Ecological Indicators 13 (2012) 275–283 Contents lists available at ScienceDirect Ecological Indicators jo ur nal homep age: www.elsevier.com/locate/ecolind Spatial patterns of aquatic habitat richness in the Upper Mississippi River floodplain, USA Nathan R. De Jager ∗ , Jason J. Rohweder U.S. Geological Survey, Upper Midwest Environmental Sciences Center, 2630 Fanta Reed Road, La Crosse, WI 54603, United States a r t i c l e i n f o Article history: Received 21 March 2011 Received in revised form 1 June 2011 Accepted 10 June 2011 Keywords: Aquatic habitat richness Floodplain river Fractal Habitat connectivity River restoration Secondary channel a b s t r a c t Interactions among hydrology and geomorphology create shifting mosaics of aquatic habitat patches in large river floodplains (e.g., main and side channels, floodplain lakes, and shallow backwater areas) and the connectivity among these habitat patches underpins high levels of biotic diversity and produc- tivity. However, the diversity and connectivity among the habitats of most floodplain rivers have been negatively impacted by hydrologic and structural modifications that support commercial navigation and control flooding. We therefore tested the hypothesis that the rate of increase in patch richness (# of types) with increasing scale reflects anthropogenic modifications to habitat diversity and connectivity in a large floodplain river, the Upper Mississippi River (UMR). To do this, we calculated the number of aquatic habitat patch types within neighborhoods surrounding each of the ≈19 million 5-m aquatic pixels of the UMR for multiple neighborhood sizes (1–100 ha). For all of the 87 river-reach focal areas we examined, changes in habitat richness (R) with increasing neighborhood length (L, # pixels) were characterized by a fractal-like power function R = L z (R 2 > 0.92 (P < 0.05)). The scaling exponent (z) measures the rate of increase in habitat richness with neighborhood size and is related to a fractal dimension. Variation in z reflected fundamental changes to spatial patterns of aquatic habitat richness in this river system. With only a few exceptions, z exceeded the river-wide average of 0.18 in focal areas where side channels, con- tiguous floodplain lakes, and contiguous shallow-water areas exceeded 5%, 5%, and 10% of the floodplain respectively. In contrast, z was always less than 0.18 for focal areas where impounded water exceeded 40% of floodplain area. Our results suggest that rehabilitation efforts that target areas with <5% of the floodplain in side channels, <5% in floodplain lakes, and/or <10% in shallow-water areas could improve habitat diversity across multiple scales in the UMR. Published by Elsevier Ltd. 1. Introduction Conceptual models that seek to explain patterns of diversity and productivity in large floodplain rivers are primarily derived from the concept of connectivity (Vannote et al., 1980; Minshall et al., 1982, 1985; Ward and Stanford, 1983; Junk et al., 1989). Connec- tivity can be expressed as upstream-downstream connections (e.g., the river continuum concept (Vannote et al., 1980; Minshall et al., 1982, 1985)), or as the absence thereof (e.g., the serial discontinu- ity concept (Ward and Stanford, 1983)), or as lateral connections between the river and floodplain (e.g., the flood pulse concept (Junk et al., 1989)). Current, broadened models of riverine biodiversity and produc- tivity (e.g., Thorp and Delong, 1994; Ward et al., 1999; Thorp et al., ∗ Corresponding author. Tel.: +1 608 781 6232; fax: +1 608 783 6066. E-mail address: ndejager@usgs.gov (N.R. De Jager). 2006) integrate various concepts from previous models but also incorporate concepts derived from hierarchy theory and landscape ecology. These concepts suggest that higher levels of biological organization (e.g., habitat/landscape elements) constrain lower levels (e.g., communities and populations) (Allen and Starr, 1982; O’Neill et al., 1986). At higher levels of biological organization, large river floodplains have been characterized as diverse mosaics of habitat patches with pronounced yet dynamic ecotones (Church, 2002; Amoros and Bornette, 2002; Tockner et al., 1998; Ward, 1998). The type, number, and connectivity among patch types in these systems is a function of large-scale and long-term feedbacks among hydrology, geomorphology, ecology, and human land and water use (Ward et al., 1999) and strongly influence biodiversity patterns (Ward et al., 1999; Koel, 2004; Thorp et al., 2006). The geometry and flow dynamics of most large floodplain rivers have been fundamentally altered by navigation and flood–control structures (National Research Council, 1992) with concomitant effects on aquatic habitat connectivity and biodiversity (Ward 1470-160X/$ – see front matter. Published by Elsevier Ltd. doi:10.1016/j.ecolind.2011.06.013