Burning is often used to manage North American prairies. However, few studies have assessed the importance of time (season) of burning on vegetation responses. In this study, long-term effects (1926-1982) of annual burning at different seasons in ungrazed tallgrass prairie in the Flint Hills of Kansas (central USA) were investigated.
The site from the early 1900’s, was cattle-grazed at moderate intensity and burned in late winter. In 1926, prevalent species were little bluestem Andropogon scoparius (32.6%), prairie junegrass Koeleria pyramidata (20.2%), big bluestem A.gerardii (16.7%)and indiangrass Sorghastrum nutans (11.0%), with perennial herbs (6.1%). Annual grasses and herbs were rare (<1% cover).
In 1926, a 0.4 ha area was fenced and divided into 10, 20 x 10 m plots, 0.9 m apart. Treatments (beginning December 1926) were: winter burn (1 December), early-spring burn (20 March), mid-spring burn (10 April), late-spring burn (1 May), and unburned (controls). Four plots were burned annually and four burned every other year, but from 1950 these latter plots were burned yearly. Burning was not undertaken in 1935-1937 and 1945-1949.
Unburned plots were mown and cuttings removed every year before initiation of spring growth until 1968 (unmown thereafter). All plots were burned by wildfires in January 1951 and March 1966.
Herbage production in each plot was estimated annually at the end of the growing season (usually early-October). Botanical composition was estimated (usually by the end of June) from 1928-1982.
Time of burning greatly influenced vegetation composition; only 3 weeks difference in time of spring burning had a dramatic long-term influence.
Annual late-spring burning (coinciding with emergence of the warm-season perennial grasses), maximized grass production and big bluestem and indiangrass prevalence, but significantly reduced perennial forbs (c.1.8% vs. 3.8-5.8% in other treatments). It also reduced sedges (3%) compared to other burn periods (9-12%) as it coincided with initiation of (dominant) Carex growth. Overall, it was not detrimental to species composition or total basal cover.
Burning in winter, early- or mid-spring reduced herbage production and shifted vegetation composition by favouring other species and increased species diversity. Little bluestemincreased with mid- and early-spring burning, while perennial forbs and sedges increased under early-spring and winter burning regimes.
In unburned plots Kentucky blue-grass Poa pratensis increased (13.8% vs. around 0.9-1.1% in burned plots), whilst invading trees reduced grass production and species diversity declined. Annual grasses never exceeded 1% cover (although average percentage highest in unburned and lowest in late-spring burned plots). Annual forbs were similarly insignificant, with no significant difference among treatments.