Regionalization at the CDR

Authors: Peter B. Reich

Scientific Basis for Regionalization at the CDR LTER Site

The Cedar Creek LTER site lies at the prairie-forest border in central North America and has been a focal point for research on successional processes in grassland ecosystems on old fields, as well as for research on causes and consequences of biodiversity. Along the so-called tension zone in the upper Midwest (also known as the regional prairie-forest border), the presettlement landscape was most likely a mosaic of grasslands, forests, wetlands and mixed vegetation types, including oak-dominated woodlands and savannas. To the northeast of the tension zone, vegetation regionally shifts from oak-dominated ecosystems to northern hardwood forests to boreal forests. To the west of the tension zone the vegetation trends into tallgrass prairie that stretches across much of the Great Plains region. Thus, the Cedar Creek LTER is located at or near the crossroads of several major temperate zone biomes. As a result of its location, researchers of the Cedar Creek LTER project are becoming increasingly aware of linkages between the Cedar Creek site itself and important basic and applied regional ecological questions.

Our regionalization research to date focuses on fire in oak-dominated ecosystems. This work addresses both basic and applied issues relevant to ecology and land management.

Regionalization Research

Cedar Creek researchers are addressing regionally important issues in the following two ways: with studies that are physically linked within the upper Midwest region and with studies that conceptually address questions that cross biomes or that examine variation within a biome.

Regional studies within the Cedar Creek LTER group

Native Americans manipulated grassland, savanna and forest ecosystems by their use of fire, and in the past century, Euro-Americans largely suppressed fire. It is well known that oak-dominated ecosystems may be dependent upon fire. It is likely that activities of Native Americans promoted fire-tolerant plant species and communities, and that 20th century fire suppression has pushed those ecosystems in the opposite direction. Despite our inability to ever characterize presettlement vegetation in a way that is much more than a snapshot in time, it has been well recognized that better understanding of the effects of fire, and of its possible uses in land management, is an important area in basic and applied ecology.

No natural fire regimes exist in the tension zone ecotone anywhere in Minnesota, Wisconsin, Illinois or Indiana. Numerous researchers and land management agencies have been experimenting and/or using fire over the past several decades, and that trend is increasing. However, careful scientific examination of the effects of fire has generally not kept up with its use in management activities: in some ways practitioners have been ahead of the researchers. To improve our scientific understanding of fire and oak-dominated ecosystems, LTER scientists and a group of researchers from around the region have teamed up in a regional oak ecosystem project. Researchers within this network are experimenting using prescribed burning to better understand fire effects on these ecosystems, and as a means of restoring and maintaining ecosystems ranging from mesic red oak forest to dry oak savannas.

The Cedar Creek site houses the longest running experiment that broadly covers the range of oak ecosystems. At Cedar Creek Natural History area there is a set of sites that represent a continuum from closed canopy forest (>90% tree cover) to open savanna (<10% tree cover). The communities on these sites have been developed in response to an ongoing 32-year fire experiment, in which given sites are either unburned or burned on approx. 1 in 10, 1 in 3, 1 in 2, 2 in 3 or 4 in 5 year cycles. Research at seven other sites in the region (Figure 1) complement the work at Cedar Creek. At all eight sites, prescribed burning experiments have been conducted in oak forest or savanna to test hypotheses about the effects of fire on vegetation dynamics, soil properties and processes, plant nutrition and ecophysiology, and ecosystem structure and function (see Table 1). Except for the work at Cedar Creek carried out in large part as part of the LTER project, research at all other sites has been funded in other ways. Thus the regionalization taking place at the Cedar Creek LTER represents a collaborative effort beyond the LTER network.

Collectively, the results from these eight sites provide empirical data that is in short supply compared to a wealth of non-replicated managed sites and a rich anecdotal data base. What has been found so far?

1) Acorn predation

Predation of acorns (once on the ground) is extremely high. In all cases, 60 to 95% of unprotected acorns are removed by animals in savannas or forests. Interannual variation in this is surprisingly modest after 4 years at one site.

2) Acorn germination, escape from predation and early survival are positively affected by fire in all cases

In mesic forest on medium textured soils, the number of acorn germinants and young seedlings is higher in infrequently burned than unburned forest. On sandy soils in a dry oak forest to savanna continuum, germination and early survival increase with fire frequency. The numbers of germinants and young seedlings reflect escapees from seed predation that managed to germinate and survive. Thus, no one single factor controls the level of recruitment. We have no data yet on which animals are responsible for the acorn removal, nor on the proportion of acorns that are merely moved to new sites as opposed to being eaten.

3) Oak seedling/sapling stage

Periodic fire favors oak seedlings/saplings over fire-intolerant tree species in xeric or mesic forest. In studies to date we have found that following fire there is no change in population for oaks, but a sharp decline in other woody species. In contrast, frequent fire favors fire-tolerant grasses and herbs over oak seedlings/saplings.

4) Nitrogen dynamics

Infrequent fire in forest has little affect on soil net N mineralization or in N pool sizes, but plants have a brief period of enhanced N availability due to reduced competition. Again in strong contrast, 32 years of frequent fire in sandy oak ecosystems at Cedar Creek strongly decreased soil N mineralization, due to long-term effects of fire itself, plus a change in vegetation type. Soil N mineralization is higher in forests than in woodlands and lowest in grasslands, even when sites are standardized for fire treatments.

5) Critical stage in community development?

Recruitment of stems into young pole-size tree class appears to be the critical process determining whether a community moves toward non-oak hardwood forest, oak forest, oak woodland, oak savanna or prairie. Fire, deer, deep shade or other stresses can almost entirely eliminate this step in some cases.

---In mesic forest, lack of fire promotes stiff competition, and more shade-tolerant tree species win out, eliminating oak from the next generation of canopy dominants.

---In mesic forest, periodic fire may (data are incomplete) enable oak to remain an important component in next generation of canopy dominants

---On sandy sites, lack of fire enables xerophytic oaks to regenerate (stands have moved from woodland to closed forest in 32 years at Cedar Creek). Once the canopy is closed, the stand will continue to move towards more shade-tolerant tree species, and the balance of oak vs. non-oak tree species will depend largely on just how xeric a given site is.

---On xeric or mesic sites, frequent regular fire will likely suppress meaningful recruitment into the pole-size stage. Thus a community may pass through several stages: from a forest to woodland to savanna to prairie over a long time period (stands have moved from dense woodland to open savanna in 32 years at Cedar Creek).

Continued studies will focus both on the mechanisms whereby fire influences resources and vegetation dynamics and on the implications of fire regimes for long-term stand structure and function.

In addition to studies within the oak-dominated ecotone, future research will likely expand to other biomes. Thus, increasing regionalization will also likely take the form of studies to increase our understanding of the controls on ecosystem patterns and processes as one moves perpendicular to the tension zone and as vegetation rapidly shifts from grassland to boreal forest in a space of only several hundred km.

Cross-biome comparative studies at the Cedar Creek site

A number of studies are ongoing at Cedar Creek that address regional issues, such as comparisons among grasslands or forests, or contrasts between grasslands, savannas and forests.

Are Cedar Creek forests, growing on very course sands, different in important ways from forests elsewhere in the region? Does net primary production, litterfall N cycling or soil N mineralization differ among regional forests? Cedar Creek LTER scientists are addressing this issue by comparative studies from sites on rich and poor soils in Minnesota and Wisconsin.

Another project examines the relationships between leaf area index, net primary production and land cover classes at a range of spatial scales. Understanding the regional implications of findings from any one site, such as an LTER site, partially depends on whether patterns and processes translate in any understandable way at different sites, and or once aggregated at landscape scales. Researchers at Cedar Creek are involved as part of an LTER network project, called MODLERS, addressing issues of scale in these fundamental ecosystem attributes. The project uses both ground data and remotely sensed data to ask whether a 10 x 10 km area centered on Cedar Creek looks, feels and acts the same in terms of ecosystem functioning as the aggregated responses of the 160,000 25 x 25 m cells that make up the 100 km2 landscape.


Harrington, R.A., B.J. Brown & P.B. Reich. 1989. Ecophysiology of exotic and native shrubs in southern Wisconsin. I. Relationship of leaf characteristics, resource availability and phenology to seasonal patterns of carbon gain. Oecologia 80: 356-367.

Reich, P.B., M.D. Abrams, D.S. Ellsworth, E.L. Kruger & T.J. Tabone. 1990. Fire affects ecophysiology and community dynamics of central Wisconsin oak forest regeneration. Ecology 71: 2179-2190.

Faber-Langendoen, D and J.R. Tester. 1993. Oak mortality in sand savannas following drought in east-central Minnesota. Bulletin of the Torrey Botanical Club 120:248-256.

Ko, L.J. & P.B. Reich. 1993. Oak tree effects on soil and herbaceous vegetation in savannas and pastures in Wisconsin. American Midland Naturalist 130:31-42.

Cole, KL and Taylor, RS. 1995. Past and current trends of change in a dune prairie/oak savanna reconstructed through a multiple-scale history. Journal of Vegetation Science 6:399-410.

Reich, P.B., D.F. Grigal, J.D. Aber & S.T. Gower. 1996. Nitrogen mineralization and productivity in 50 temperate forest stands of differing community and soil type. Ecology (in press).

Kruger EL and P.B. Reich. Responses of Quercus rubra and competing tree regeneration to fire in mesic forest openings. I. Community dynamics (submitted to Can J Forest Research)


USDA Competitive Grants Program, National Park Service, NASA, Minnesota Legislative Commission on Minnesota's Resources

Go to regionalization document
Links to LTER Sites

Submitted by: