Declines in abundance and species richness of birds following a major flood on the Upper Mississippi River

Auk, The, Jul 1997 by G, Melinda, E, Erwin

MELINDA G. KNUTSON1 AND ERWIN E. KLAAS

ABSTRACT.-We examined the abundance and species richness of birds breeding in floodplain forests of the Upper Mississippi River from 1992 to 1994 to identify effects of a major flood in 1993 on the bird assemblage. Sixty-five study plots were divided into treatments and controls based on whether they were flooded in 1993. Among 84 species observed on all plots, 41 species decreased in abundance from 1992 to 1994, 13 species increased, and numbers of 5 species were unchanged. Sample sizes were inadequate to evaluate trends for 25 species. Species richness declined over the three-year period. The main effect "year" was significant in 20 of the 36 species tested. Cool, wet conditions may have contributed to poor reproductive success in 1993 and resulted in widespread declines in abundance during the year following the flood. Bird abundance increased on most unflooded plots in 1993, probably because birds were displaced from flooded plots. This pattern was most striking for Neotropical migrants, species preferring habitat edges, lower-canopy nesters, and species that forage in the air. We suggest that periodic major flooding maintains suitable floodplain habitat for Prothonotary Warblers (Protonotaria citrea) in the face of competition with House Wrens (Troglodytes aedon) for nest sites. Received 30 November 1995, accepted 22 January 1996.

FLOODPLAINS ARE DYNAMIC HABITATS where periodic flooding creates natural disturbances that set back succession and provide new sites for plant establishment (Peck and Smart 1986, Kupfer and Malanson 1993, Jones et al. 1994, Sparks 1995, Yin and Nelson 1995). Large floodplain forests like those of the Upper Mississippi River are uncommon in northern landscapes (Mitsch and Gosselink 1993) and provide unique habitats that support a high abundance and diversity of birds (Grettenberger 1991, Best et al. 1995, Knutson et al. 1996). Few studies have examined the avifauna of large floodplain forests (Emlen et al. 1986, Decamps et al. 1987, Knutson 1995), and studies of the effects of floods that include pre-flood data are even less common (Knopf and Sedgwick 1987).

Direct effects of flooding on birds nesting in floodplains tend to be minor because spring floods usually recede by mid- to late May. Summer floods are rare and usually of short duration. However, major floods of long duration or high amplitude can create large-scale disturbances that dramatically change vegetative cover (Yin et al. 1994). Bird assemblages may respond to changes in availability of nest sites and food resources caused by these disturbances. Flooding had a major effect on a riparian bird community in Arizona by eliminating cottonwood and willow nesting habitat (Hunter et al. 1987). Depending upon the severity of the disturbance, some species may increase in abundance while others decline. In addition, flooding may maintain suitable habitat for some bird species that are uncommon in drier, less-disturbed habitats.

Rainfall during May to August 1993 in the Upper Mississippi River basin reached historically high levels (ca. twice normal), approximating 75- to 300-year recurrence intervals (Wahl et al. 1993, Interagency Floodplain Management Review Committee 1994). River water levels were well above normal throughout the breeding season (Parrett et al. 1993). We examined the effects of this major flood on the abundance and species richness of birds breeding in floodplain forests by comparing data collected from 1992 to 1994. We considered our data in relation to two hypotheses of how the flood and associated weather patterns affected the floodplain bird assemblage. The Flood Disturbance Hypothesis proposes that the pattern of change in abundance will differ between flooded and unflooded plots because structural changes in habitat will occur on flooded plots but not on unflooded plots. Bird abundance may increase on unflooded plots the year of the flood as birds seek refuge from adjacent flooded sites. Other possible effects include lower abundance on flooded plots in the year of the flood or the following year. We predicted that edge species, ground and lower-canopy nesters, and ground and lower-canopy feeders would be most affected by these structural changes. We also predicted that species preferring habitats near water, and piscivores, would increase on flooded plots. The Weather Hypothesis proposes that multispecies declines in abundance would be observed on both flooded and unflooded plots, resulting from reduced productivity in wet, cool weather associated with a flood (and not from the flood disturbance itself). We predicted that insectivores would be most affected if weather conditions depressed insect abundance (Rotenberry et al. 1995).

STUDY AREA AND METHODS

The study area included forested floodplain habitat of the Upper Mississippi River (Fig. 1) from near Winona, Minnesota downriver to Guttenburg, Iowa, a distance of about 177 km along the main channel of the river. This section of the river is unrestricted by levees; during high water the river floods some or all of the floodplain forests, depending upon their elevation. The forest community is dominated by silver maple (Acer saccharinum). Elm (Ulmus spp.), green ash (Fraxinus pennsylvanica), swamp white oak (Quercus bicolor), cottonwood (Populus deltoides), hackberry (Celtis occidentalis), and river birch (Betula nigra) are less abundant (Knutson 1995). The shrub layer is dominated by green ash seedlings, poison ivy (Toxicodendron radicans), silver maple seedlings, and prickly ash (Zanthoxylum americanum). Wood nettle (Urtica dioica) and reed canary grass (Phalaris arundinaceae) dominate the herb layer.