2011 SOUTHEASTERN NATURALIST 10(3):459–476 Short-term Assessment of Morphological Change on Five Lower Mississippi River Islands James E. Moore1,2,3,*, Scott B. Franklin2,4, Daniel Larsen5, and Jack W. Grubaugh2,6 Abstract - To examine short-term morphological changes and variables controlling this change, we conducted a study on five lower Mississippi River islands in 2007 and 2008. The southernmost island was located 25 km north of Memphis, TN, and the northernmost island was located 105 km north of Memphis. We surveyed elevations along six transects for each of the five islands to examine morphological changes between years and measured river velocity and backscatter intensity (sediment load) at the head, middle, toe, zee, and fore sides of the islands. We predicted the impacts of flooding on island geomorphology would be greatest at the head (upstream edge) and fore (thalweg) side. In addition, we predicted the morphology of islands present in the main channel in the 1890s would differ in morphology from contemporary islands. Results indicated morphological change during the study period but only in width. Islands showed general lateral migration, with one side aggrading while the other side eroded. Changes in morphological characteristics were consistent with intensity of river flow around the islands, with the highest velocity and sediment load observed near the island head and to a lesser degree the thalweg side of islands. In reference to islands present in the late 1890s, contemporary islands appear to be more elongated and have a greater surface slope. Since these islands provide beneficial habitat for flora and fauna, it is crucial to understand the dynamic nature of these landscape features. Introduction The Mississippi River is one of the world’s major river systems in size, habitat diversity, and biological productivity, and is also one of the largest regulated rivers. The lower Mississippi River (LMR) has undergone extensive modifications for navigation and flood management (Hudson et al. 2008, Wasklewicz et al. 2004a). Research on this section of the Mississippi River has focused on channel processes and form; however, little work has focused on riverine islands. Islands are easily identifiable, with water surrounding them for a majority of the year (Lekarczyk and Wasklewicz 2003, Osterkamp 1998). For our study, islands were considered to be in-channel land features surrounded by water at least nine months out of the year, with permanence not being a concern. Because of their 1Department of Biological Sciences, University of Memphis, 3700 Walker Avenue, Memphis, TN 38152. 2Edward J. Meeman Field Station, University of Memphis, Millington, TN 38058. 3Current address - Department of Biology, Christian Brothers University, Memphis, TN, 38104. 4School of Biological Sciences, The University of Northern Colorado, Greeley, CO 80639. 5Department of Earth Sciences, The University of Memphis, 201 Johnson Hall, Memphis, TN 38152. 6Department of Biological Sciences, The University of Tennessee at Martin, Martin, TN 38238. *Corresponding author - jmoore25@ cbu.edu. 460 Southeastern Naturalist Vol. 10, No. 3 placement within the main channel, these islands can be impacted by the smallest change in hydrology (Osterkamp 1998). Changes in the islands reflect variations in erosional and depositional processes, which are ultimately affected by natural processes and variables (flow velocity, water depth, sediment size and content) as well as river control structures and vegetative cover, which is tightly linked with flood intensity and duration of inundation. As such, a closer examination of changes in morphometry (i.e., elevation and bank slopes) of riverine islands can provide insight into river dynamics (Lane et al. 1998). Furthermore, evaluation of island morphometry over relatively short durations (months to years) may illustrate how these crucial habitats, and their associated floral and faunal communities, may be altered. Vegetative communities inhabiting islands along the LMR support many migratory waterfowl for food and cover, and provide critical habitat for the endangered Sterna antillarum (Lesson) (Least Tern; Smith and Renken 1991). The vegetative communities on these islands, similar to their floodplain counterparts, are intimately linked with hydrologic patterns. Periodic floods of varying frequency, duration, and erosive capability affect vegetation by: creating new land areas, changing bank stability, and ultimately forcing a gradient of flood intensities as elevation on the island increases (Hupp and Osterkamp 1996). Personal observations (J.E. Moore; over four years) on the islands indicate that where dense monotypic stands of Salix nigra Marsh (Black Willow), Salix interior Rowlee (Sandbar Willow), and Populus deltoides Bartram ex Marsh (Cottonwood) are found, islands seem to persist and are less affected by low-to moderate-magnitude events. Historical surveys of depositional islands in the LMR show that once vegetated, they can become part of the mainland within 20 years, retarding the movement of water during floods and causing rapid sediment deposition along their landward margin (Shull 1944). In the LMR, there are two major types of islands. The first is created via chute cut-off (Osterkamp 1998), either formed by deflected flow around woody flood debris (Maser and Sedell 1994) or, more recently, wing dams; the second is created from an increase in bed roughness that leads to sediment building through time to form a mid-channel bar (island) (Osterkamp 1998). In both cases, aggradational and degradational processes are event driven, with more severe floods resulting in greater aggradation or degradation (Shull 1922, 1944). Thus, islands remain dynamic, developing and disappearing over time if sparsely vegetated (Collins and Knox 2003, Gurnell and Petts 2002, Lekarczyk and Wasklewicz 2003). Some sections of the Mississippi River have shown a net loss in the number of islands over historic time (Theiling 1998), while other sections have increased in island number (Collins and Knox 2003). The main factors determining island formation or loss are flow intensity, sediment movement, duration of flow (Collins and Knox 2003, Nanson and Croke 1992), and vegetation characteristics; all of these have changed along the Mississippi River over time (Nielson et al. 1984, Shull 1944). Osterkamp (1998) suggests that the morphology of the island may be indicative of processes by which they were formed (e.g., a teardrop shape is typical of snag islands). Island degradation should occur where flow intensity and 2011 J.E. Moore, S.B. Franklin, D. Larsen, and J.W. Grubaugh 461 scouring effects are highest, i.e., head (upstream) portions and sides facing the main channel (fore), and aggradation should occur at the toe (downstream) and backwater sides (zee), resulting in downstream movement of islands over time (Lekarczyk and Wasklewicz 2003, Osterkamp 1998); however, anthropogenic modifications to the LMR may affect such relationships. The objective of this project was to examine five islands over a two-year period to determine how Mississippi River island morphology changes over short time periods and to determine which measured factors correlate with island change. We hypothesized that island morphology (assessed by measuring multiple topographic transects) would change significantly within a two-year period following a dry year (2007) and a flood year (2008). More specifically, we hypothesized that fore (thalweg) sides would be more affected than zee (backwater; away from main channel) sides due to differential scouring effects and that island heads would be more affected than island toes. Finally, we also hypothesized that the morphology of islands present in the main channel in the 1890s would differ from contemporary islands. Study Sites The study area covers a portion of the LMR from river mile (rm) 754 to rm 804 and comprises ≈8.0% of the LMR channel length (Table 1, Fig. 1). This reach is characterized by numerous wing dams and bank revetments (Hudson et al. 2008, Wasklewicz et al. 2004a). Five islands were chosen to assess morphological changes for a two-year period (2007–2008). The criteria for choosing islands included: 1) a separation from shore for at least nine months of a year, 2) a location not along a major bend in the river, and 3) logistic ability to access the island by boat. All islands were at least forty years old according to aerial photography and Table 1. Island location, approximate area, % bare sand, % woody cover, maximum elevation for 2007 and 2008, and length/width ratio at high water. The final row gives the range of values (with means in parentheses) from 10 randomly selected islands found in the 1890 Mississippi River Commission navigational charts within the study area (Mississippi River Commission 1890). Length/ Transect 1 Approximate % bare % woody Max width ratio Island lat/long sample area (m2) sand cover elev. (m) total island Densford Bar 35°23'16.3"N, 138,000 >90.0 1.5 10 6.28 90°04'14.6"W Dean Island 35°25'54.9"N, 550,000 5.0 60.0 11 4.80 90°00'28.0"W Sunrise Towhead 35°36'55.4"N, 25,000 30.0 44.0 8 5.29 89°54'11.5"W Keyes Point 35°43'39.6"N, 84,000 34.6 29.9 8 5.80 89°54'22.5"W Wardlow Bar 35°51'00.3"N, 83,400 46.7 24.0 11 7.08 89°43'34.1"W 1890 Islands (n = 10) 5250–1,950,000 3.0–95.0 5.0–75.0 3.1–6.1 2.75–7.5 (429,075) (4.6) (3.9) 462 Southeastern Naturalist Vol. 10, No. 3 river mapping data. The southernmost island (Densford Bar) was located at rm 754, and the northernmost island (Wardlow Bar) was located at rm 804. Dean Island (rm 758), Sunrise Towhead (rm 778), and Keyes Point (rm 790), along with Figure 1. Lower Mississippi River map showing research islands and gages in relation to river mile with latitude/longitude coordinates for each island. 2011 J.E. Moore, S.B. Franklin, D. Larsen, and J.W. Grubaugh 463 Densford Bar and Wardlow Bar, were disconnected from the mainland during low-moderate flow and were only connected to the mainland during extremely low flow (<-5 m stage at Osceola, AR; Fig. 1). Each island differed in % bare sand, % woody cover, and maximum elevation (Table 1). Methods Stage and discharge The Mississippi River possesses an extensive network of gages that have been recording river stage since the late 1800s (Wasklewicz et al. 2004b). For this study, we used only the Osceola gage (gage 152), located near Osceola, AR at rm 783.5 with an elevation of 63.8 m (209.43 ft), for river-stage data and estimating flood recurrence intervals. Gage data are uploaded daily from this gage, and each daily recording is an average of three readings taken for that particular day (8-hour intervals). River discharge is also reported based on readings from two gages north and south of the study islands. The Hickman, KY gage is located at rm 922 and has an elevation of 80.7 m (264.7 ft), whereas the Memphis, TN gage is located at rm 734.4 and has an elevation of 56.1 m (183.9 ft). Hydrology assessment Elevations of five river gages (including the Osceola river gage) were used to “zero” island elevation with river stage and provide an accurate estimate of hydrological impacts on each island. Elevations of each gage were used in a regression equation to obtain the slope of the river, a simple technique that allowed us to determine a zero-point elevation for each island so hydrologic data would match island elevation. To determine potential flow impacts (velocity and sediment load) near each island, we used a Teledyne Workhorse Rio Grande acoustic doppler current profiler (ADCP), which is an accurate rapid-sampling currentprofi ling system designed to operate from a moving boat. The acoustic doppler was employed to determine river flow velocity and backscatter intensity (used to gain an understanding of how much sediment was being carried at different locations around the islands) during high (>7.6 m) river stage events in 2008. Three flow transects were established at each island: one at the head, one in the middle (thalweg side), and one at the toe (Fig. 2a). These measurements were recorded near the water-sediment interface at distances of 1 to 5 m away from island sides (when the boat could not float over the top, as in middle sections) or island center to understand how flow velocity and sediment load differed on fore and zee sides and on head, middle, and toe sections of islands. Each measurement is a result of multiple acoustic soundings (pings) sent through the water column by the doppler. The doppler receiver resolves the reflected signals into multiple depth cells, each 5 cm in increment (high resolution), that are used to calculate an average velocity for that particular depth cell. Our recorded measurements were from the bottom depth cell (i.e., channel bed, bottom 5 cm of water column) at 1 to 5 m distance from the island (i.e., five measurements per transect). The velocity data from the bottom depth cells are used as a proxy for bedding-plane shear stress at the sediment-water interface to 464 Southeastern Naturalist Vol. 10, No. 3 Figure 2. Aerial photos of Densford Bar during (a) a high stage event (2010) and (b) a low-stage event (2004). Morophological terms applied to islands are show in (a). The sampling scheme used to collect flow velocity and sediment backscatter data using the ADCP doppler is also shown. Approximate island area at low flow is also illustrated. 2011 J.E. Moore, S.B. Franklin, D. Larsen, and J.W. Grubaugh 465 assess the potential for sediment movement at a given transect. The backscatter signal allows for the interpretation of sediment transport and provides an estimate of suspended sediment concentration; however, there is a poor understanding of the sensitivity of the doppler to backscatter particle size (Kostaschuk et al. 2004). Islands were sampled consistently during high-stage events, but only four (Densford Bar, Dean Island, Sunrise Towhead, and Keyes Point) of the five islands were sampled due to logistics and time constraints. Morphological change Six elevation transects were completed on each island extending from the water’s edge on the fore side (main channel side) to that on the zee side (backwater; away from main channel) and crossing the island perpendicular to overall flow. A permanent stake was positioned at the highest elevation point of the island encountered in each transect (typically near the center of the transect). Elevations were determined using a David White® autolaser 300 and LD-12 detector (David White, Germantown, WI) attached to a surveyor’s pole. Measurements were taken at topographical deviations along transects from center stakes to water’s edge on both sides of the islands (Fig. 2b) and transformed to 1-m resolution (used in a companion study that examined plant community composition in response to variable frequency and intensity of floods) (Moore et al., in press). Transect length was used as a proxy to assess morphological change of each island. Changes in transect length, after correction for deviation from zero water level at each island, indicate changes in width and length of islands. Relative changes in transect length are used to standardize comparison of island morphology. Bare sand and percent woody vegetation cover along with length/width ratio were quantified using Google Earth images (10 January 2007) (Moore 2011). In an effort to compare contemporary island geomorphology with 1890 island morphology, we randomly selected ten of the 13 islands mapped in 1890 that were similar in area to our islands and within the same river section (none of the islands from this study were present in 1890) from the navigation charts developed in 1890 by the Mississippi River Commission (Mississippi River Commission 1890). These topographical maps exhibit 1.5-m (5-ft) contour intervals and are easily comparable to our data, i.e., 1-m intervals. The scales of the aerial photos, however, differed by a scale factor of four; 1:1524 m (1:5000 ft) for contemporary islands and 1:6096 m (1:20,000 ft) for historical islands. Data analysis Maximum transect length on fore and zee sides between years were compared. Data were analyzed using paired t-tests and general linear models (GLMs) using SAS v.9.1.3 (SAS 2000). A Cochran-Armitage trend test was used to determine if island elevation increased or decreased for the two-year study period (i.e., positive trend statistic = elevation increase, negative trend statistic = elevation decrease).To assess differences in flow velocity and sediment load on different sides and areas (head, middle, and toe) of the islands, we used multiple analysis of variance (MANOVA). We also assessed angle of grade (slope) to confirm 466 Southeastern Naturalist Vol. 10, No. 3 differences on fore and zee sides and between years. We used a simple rise/ run technique in which we divided the elevation by transect length. Slope data were arcsine transformed and analyzed using a GLM ANOVA. To compare 1890 and contemporary islands, we compared basic metrics on island elevation, area, length/width ratio, and slope. Results Hydrology assessment River stage in 2007 was on average lower than the twenty-year mean (Fig. 3a). In 2007, the Mississippi River had 93 days of flow less than 0 m on the Osceola gage (i.e., stage was negative) and only 10 days of flow greater than 7.62 m (25 ft). Stage patterns in 2008 were on average higher than the twenty-year mean and showed 59 days of flow with a negative stage and 81 days of flow greater than 7.62 m (Fig. 3a). Discharge data were also higher in 2008 compared to 2007 (Fig. 4a, b), with 2008 showing the sixth highest discharge in twenty years (Fig. 3b). 2008 peak flow had a modest recurrence time of 5.25 years, while 2007 had a recurrence time of 1.31 yrs (Fig. 4c). Doppler flow velocity showed no statistical differences in regards to individual island, area of the island, or side; however, head portions and fore sides appeared to have slightly higher velocities (Fig. 5a, b). Also, suspended sediment concentration showed no statistically signifi cant differences in area of the island, or side; however, head portions and fore sides again appeared to have slightly higher sediment movement (Fig. 5c, d). Morphological change Historical islands averaged gentler slopes (mean slope = 0.07) with a maximum slope of 0.23 (Table 1). Contemporary islands had greater elevation gradients and, to a lesser extent, length/width ratios than their historic counterparts. Thus, contemporary islands appear to have greater relief and are longer than historical islands (Table 1). Transect length, however, showed significant interactions for transect x side (P = 0.01). Zee sides had significantly longer transects than fore sides (mean zee = 120 m; mean fore = 94 m) (Fig. 6). The greatest relative changes in transect length occurred toward the head of the islands on fore sides and on both the head and toe of islands on zee sides, but zee sides were not noticeably more affected than fore sides (Fig. 7). Densford Bar and Sunrise Towhead showed the greatest mean difference in transect length from 2007 to 2008 (Fig. 8). For all islands, a gain on one side meant a loss on the other, but such dynamics were island- and transect-dependent; e.g., fore sides showed greater positive differences compared to zee sides, with Densford Bar, Dean Island, and Keyes Point increasing in overall transect length (Fig. 6). Figure 3 (opposite page). (a) Daily Stage data for 2007, 2008, and the twenty-year average (m). Asterisks indicate dates doppler data were collected (15 Feb 2008, 28 April 2008). (b) Annual peak discharge for Hickman, KY and Memphis, TN gages for twentyyear period (1987–2008). Boxes represent peak discharge events for 2007 and 2008 respectively for the two gages. 2011 J.E. Moore, S.B. Franklin, D. Larsen, and J.W. Grubaugh 467 468 Southeastern Naturalist Vol. 10, No. 3 2011 J.E. Moore, S.B. Franklin, D. Larsen, and J.W. Grubaugh 469 No significant correlation between transect length and slope was observed (not shown). On average, fore sides had greater slopes than zee sides (Fig. 9) (mean fore slope = 0.34, mean zee slope = 0.26). Figure 4 (opposite page, top). Discharge (m3/s) vs. River Stage (m) for Hickman, KY, and Memphis, TN, gages in (a) 2007 and (b) 2008; labels indicate days near sampling period for which discharge data were collected by US Army Corp of Engineers. (c) Recurrence time for peak stages; + indicates 2007 peak discharge with a recurrence interval of 1.31 yrs., asterisk indicates 2008 peak discharge with a recurrence interval of 5.25 yrs. Y-axis stage values are log scaled. Figure 5 (opposite page, bottom). Mean (a) velocity (m/s) for head, middle, and toe areas of islands; (b) velocity (m/s) for fore and zee sides; (c) backscatter (dB) for north, middle, and toe areas of islands; and (d) backscatter for fore and zee sides. Error bars represent (± 1 standard error). Figure 6 (above). Transect (from island head [i.e., transect 6] to toe [i.e., transect 1]) length by island for five islands in the Mississippi River. Wide bars represent 2007 lengths, whereas thin bars represent 2008 lengths. Positive values represent fore side and negative values represent zee side. Thin bars longer than wide bars = aggrading transects. 470 Southeastern Naturalist Vol. 10, No. 3 2011 J.E. Moore, S.B. Franklin, D. Larsen, and J.W. Grubaugh 471 Figure 7 (opposite page, top). Mean relative change in transect (from island head [6] to toe [1]) length for five islands in the Mississippi River. Positive represents fore side and negative represents zee side. Figure 8 (opposite page, bottom). Mean difference in overall transect length for each of five Mississippi River islands combining fore and zee sides. Error bars represent (± 1 SE). Figure 9 (above). Mean side slopes of five Mississippi River Islands for 2007 and 2008: (a) fore sides and (b) zee sides. 472 Southeastern Naturalist Vol. 10, No. 3 No statistically significant changes in island elevation were witnessed from 2007 to 2008 (F = 0.59; P = 0.44). A Cochran-Armitage trend test was used to determine the direction of change (increasing or decreasing) in elevation from 2007 to 2008 and showed that all islands collectively were decreasing in elevation (Trend test statistic = -4.31), especially at lower elevations. However, independently, peak-elevation increases of three islands were not statistically significant: Densford Bar (9.79 to 10.21 m), Sunrise Towhead (6.51 to 7.28 m), and Wardlow Bar (8.53 to 10.3 m). Dean Island and Keyes Point decreased, also insignificantly, in elevation from 10.93 to 10.11 m and from 8.19 to 7.32 m, respectively. Overall, morphological changes ranged from minor, on Dean Island, to substantial changes on all transects on Densford Bar (Fig. 6). Discussion We documented the short-term changes in morphology of five islands in the Mississippi River from 2007 (a low-water year) to 2008 (a high-water year). As hypothesized, island morphology changed significantly, but only in width (i.e., transect length). While there were elevational changes, the observed changes in peak elevation were not statistically significant. This result is not surprising given that flooding events were of relatively low magnitude in 2008 and did not affect upper island elevations. Biologically, however, the elevation change on the lower parts of the islands had dramatic impacts on vegetation. The three islands that increased in elevation (Densford Bar, Sunrise Towhead, and Wardlow Bar) experienced a minimum of 42 cm aggradation during this study period, which could be detrimental to perennial plant species now buried under recent sedimentation. However, these results are not consistent with % woody vegetation cover. Based on contemporary low-water images, Densford Bar had the highest % bare sand of any island and is only sparsely vegetated (≈1.5% woody vegetation; Table 1), but gained elevation. In contrast, Sunrise Towhead had the second highest % woody cover (44.0%) and showed a gain in elevation (aggradation). Previous studies have suggested that high % woody vegetation cover on islands leads to greater sediment retention (see Shull 1922, 1944); however, contemporary lowwater images suggest that other factors are important for sediment retention on islands (i.e., distance to mainland, distance from major turn in river, dike field proximity, etc.). To test possible correlative factors affecting island morphology, we assessed proximity to nearest wing dam field and distance to the mainland with aerial photos. No apparent correlation with proximity to wing dams was observed; however, one striking similarity of islands that gained elevation was their respective distance to the mainland. Based on contemporary low-water images, Densford Bar was approximately 167 m from the mainland, Sunrise Towhead was 50 m, and Wardlow Bar 129 m. The two islands that degraded were Dean Island (344 m away) and Keyes Point (229 m away). We speculate that close proximity to the mainland reduces flow and ultimately affects sediment retention as high water moves over the floodplain. Kesel et al. (1974) found that sediment thickness and texture decreased rapidly away from the main channel following the 1973 flood 2011 J.E. Moore, S.B. Franklin, D. Larsen, and J.W. Grubaugh 473 in the Mississippi River Valley. It is likely that islands close to the mainland are indeed extensions of the nearby floodplain and receive floodplain-like deposits. Overbank sedimentation has shown similar responses in other portions of the Mississippi River. Kesel et al. (1974) found sediment deposition as much as 53 cm thick on natural levees following the 1973 event, and Gruelich et al. (2007) found 15 cm of sediment deposition in a low-lying area of Sunrise Towhead (different area from this study). European rivers showed similar trends, but differed signifi- cantly in terms of catchment area; therefore, comparisons are difficult (Walling and He 1998). One other factor that deserves scrutiny is the grain size of the sediment that comprises these islands. Personal observations (J.E. Moore) indicate that sand is the dominant substrate, at least in the upper 30 cm of soil. Below 30 cm, there appears to be a silty/clay layer that is likely the result of slow receding waters laiden with sediment from previous floods. Thus, more intense floods may be needed to move larger sediment, such as the upper sandy layer, away from islands. Morphological changes were island specific, but overall patterns occurred. Although factors controlling fore-side versus zee-side sedimentation and erosion are complex, the islands all showed that aggradation on one side was balanced by erosion on the other. In an extreme case, the entire mass of Keyes Point island shifted toward shore (in this case, away from the main channel), which is supported by higher sediment load in the water column at the closest distance to the islands (Fig. 5c, d). In addition, the river velocity and sediment transport at the heads of islands and, to a lesser extent, fore sides appeared greater than at the toe or zee sides, suggesting increased intensity of flooding and sediment movement in these areas. Although not statistically significant (due to a small sample size and only moderate flow), the difference may be ecologically relevant (see below). For example, the greatest relative changes in transect length occurred at the head end of the islands. In addition, the velocity data corroborate observations of greater slope angles on fore sides. Such dynamics might be expected to lead to a shift downstream for islands, as in bar migration (Osterkamp 1998), but our islands only shifted laterally. The pattern suggests a shift in flow intensity from one side to the other (lateral movement), but a constant structure (spur dike or wing dam) that holds the island in transverse place. While our data are notably of short duration, this concept deserves further scrutiny with more data. Island changes in transect lengths were similar for fore and zee sides, suggesting island-specific shifting. However, contemporary islands were notably longer than historical (1890) islands as would be expected from accumulation of finer sediments on downstream margins (Cobb 1999). The observed changes in island morphology relative to historic data may be the result of anthropogenic changes in the Mississippi River. Flood stage for a given discharge event has increased 2–4 m over the past century (Criss and Shock 2001). In modified river systems, such as the Mississippi, it has been shown that the most severe effects of human modification were in stretches that include cut-offs, spur dikes, and levees (Belt 1975, Pinter et al. 2000, Remo et al. 2009, Stevens et al. 1975), all of which were found in our study reach. Cut-offs lead to 474 Southeastern Naturalist Vol. 10, No. 3 increased slope, increased power, and a widening and deepening of the channel. Cobb (1999) observed a degradation of river thalweg between rm 630 and 780, a reach encompassing our study area. At the same time, dike systems lead to a localized flattening of channel slope, increased channel roughness, vertical accretion of bars (and subsequently islands), increased main channel volume, and stage reductions at low discharge. For nearly the entire lower Mississippi River (rm 315–954), an overall decrease in sandbar area was observed from 1948 to 1988 (Cobb 1999). The lack of elevation change during the two-year study is not surprising. Islands tend to build in height (rapid deposition) until they reach the floodplain level (Kellerhall 1976), at which time deposition will decrease. Aerial photographs indicate this is the case for at least two of the islands (Sunrise Towhead and Wardlow Bar). If islands build to the level of the floodplain and the thalweg has incised as described above (Cobb 1999), i.e., lowering the location of the thalweg relative to floodplain height, then contemporary islands should have greater relief and steeper slopes than 1890 islands. Thus, a threshold exists and once reached influences the rate of island morphological change. Our data compared to islands in 1890 corroborate this major change in island morphology. Because islands may serve as important landscape elements for biodiversity, understanding their dynamics is important. It is obvious that a major flood event leads to significant changes in island morphology, but the change seems to be a shift rather than overall loss. Herbaceous plant richness decreased drastically when examining the total species pool on all five islands (77 to 34 species) and average island species pool (30 to 17 species) from 2007 to 2008 (Moore et al. 2011). Thus, islands do serve as dynamic but continuous morphological structures in the riparian landscape that provide open areas for pioneer plant and animal taxa. For example, Least Terns are highly impacted by increases in river stage that decrease available nesting site area; thus, only higher elevation islands may provide useful nesting grounds for this endangered species (Dugger et al. 2002). Likewise, a strong relation exists between vegetation and fluvial dynamics (Cooperman and Brewer 2005), with the two main driving variables of vegetation being the flood/flow pulse and sediment dynamics (Steiger et al. 2005). Thus, the study of island morphological dynamics related to their ecological services as habitat is of management concern and is an important direction for future research. Equal importance should also be placed on the effects of lateral “movement” of islands within the main channel. Acknowledgments The authors wish to thank Jerry Garrett from the USGS for permitting the use of the ACDP and for also helping with the interpretation of data. Michael Thron of the US Army Corps of Engineers provided discharge data. We also thank the Meeman Biological Field station and Darrell Van Vickle for logistic support and boat travel. This manuscript was also greatly improved by the comments of Dr. Jerry Miller and two anonymous reviewers. 2011 J.E. Moore, S.B. Franklin, D. Larsen, and J.W. Grubaugh 475 Literature Cited Belt, C.B. 1975. The 1973 flood and man’s constriction of the Mississippi River. Science 189:681–684. Cobb, S.P. 1999. Biological Assessment: Interior Population of the Least Tern, Sterna antillarum, Regulating Works Project, Upper Mississippi River (River Miles 0–195) and Mississippi River and Tributaries project, Channel Improvement Feature, Lower Mississippi River (River Miles 0–954.5, AHP). US Army Corps of Engineers Mississippi Valley Division, Mississippi River Commission, Vicksburg, MS. Collins, M.J., and J.C. Knox. 2003. Historical changes in Upper Mississippi River water areas and Islands. Journal of the American Water Resources Association 39:487–500. Cooperman, M.S., and C.A. Brewer. 2005. Relationship between plant distribution patterns and the process of river island formation. Journal of Freshwater Ecology 20:487–501. Criss, R.E., and E.L. Shock. 2001. Flood enhancement through flood control. Geology 29:875–878. Dugger, K.M., M.R. Ryan, D.L. Galat, R.B. Renken, and J.W. Smith. 2002. Reproductive success of the interior Least Tern (Sterna antillarum) in relation to hydrology on the Lower Mississippi River. River Research and Applications 18:97–105. Gruelich, S., S.B. Franklin, T. Wasklewicz, and J.W. Grubaugh. 2007. Hydrogeomorphology and forest composition of sunrise towhead island in the Lower Mississippi River. Southeastern Naturalist 6:217–234. Gurnell, A.M., and G.E. Petts. 2002. Island-dominated landscapes of large floodplain rivers: A European perspective. Freshwater Biology 47:581–600. Hudson, P.F., H. Middelkoop, and E. Stouthamer. 2008. Flood management along the lower Mississippi and Rhine rivers (The Netherlands) and the continuum of geomorphic adjustment. Geomorphology 101:209–236. Hupp, C.R., and W.R. Osterkamp. 1996. Riparian vegetation and fluvial geomorphic processes. Geomorphology 14:277–295. Kellerhall, R. 1976. Stable channels and gravel-paved beds. American Society of Civil Engineers Proceedings, Journal of Waterways and Harbors 93:63–84. Kesel, R.H., K.C. Dunne, R.C. McDonald, and K.R. Allison. 1974. Lateral erosion and overbank deposition of the Mississippi River in Lousiana caused by 1973 flooding. Geology 2:461–464. Kostaschuk, R., J. Best, P. Villard, J. Peakall, and M. Franklin. 2004. Measuring flow velocity and sediment transport with an acoustic doppler current profiler. Geomorphology 68:25–37. Lane, S.N. 1998. The use of digital terrain modeling in the understanding of dynamic river channel systems. Pp. 311–341, In S.N. Lane, J.H. Chandler, and K.S. Richards (Eds.). Landform Monitoring, Modeling, and Analysis. Wiley, Chichester, UK. Lekarczyk, M.T., and T.A. Wasklewicz. 2003. Island morphometry: 55 years of change on the northern lower Mississippi. M.Sc. Thesis. The University of Memphis, Memphis, TN, 36pp. Maser, C., and J.R. Sedell. 1994. From the Forest to the Sea: The Ecology of Wood in Streams, Rivers, Estuaries, and Oceans. St. Lucie Press, Delray Beach, FL. Mississippi River Commission. 1890. Surveys of the Mississippi River. Available online at http://alabamamaps.ua.edu/historicalmaps/MississippiRiver/index.html. Accessed 3 January 2011. 476 Southeastern Naturalist Vol. 10, No. 3 Moore, J.E. 2011. Using Mississippi River islands to understand plant community dynamics. Ph.D. Dissertation. The University of Memphis, Memphis, TN. Moore, J.E., S.B. Franklin, and J.W. Grubaugh. 2011. Herbaceous plant community responses to fluctuations in hydrology: Using Mississippi River islands as models for plant community assembly. Journal of the Torrey Botanical Society 38:175–189. Nanson, G.C., and J.C. Croke. 1992. A genetic classification of floodplains. Geomorphology 4:459–486. Nielsen, D.N., R.G. Rada, and M.M. Smart. 1984. Sediments of the Upper Mississippi River: Their sources, distribution, and characteristics. Pp. 67–98, In J.G. Wiener, R.V. Anderson, and D.R. McConville (Eds.). Contaminants in the Upper Mississippi River. Butterworth Publishers, Stoneham, MA. Osterkamp, W.R. 1998. Processes of fluvial island formation, with examples from Plum Creek, Colorado and Snake River, Idaho. Wetlands 18:530–545. Pinter, N., R. Thomas, and J.H. Wlosinski. 2000. Regional impacts of levee construction and channelization, middle Mississippi River, USA. Pp. 351–361, In J. Masalek, W. Ed Watt, Evzen Zeman, and Friedhelm Sieker (Eds.). Flood Issues in Contemporary Water Management. Kluwer Academic Publishers, Boston, MA. Remo, J.W.F, N. Pinter, and R. Heine. 2009. The use of retro- and scenario-modeling to assess effects of 100+ years of river engineering and land-cover change on Middle and Lower Mississippi River flood stages. Journal of Hydrology 376:403–416. SAS 2000. SAS/STAT User’s Guide, Version 8.1. SAS Institute, Inc., Cary, NC. Shull, C.A. 1922. The formation of a new island in the Mississippi River. Ecology 3:202–206. Shull, C.A. 1944. Observations of general vegetational changes on a River Island in the Mississippi River. American Midland Naturalist 32:771–776. Smith, J.W., and R.B. Renken. 1991. Least Tern nesting habitat in the Mississippi River Valley adjacent to Missouri. Journal of Field Ornithology 62:497–504. Steiger, J., E. Tabacci, S. Dufour, D. Corenblit, and J.L. Peiry. 2005. Hydrogeomrphologic processes affecting riparian habitat within alluvial channel-floodplain river systems: A review for the temperate zone. River Research and Applications 21:719–737. Stevens, M.A., D.B. Simons, and S.A. Schumm. 1975. Man-induced changes of middle Mississippi River. American Society of Civil Engineers, Journal of the Waterways, Harbors, and Coastal Engineering Division 101:119–133. Theiling, C. 1998. River geomorphology and floodplain habitats. Pp. 4-1–4-21, In R.L. Delaney, K. Lubinski, and C. Theiling (Eds.). Ecological Status and Trends of the Upper Mississippi River System 1998. A Report of the Long Term Resource Monitoring Program, US Geological Survey, LaCrosse, WI. Walling, D.E., and Q. He. 1998. The spatial variability of overbank sedimentation on river floodplains. Geomorphology 24:209–223. Wasklewicz, T.A., P.S. Liu, and S. Anderson. 2004a. Geomorphic context of channel locational probabilities along the Lower Mississippi River, USA. Geomorphology 63:145–158. Wasklewicz, T.A., J.W. Grubaugh, S.B. Franklin, and S. Greulich. 2004b. 20th-century stage trends along the Mississippi River. Physical Geography 25:208–224.