species richness between natural marshes and each of the natural forested wetland types,
 it is not surprising that the herbaceous species richness of constructed wetlands shows a
 decreasing trend with age. Young constructed forested wetlands bear a closer
 resemblance to marshes than forests and have greater herbaceous richness during early
 marsh-like conditions with decreasing richness as they begin to resemble a forest.

 Evaluating the occurrence of vegetation from structural categories other than
 herbaceous can be valuable. The potential for future shifts in species dominance may be
 reflected in species establishment in the understory. Figure 7.25a shows the frequency of
 occurrence of canopy tree species in the understory of constructed forested wetlands.
 Although canopy tree seedlings occur throughout the age-range; the two, five and ten
 year old sites are located immediately adjacent to an unmined floodplain forest.
 Seedlings in the two-year-old and five-year-old sites result from recruitment from the
 adjacent forest since there are no mature trees in either site or they may have resulted
 from supplemental planting.

 The occurrence of subcanopy tree seedling other than Myrica sp. and Salix sp. is
 rare except in the two oldest sites (Figure 7.25). There appears to be a shift from
 recruitment of Salix sp. to Myrica sp. between 8 and. 12 years. Prior to 8 years, canopy
 cover is less than 50%, resulting in light transmittance conditions, which may be more
 conducive to Salix sp. Between 8 and 10 years, average canopy cover is slightly less than
 75%, and both Salix sp. and Myrica sp. are recruited. After 12 years, canopy cover has
 exceeded 80% and Myrica sp. appears to be favored. The only apparent exception to this
 trend with age is the 17-year-old site, but with only 57% canopy cover, this site still
 provides conditions favoring Salix sp.

 Early successional shrub species continue to be recruited in the understory of
 constructed wetlands of all ages. Only two late successional species occurred in the
 understory, Cephalanthus occidentalis and Itea virginica. C. occidentalis occurred in
 36% of the sites, while I. virginica occurred in only one.

 Figure 7.26a provides the frequency of occurrence of Parthenocissus
 quinquefolia, Smilax sp. Toxicodendron radicans, and Vitis rotundifolia in natural
 ecosystems in Florida (Davis and others 1991). Frequency of occurrence represents the
 fraction of the total number of transects sampled in which each species occurred. Figure
 57b shows the frequency of occurrence, the fraction of total number of quadrats in which
 each of these four species occurred, of vines in constructed forested wetlands. Natural
 and constructed forested wetlands show some similarities with respect to these four
 species, but Figure 7.25c shows the frequency of occurrence of six other vine species
 found in constructed forested wetlands. It is not clear if these species are found in natural
 systems to the extent to which they occur in constructed systems.


 Soil Development

 Soil properties play a deterministic role in the self-organization of ecosystems. It
has been suggested that soil properties influence the type of community that develops in a


7-88