Projected changes in land cover, age class, and species composition of forests on non-channelized segments of the Missouri River

Background/Question/Methods

The Missouri River typically floods each spring and summer.  Fluvial geomorphological processes associated with stronger floods cause the river to meander across its floodplain over time. In this dynamic landscape, Plains Cottonwoods (Populus deltoides) colonize recently formed sandbars and, unless eroded away during subsequent floods, grow into the tall cottonwood-dominated. As individual trees die they are replaced by non-cottonwood species, resulting in a floodplain containing a diverse array of forest stands varying in age, structure, and species composition. The construction of six dams on the main stem of the river in the mid twentieth century drastically altered the natural flood regime of the river, reducing peak flows and lengthening flood durations. Fluvial processes now create fewer sandbars suitable for cottonwood recruitment, which has resulted in a gradual increase in the average age of forest stands. We digitized past and present floodplain land cover for three remnant floodplain segments on the Missouri River and used field reconnaissance to determine age class and forest type (cottonwood vs. non-cottonwood) of forest stands. We then developed a landscape dynamics model to project future trends in forest area, age class distribution, and forest type.

Results/Conclusions

We project that, following current trends, the total area of forest cover will decline slightly over the next one hundred years. These declines vary by river segment: the Garrison segment (below Garrison dam in North Dakota) will decline from 34% to 29% forest cover; the Fort Randall segment (below Fort Randall dam in South Dakota and Nebraska) will decline from 35% to 32% forest cover; and the Gavins Point segment (below Gavins Point dam in South Dakota and Nebraska) will decline from 7.5% to 6.8% cover. More importantly, we project that the age class distribution and species composition of forests will change markedly over the same time period. On the Garrison segment, cottonwood forest stands in our middle age-classes (25 to 115 years old) will decline from 87% of existing forest to 38% while non-cottonwood forests of all ages will increase from 7% of existing forest to 40%. The results are similar for both river segments in South Dakota and Nebraska. Our results suggest that under current trends of land conversion and flood control, native cottonwood forests will decline markedly over the next 100 years, to the point that much of the floodplain forest along the Missouri River will be dominated by non-cottonwood species.