diameter than its progenitor, S. lutea. For the endemic S. turnorii, a wider lumen may be a structural adaptation to the open sand environment.
Three endemic willows had VE densities less than that of their associated sister species; however, none of these differences was statistically significant. The only endemic to have a greater VE density than its progenitor was Salix brachycarpa var. psammophila, but the density values were quite similar for each.
Interestingly, the sympatric progenitor from Thomson Bay, S. planifolia ssp. planifolia, had a greater VE density than both the endemic ssp. tyrrellii and the allopatric widespread ssp. planifolia. The sympatric ssp. planifolia also had the smallest VE length of the three subspecies. During collection of the sympatric ssp. planifolia from Thomson Bay, stem cuttings were taken from shrubs that had fewer stomata on adaxial leaf surfaces. The endemic ssp. tyrrellii is amphistomatic (Argus and Steele, 1979
; Cooper and Cass, unpublished data), but ssp. planifolia has few to no stomata on the adaxial leaf surface. At this site, it may be possible that we collected stem cuttings from hybrids as well as ssp. planifolia. The two subspecies are quite similar morphologically, and the stomatal character has been useful in elucidating the identity of each (Argus and Steele, 1979
; R. L. Cooper and J. Gould, personal observation). Furthermore, Orians et al. (1999)
suggested that increased water availability may enhance relative performance of Salix hybrids. We feel that we collected pure ssp. planifolia from this shoreline site; however, the structural data from hybrid stem cuttings may explain the variation seen in the data sets for VE density and VE length. If the latter is true, additional morphological characters need to be evaluated to properly identify the subspecies of S. planifolia at the shore of Thomson Bay.
Taxonomic implications
The similarities of VE character data values within endemicprogenitor pairs illustrate the close taxonomic relationships between the Lake Athabasca sand dune endemics and their widespread boreal or arctic putative sister species. However, there were some notable differences among endemicprogenitor pairs. Salix brachycarpa var. psammophila and its progenitor had the highest VE density values of all willows in this study. In our opinion, this is most likely indicative of the taxonomic position of the S. brachycarpa endemicprogenitor pair. In a recent classification of New World Salix, Argus (1999)
divided Salix into four subgenera. Salix brachycarpa var. psammophila and var. brachycarpa were placed in Salix subgenus Chamaetia (Dumort.) Nasarov, while the other three endemicprogenitor pairs were placed in Salix subgenus Vetrix (Dumort.) Dumort (Argus, 1999
). We believe that little to no detectable variation in VE characters reflects the close taxonomic relationships within endemicprogenitor pairs. Furthermore, internal structural similarities support the putative evolutionary relationship between derived, endemic willows from the Athabasca sand dunes and their widespread sister species.
| FOOTNOTES |
|---|
2 Author for reprint requests. ![]()
| LITERATURE CITED |
|---|
|
|
|---|
, and J. W. Steele. 1979 A reevaluation of the taxonomy of Salix tyrrellii, a sand dune endemic. Systematic Botany 4: 163177[CrossRef][ISI]
Arnold, D. H., and J. D. Mauseth. 1999 Effects of environmental factors on development of wood. American Journal of Botany 86: 367371
Carlquist, S. 1975 Ecological strategies of xylem evolution. University of California Press, Berkeley, California, USA
. 1988 Comparative wood anatomy: systematic, ecological, and evolutionary aspects of dicotyledon wood. Springer-Verlag, New York, New York, USA
Hermesh, R. 1972 A study of the ecology of the Athabasca sand dunes with emphasis on the phytogenic aspects of dune formation. M.Sc. thesis, Department of Plant Ecology, University of Saskatchewan, Saskatoon, Saskatchewan, Canada
Johansen, D. A. 1940 Plant microtechnique. McGraw-Hill, New York, New York, USA
Lovisolo, C., and A. Schubert. 1998 Effects of water stress on vessel size and xylem hydraulic conductivity in Vitis vinifera L. Journal of Experimental Botany 49: 693700
Orians, C. M., D. I. Bolnick, B. M. Roche, R. S. Fritz, and T. Floyd. 1999 Water availability alters the relative performance of Salix sericea, Salix eriocephala, and their F1 hybrids. Canadian Journal of Botany 77: 514522[CrossRef]
Purdy, B. G., and R. J. Bayer. 1995a Allozyme variation in the Athabasca sand dune endemic, Salix silicicola, and the closely related widespread, S. alaxensis. Systematic Botany 20: 179190
, and . 1995b Genetic diversity in the tetraploid sand dune endemic Deschampsia mackenzieana and its widespread progenitor D. cespitosa (Poaceae). American Journal of Botany 82: 121130[CrossRef][ISI]
, and . 1996 Genetic variation in populations of the endemic Achillea millefolium ssp. megacephala from the Athabasca sand dunes and the widespread ssp. lanulosa in western North America. Canadian Journal of Botany 4: 11381146
, , and S. E. Macdonald. 1994 Genetic variation, breeding system evolution, and conservation of the narrow sand dune endemic Stellaria arenicola and the widespread S. longipes (Caryophyllaceae). American Journal of Botany 81: 904911[CrossRef][ISI]
Raup, H. M. 1936 Phytogeographic studies in the Athabaska-Great Slave Lake region. I. Catalogue of the vascular plants. Journal of the Arnold Arboretum 17: 180315
, and G. W. Argus. 1982 The Lake Athabasca sand dunes of northern Saskatchewan and Alberta, Canada: I. The land and vegetation. National Museum of Natural Sciences Publication number 12, Ottawa, Ontario, Canada
Smith, D. G. 1978 The Athabasca sand dunes: a physical inventory. Indian and Northern Affairs National Parks Branch Contract number 77-31, Ottawa, Ontario, Canada
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |