Dendrochronologists often examine annual variation in site-averaged tree-ring widths, but spatial growth variation within a site is less well-studied. Strong correlation of river flow with the average annual ring width of floodplain trees in dry regions suggests that temporal variation in growth is limited by water availability. We tested the hypothesis that spatial variation in growth is similarly controlled by the topographic factors associated with water availability: height above water surface and distance from channel. We analyzed variation in growth of 316 plains cottonwoods (Populus deltoides subsp. monilifera) aged 30 to 369 years at randomly selected locations on the floodplain of the Little Missouri River in the North and South units of Theodore Roosevelt National Park, North Dakota, USA. We averaged ring widths of two cores from every tree and corrected for tree age by dividing each annual ring width by the average value for trees of the same age at the site (regional curve standardization). We related age-corrected ring widths to height above water surface and distance from channel determined from Lidar, annual streamflow (log of mean annual April-July flow) and stem density determined from field measurements and above-ground Lidar returns.
Results/Conclusions
Site-averaged ring width was strongly positively correlated with annual flow, indicating that variation in growth over time is limited by water scarcity. Variation in growth over space, however, was not strongly related to the topographic factors that would be expected to control water availability. Height above water surface (ranging from 2 to 7 m) and distance from channel were at most weakly correlated with ring width and with the correlation between ring width and flow. The strongest influence on spatial variation in ring width was competition; stem density was negatively correlated with ring width and the flow-ring width correlation, especially in stands less than 70 years old. Growth rate and correlation between growth and flow varied among point bars, and these differences changed from one 30-year period to the next. We conclude that growth of cottonwood is controlled more by year-to-year variation in water availability and intraspecific competition than by topographic position on the floodplain.