A chronosequence (1 to 250+ years) of 160 woody species plots was established throughout northern Missouri riparian forests to explain the influence of site hydrology and stand age on species distribution and forest structure. Site hydrology was estimated by the surrogate variables of watershed size, height above the channel shelf, distance to channel, stream segment gradient, depth to water table, and percent clay. Stand age was found to be the dominant factor affecting species distribution and forest structure across the chronosequence. Watershed size was found to influence only species distribution; forest structure (tree height, coarse woody debris, size and age distributions) was not affected by watershed size. A conceptual model was developed and tested on the importance of site hydrology to species distribution and forest structure. Tree data were grouped into the flood tolerant and wetland indicator functional groups and these groups were used for model validation. The importance of very flood tolerant and obligate wetland species increased in larger watersheds suggesting that site hydrology shapes northern Missouri riparian forest species distribution. The flood-pulse hypothesis suggests that flood duration increases in larger watersheds and the conceptual model was based on this hypothesis; however, gage analysis suggested flow duration on the floodplain did not increase with watershed size. In addition, gage analysis indicates that flow duration is not biologically significant on these floodplains. Thus, it is hypothesized that other flooding variables or soil moisture content (groundwater) are important for the increase in the very flood tolerant or obligate wetland species in larger watersheds; additionally, the larger size of the geomorphic features in larger watersheds could increase the colonization of the very shade intolerant cottonwood and black willow. Stand age was found to be important in structuring forest communities, explaining species replacement patterns, and species richness. Early successional communities were found to differ in species composition between watershed sizes. Small watersheds (< 1000 km2) had silver maple and boxelder as the early successional community, while early successional communities in larger watersheds (> 1000 km2) were dominated by cottonwood and black willow. It is hypothesized that larger watersheds have the sediment transport capacity to create the bare mineral soil regeneration sites required for tree species in the family Salicaceae. Correlations of species importance values and stand age indicated that 12 of 18 species were influenced by stand age and five of the six remaining species could be considered midsuccessional, thus not expected to increase montonically with stand age. Watershed size was also associated with species distribution and tree species diversity patterns in these riparian forests. Correlation analysis indicated that the distributions of 12 of the 18 most abundant species were influenced by watershed size. Non-metric multidimensional or...

A stocking diagram for Midwestern bottomland eastern cottonwood (Populus deltoides Bartram ex Marsh.)-silver maple (Acer saccharinum L.)-American sycamore (Platanus occidentalis L.) forests was developed following the methods of S.F. Gingrich (1967 Measuring and evaluating stocking and stand density in upland hardwood forests in the Central States. For. Sci. 13:38‐53). The stocking diagram was derived from forest inventory data from two different studies of bottomland forests that covered a wide range of soil and hydrologic site characteristics found throughout the central Midwest, including Missouri, Iowa, Illinois, and southern Wisconsin. The minimum of full stocking (B-level) was determined from measurements on open-grown trees. The maximum of full stocking (A-level) in our study was almost one-third higher in stand basal area than A-level stocking, as determined by J.C.G. Goelz (1995. A stocking guide for southern bottomland hardwoods. South. J. Appl. For. 19:103‐104) for southern hardwood bottomlands dominated by cherrybark oak (Quercus pagoda Raf.), Nuttall oak (Quercus nuttallii Palmer), and sweetgum (Liquidambar styraciflua L.,), or by C.C. Myers and R.G. Buchman (1984. Manager's handbook for elm-ash-cottonwood in the north central states. GTR-98. US Forest Service, North Central For. Exp. Stn., St. Paul, MN. 11 p) for elm (Ulmus spp.), ash (Fraxinus spp.), and eastern cottonwood forests in the north central states. However, A-level stocking in this study was only slightly higher than guides developed for northern red oak (Quercus rubra L.) in Wisconsin (McGill, D.W., R. Rogers, A.J. Martin, and P.S. Johnson. 1999. Measuring stocking in northern red oak stands in Wisconsin. North. J. Appl. For. 16:144‐150). Differences in stocking among these forest types are due to variation in species composition, species silvical characteristics, and possibly the data sources used to construct the stocking diagrams. This stocking diagram can be used by forest managers to make decisions related to stand stocking for management of eastern cottonwood, silver maple, and American sycamore bottomland forests in the Midwest.

Hydrologic connectivity between the channel and floodplain is thought to be a dominant factor determining floodplain processes and characteristics of floodplain forests. We explored the role of hydrologic connectivity in explaining floodplain forest community composition along streams in northern Missouri, USA. Hydrologic analyses at 20 streamgages (207–5827 km2 area) document that magnitudes of 2‐year return floods increase systematically with increasing drainage area whereas the average annual number and durations of floodplain‐connecting events decrease. Flow durations above the active‐channel shelf vary little with increasing drainage area, indicating that the active‐channel shelf is in quasi‐equilibrium with prevailing conditions. The downstream decrease in connectivity is associated with downstream increase in channel incision. These relations at streamflow gaging stations are consistent with regional channel disturbance patterns: channel incision increases downstream, whereas upstream reaches have either not incised or adjusted to incision by forming new equilibrium floodplains. These results provide a framework to explain landscape‐scale variations in composition of floodplain forest communities in northern Missouri. Faust () had tentatively explained increases of flood‐dependent tree species, and decreases of species diversity, with a downstream increase in flood magnitude and duration. Because frequency and duration of floodplain‐connecting events do not increase downstream, we hypothesize instead that increases in relative abundance of flood‐dependent trees at larger drainage area result from increasing size of disturbance patches. Bank‐overtopping floods at larger drainage area create large, open, depositional landforms that promoted the regeneration of shade‐intolerant species. Higher tree species diversity in floodplains with small drainage areas is associated with non‐incised floodplains that are frequently connected to their channels and therefore subject to greater effective hydrologic variability compared with downstream floodplains. Understanding the landscape‐scale geomorphic and hydrologic controls on floodplain connectivity provides a basis for more effective management and restoration of floodplain forest communities. Published 2013. This article is a U.S. Government work and is in the public domain in the USA.