Holding the line: three decades of prescribed fires halt but do not reverse woody encroachment in grasslands
Encroachment of woody vegetation represents a significant global threat to biodiversity in grasslands, but practices used to reverse encroachment are rarely evaluated comprehensively. Several factors may drive encroachment, such as land use history, alteration of disturbance regimes, and local environment, but their relative importance is poorly understood. Another complicating factor is that encroachment may proceed via positive feedbacks that result in thresholds, beyond which its reversal is difficult.
We ask what impact reintroducing frequent fire has on encroachment relative to the influences of landscape context and historical vegetation. We investigate whether woody cover frequency distributions suggest that feedbacks reinforce encroachment after a threshold of woody cover is surpassed.
We analyze aerial photos in glade grasslands in Missouri, USA, to assess encroachment patterns over a 75-year period. Fire was excluded from this landscape for the first 45 years, and then reintroduced at varying frequencies in the last 30 years.
Woody vegetation cover increased sevenfold from 1939 to 2014 overall. After the reintroduction of prescribed fire, woody cover stayed approximately constant in burned glades, but continued increasing in unburned glades. Woody cover followed bimodal frequency distributions in burned areas. Fire-tolerant vegetation tended to encroach near historically wooded areas, while fire-sensitive vegetation responded more to fire history.
Altered disturbance regimes, in addition to numerous recognized drivers, can cause ecosystem state changes associated with losses to biodiversity. Conducting management early in the encroachment process and restoring grasslands at broad landscape scales may help counteract local feedbacks that promote encroachment.
KeywordsBistability Glades Hysteresis Juniperus virginiana Nonlinear responses Regime change Ozarks Resilience State changes Global change
We thank Mark Twain National Forest and the Missouri Spatial Data Information Service for providing data and other support. Anisa Fadhil provided assistance with GIS analyses, and we are grateful to Phil Hahn and Tony Ives for their wizardly statistical guidance. We thank Monica Turner and the Damschen lab for their assistance. This material is based upon work supported by National Science Foundation (NSF) DEB-0947432 and the NSF Graduate Research Fellowship under Grant No. 2012149884. ZR was supported by NSF DBI-1402033.
- Bartoń K (2015) MuMIn: multi-model inferenceGoogle Scholar
- Dey DC, Guyette RP, Stambaugh MC (2004) Fire history of a forest, savanna, and fen mosaic at White Ranch State Forest. In: Spetich MA (ed) Upland oak ecology symposium: history, current conditions, and sustainability. U.S. Department of Agriculture, Asheville, pp 132–137Google Scholar
- Dollison RM (2010) The National Map: new viewer, services, and data download: U.S. Geological Survey Fact Sheet 2010–3055Google Scholar
- ESRI (2011) ArcGIS Desktop: Release 10Google Scholar
- Gelman A, Hill J (2007) Data analysis using regression and multi-level/hierarchical models. Cambridge University Press, New YorkGoogle Scholar
- Guyette RP, Muzika RM, Dey DC (2002) Dynamics of an anthropogenic fire regime. Ecosystems 5:472–486Google Scholar
- Kimmel VL, Probasco GF (1980) Change in woody cover on limestone glades between 1938 and 1975. Trans Mo Acad Sci 14:69–74Google Scholar
- Kuznetsova A, Brockhoff PB, Christensen RHB (2015) lmerTest: tests in linear mixed effects modelsGoogle Scholar
- Ladd D (1991) Reexamination of the role of fire in Missouri oak woodlands. Proceedings of the Oak Woods Management Workshop. Eastern Illinois University, Charleston, pp 67–80Google Scholar
- Nelson P (2005) Terrestrial natural communities of Missouri. Missouri Department of Natural Resources, Jefferson CityGoogle Scholar
- Nelson P (2014) MO 2014 natural glades. Department of Natural Resources, MissouriGoogle Scholar
- R Core Team (2014) R: a language and environment for statistical computing. R Foundation for Statistical Computing, ViennaGoogle Scholar
- Rice JM (2009) Forest-meadow dynamics in the central western Oregon Cascades: topographic, biotic, and environmental change effects. Oregon State University, Oregon, p 245Google Scholar