Twenty‐five years of tree demography in a frequently burned oak woodland: implications for savanna restoration

Abstract

Due to decades of fire suppression, much of the Upper Midwest savanna habitat has converted to oak woodland. In efforts to restore oak savanna habitat, fire has been re-introduced in many of these woodlands. A primary purpose of these burns is to kill the fire-sensitive mesophytic tree species, which had established themselves during the decades of fire suppression, reduce the number of understory trees, and preserve the larger more widely spaced oaks. It is clear from ongoing efforts that restoring oak savannas will require frequent fires over decades, but frequent fires over the long term can also threaten the desirable oaks. Long-term demographic studies at savanna restoration sites experiencing frequent fires are necessary to determine the extent to the frequent burns are supporting and/or confounding restoration goals. Results presented here are from a twenty-five-year demographic study of an Upper Midwest bur oak (Quercus macrocarpa) savanna/woodland experiencing frequent fire, during which both the survival and growth of more than 9000 trees were documented. Survival was assessed annually and growth every five years. In the face of frequent fires, stem survival was found to be strongly associated with tree species, stem size, and stem growth. In turn, stem growth was found to be related to tree species and stem size. Decades of frequent burning in this oak woodland have substantially reduced the abundance of unwanted trees, specifically mesophytic species and Quercus ellipsoidalis, the latter which outcompetes Q. macrocarpa in the absence of fire. While Q. macrocarpa mid-sized (10-25 cm dbh) and large (≥25 cm dbh) trees are quite resistant to fire and now dominate the savanna landscape, they are not immune from fire-induced mortality. It is recommended that the number and density of these trees should be re-evaluated every few years to ensure that desirable numbers remain. If necessary, fires should be suspended for a period of time. This will give smaller Q. macrocarpa trees time to grow larger and become more fire-resistant, thereby ensuring successive generations of Q. macrocarpa.

Neighbours consistently influence tree growth and survival in a frequently burned open oak landscape

“…In the study reported here, due to persistent neighbour facilitation, trees survived best growing in groups. Since the frequent fires prevented any significant tree regeneration in the study area, either through sprouting or seedling recruits (Davis, 2021), the study showed that aggregated tree survival, by itself, can also be an important driver of post‐fire tree clumping. The results reported here, along with those by Lutz et al (2014) and Larson et al (2015), emphasize the importance of documenting the spatial patterns of surviving trees following fires, in addition to those from regeneration, to gain a fuller understanding of the processes creating the observed tree spatial patterns in fire‐frequent landscapes.…”

“…Although growth in large Q. macrocarpa trees was not affected by conspecifics, it was negatively affected by heterospecific neighbours. This is likely due to interspecific competition with the much faster growing Q. ellipsoidalis , which grows (increase in dbh) three times faster than Q. macrocarpa (Davis, 2021). Also, at this site Q. ellipsoidalis typically grows taller than Q. macrocarpa , meaning that even large Q. macrocarpa trees can end up growing in the shade of Q. ellipsoidalis .…”

“…In 1995 and 1996, nearly 9,000 trees with a dbh (diameter at breast height) of at least 2 cm were tagged, each with a unique number. Their dbh and x , y location within the grid were also recorded (Davis, 2021). If a tree possessed multiple stems, the tag was attached only to the largest stem.…”

“…However, dbh was measured and recorded for all stems of multiple‐stemmed trees. A burn programme was instituted in the grid area in 1987, and the study grid contained three different burn units, with burn numbers in the units during the 25‐year study (1995–2020) ranging from 9 to 11 (Davis, 2021) and mean burn frequencies ranging from one in 2.3 years to one in 2.75 years. Burns took place in late April or early May and were surface fires.…”

“…All trees, including each of the stems of multiple‐stemmed trees, were visited annually and their statuses (live, dead standing or dead fallen) recorded. The dbh of live stems was measured every 5 years (Davis, 2021).…”