Ecological
Archives E085-100-A1
Mark Vellend. 2004. Parallel effects
of land-use history on species diversity and genetic diversity of forest herbs.
Ecology VOL: pp-pp.
Appendix A. Additional details on
the study system and methods.
Trillium grandiflorum
biology
T. grandiflorum has a widespread
distribution in eastern North America, extending from Minnesota, Ontario and
Québec in the north, to Georgia and Alabama in the south (Gleason and
Cronquist 1991). On each plant, typically one shoot (occasionally two)
emerges from a rhizome in early spring, and each shoot may (or may not) bear
a single flower. Pollination is mainly by bees (Irwin 2000), and individuals
are largely self-incompatible, though the mechanism of self-incompatibility
is considered "leaky" in that some seeds can be produced by self-pollination
(Sage et al. 2001). Each flower produces a single capsule usually with
about 15–30 seeds; each seed bears a lipid-rich appendage called an elaiosome
to which ants are attracted, and ants are the primary agent of seed dispersal,
though occasional seeds can be moved long distances (several km) via ingestion
and defecation by white-tailed deer (Vellend et al. 2003). There is no
long-term seed bank (Hanzawa and Kalisz 1993).
Forest stand characteristics
During an initial survey of potential
study sites, I found 10 isolated secondary forests with T. grandiflorum present,
all of which were included in this study. These stands all occurred on
former agricultural fields abandoned sometime before 1938 (the time of the earliest
aerial photographs – see Smith et al. 1993), and evidence of pit-and-mound
microtopography in some parts of all secondary stands suggested an agricultural
history of pasturing, or at least a history that did not include plowing (Marks
and Gardescu 2001). Secondary stands represent a relatively narrow range
of areas (1.5–4.4 ha) because larger secondary stands tend to have at least
one side adjacent to primary forest. Stands were chosen so that primary
and secondary forests occurred on the same range of soil associations, and a
multivariate analysis of variance on three principal components axes extracted
from the soil nutrient data showed no significant differences between primary
and secondary stands (P > 0.5). Primary stands were
also selected to represent a wide range of areas (0.1–33 ha), encompassing
those of secondary stands. The three smallest primary stands were hedgerows,
which function in this landscape as very small elements of primary forest with
respect to forest-herb distributions (Corbit et al. 1999).
Surveys of vegetation and Trillium
populations
For the surveys of forest-herb presence
within subplots in each stand, larger stands contained six parallel transects;
smaller stands (<2 ha) contained 1–5 transects depending on the area
and shape of the stand (e.g., plots were placed along one transect down the
middle of the hedgerows).
During surveys conducted to estimate
T. grandiflorum population size, I counted only three-leaved plants (young
seedlings have only one leaf and can be difficult to locate amidst other vegetation),
and I distinguished reproductive from non-reproductive individuals. Because
estimates of the total number of individuals and the number of reproductives
were highly correlated (r = 0.92), in this paper I report only
the total population size estimates. I aimed to count 
100 individuals, with overall population size calculated as the mean density
within the plots multiplied by the area over which T. grandiflorum was
distributed. When T. grandiflorum was more-or-less evenly distributed
throughout larger stands, three 2 × 100 m belt transects were
positioned at random along evenly-spaced lines traversing the stand. If
density was high, 8–10 random positions were chosen along each transect,
and all individuals were counted in 0.5–2 m2 plots at these
positions. If density was low, all individuals were counted in the entire
2 × 100 m area of each belt transect. Shorter transects
were used in stands where linear dimensions were <100 m. For stands
in which T. grandiflorum was present in only a small area (or areas)
of the stand, these areas were delimited, and random 1-m2 plots sampled
until at least 100 individuals had been counted. In the two smallest populations,
I counted all the individuals I could find.
Allozyme analysis
Following field collection, leaf
samples were stored at 4oC for no more than 4 days prior to processing
for genetic analysis. Half of each sample was ground with mortar and pestle
in an extraction buffer with 0.2 M Tris HCl, 0.7 mM Borax, 4 mM sodium metabisulfite,
40 mM sodium diethyldithiocarbonate, 50 mM sodium ascorbate, 11 mM DTT, and
3 mM PVP-40 (Griffin and Barrett 2004). The resulting extract was absorbed
onto filter paper wicks, and frozen at -80oC prior to electrophoresis.
Allozyme variation was assayed at four loci in all 40 individuals per population
using standard methods of starch gel electrophoresis and specific enzyme activity
staining (Wendel and Weeden 1989). Malate dehydrogenase (MDH), phosphogluconate
dehydrogenase (PGD), and phosphoglucoseisomerase (PGI) were run on a morpholine
citrate gel system at pH 6.1 (Wendel and Weeden system no. 2), and menadione
reductase (MNR) was run on lithium borate – tris citrate gel
system (Wendel and Weeden system no. 6; see also Broyles et al. 1997).
Chloroplast DNA analysis
One region of the T. grandiflorum
cpDNA genome (~2.4 kbp) was amplified using the trnH-trnK primer
pair from Demesure et al. (1995). The PCR product for each individual
was digested separately with two restriction enzymes, AluI and MspI
(New England BioLabs, Inc., Beverly, Massachusetts, USA), and the resulting
fragments were run out on 1.5% agarose gels. DNA fragments revealing one
length polymorphism (best seen with AluI-created fragments), and one
restriction site polymorphism (MspI), were visualized by staining the
gels with Ethidium Bromide, and viewing under UV radiation (see also Griffin
and Barrett 2004).
Soil
nutrient analysis
The sensitivity of the soil NO3-
and NH4+ analyses was insufficient for detection in the
forests sampled in this study, though a previous, more sensitive analysis indicated
that NO3- and NH4+ were strongly
correlated with the first principal component axis describing variation in the
same soil variables across a range of forest sites in Tompkins County (|r| >
0.7; K. M. Flinn, M. Vellend, and P. L. Marks, unpublished data).
NO3- and NH4+ were not included
in the principal components analysis of soil data in this study.
The degree of soil drainage at each
site was scored on a scale from 1–4: very poor, poor, moderate, and well-drained,
respectively, using a digital map (Tompkins County Planning Department 2000)
and descriptions (Neeley 1965) of soil associations in Tompkins County.
When more than one soil association occurred at a given site, a weighted average
(based on area) was calculated as the site level drainage value.
Prior to the PCA of soil data, all
variables except pH, loss-on-ignition and drainage were either log- or square
root-transformed to reduce skew in their distributions. The first two
PCA axes were rotated using the VARIMAX method (SAS PROC FACTOR) to maximally
align variables with PCA axes. Axis I was strongly correlated with pH
(r = 0.96), Ca (0.89), Mg (0.83), Cu (0.75), Al (-0.97), Fe
(-0.87), and Zn (-0.77). Axis II had moderate to strong correlations with
drainage (-0.59) and LOI (0.82).
LITERATURE
CITED
Broyles, S. B., S. L. Sherman-Broyles,
and P. Rogati. 1997. Evidence of outcrossing in Trillium erectum and
Trillium grandiflorum (Liliaceae). Journal of Heredity 88:325–359.
Corbit, M., P. L. Marks, and S.
Gardescu. 1999. Hedgerows as habitat corridors for forest herbs in central New
York, USA. Journal of Ecology 87:220–232.
Gleason,
H. A., and A. Cronquist. 1991. Manual of vascular plants of northeastern United
States and adjacent Canada, Second edition. New York Botanical Garden, New York,
New York, USA.
Griffin, S. R., and S. C. H. Barrett.
2004. Post-glacial history of Trillium grandiflorum (Melianthiaceae)
in eastern North America: inferences from phylogeography. American Journal of
Botany, in press.
Hanzawa, F. M., and S. Kalisz 1993.
The relationship between age, size and reproduction in Trillium grandiflorum
(Liliaceae). American Journal of Botany 80:405–410.
Irwin, R. E. 2000. Morphological
variation and female reproductive success in two sympatric Trillium species:
evidence for phenotypic selection in Trillium erectum and Trillium
grandiflorum (Liliaceae). American Journal of Botany 87:205–214.
Marks, P.L., and S. Gardescu. 2001.
Inferring forest stand history from observational field evidence. Pages 177–198
in D. Egan, and E. A. Howell, editors. The historical ecology handbook:
a restorationist’s guide to reference ecosystems. Island Press, Washington,
DC, USA.
Tompkins County Planning Department.
2000. Soil Associations (ARC Export 1965). Tompkins County ITS GIS Division,
Tompkins County Planning Department, Ithaca, New York, USA.
Vellend, M., J. A. Myers, S. Gardescu,
and P. L. Marks. 2003. Dispersal of Trillium seeds by deer: implications
for long distance migration of forest herbs. Ecology 84:1067–1072.
Wendel, J. F., and N. F. Weeden.
1989. Visualisation and interpretation of plant isozymes. Pages 46–72 in
D. E. Soltis, and P. S. Soltis, editors. Isozymes in plant biology. Dioscorides
Press, Portland, Oregon, USA.
[Back to E085-100]