dc.description.abstract | The study measured landscape level diversity of the herbaceous understory of
mature, upland forests in north central Wisconsin across a moisture-nutrient gradient.
Sites were located in Region Three of the Kotar et al. (1988) habitat typing system, which
includes Vilas, Oneida, Florence, and Forest Counties. Eight habitat types were used to
segregate the landscape along a moisture-nutrient gradient. There were 40 sites, with five
replicate stands for each habitat type, except TMC which had four sites and A ViO which
had six. All sites were reproductively mature, had closed canopies, were undisturbed for
at least 20 years, and were at least 8 ha in size.
Ocular estimation of percent cover of groundlayer species was performed in
randomly located one meter square plots along two or three randomly placed 25 meter
transects during June and August of 1995. Shrub cover was estimated by line intercept.
Sampling intensity was determined from data collected in 1994. Data were analyzed at the
gamma (landscape), beta (among site), and alpha (within site) levels; early and late
summer data were treated separately. Gamma diversity was quantified using a new
method, affinity analysis, which generates a list of modal and outlier sites and calculates
mosaic diversity, a measure of landscape complexity. Modal communities are rich in
species common to the region, while outlier communities are rich in rare species, or are
species poor (Scheiner, 1992). The diversity indices used in the analysis were also
compared for their efficacy.
Generally, communities in the middle of the moisture-nutrient gradient were
modal, while those at the mesic end of the gradient were outlier. Hemlock (Tsuga
canadensis) dominated communities were exclusively outlier. Mosaic diversity values,
with jackknifed standard errors, were 2.88± 0.04 for early summer, 2.95±0.03 for late
summer, and 3.95±0.07 for all data combined. A paired t-test detected a significant
difference between early and late summer values (p</=0.01). Whittaker's Index (beta
diversity) revealed varying rates of species turnover along both the moisture and nutrient
gradients. Species densities (richness) were relatively constant. Rank abundance plots
(evenness) indicated that communities were composed mainly of species represented by a
few individuals. A one-way analysis of variance of Shannon-Weaver values uncovered no
significant differences among site types (p>/= 0.05).
Regional diversity mainly resulted from high beta values (mesic sites). The
sources of high beta values were the moisture and nutrient gradients because they
contributed significantly and approximately equally to differences in composition among
stands. The mechanisms creating differences in composition along the moisture-nutrient
gradient probably were niche partitioning, disturbance and succession, site history, and
differences in life history strategies. The influence of these factors on composition
probably varied in importance along the moisture and nutrient gradients. Overall, this
landscape was about as diverse as other temperate mixed forest ecosystems (Scheiner and
Rey Benayas, 1994).
Affinity analysis is unique because it is the first known diversity index to quantify
landscape level diversity. A recommended combination of indices is a landscape level one,
such as affinity analysis, along with species density (if sample sizes are unequal) and
evenness. These indices together provide the researcher with a working knowledge of the
landscape and key factors influencing it. | en_US |