134 266 ”THS THFSIS f. lllllllllllllllllllllllllllllllllllllllllll!Jillilllllllllllll 31293 02079 6383 K} (x ) This is to certify that the thesis entitled THE DISTRIBUTION, DIEL HIGRA'IION, AND GROWTH OF THE GRASS SHRDIP :Pmmms PALUDOSUS IN THE KISSDMEB RIVER FLOODPIAIN BCOSYSTEH presented by Kelly James Uessell has been accepted towards fulfillment of the requirements for H S degree in Fisheries and Wildlife flaw, Major professor Date 16 December 1999 0-7639 MS U is an Affirmative Action/Equal Opportunity Institution LIBRARY Michigan State University PLACE IN RETURN BOXto remove this checkout from your record. TO AVOID FINE return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE W00 elem.m 14 THE DISTRIBUTION, DIEL MIGRATION, AND GROWTH OF THE GRASS SHRIMP PALAEMONETES PALUDOSUS IN THE KISSININIEE RIVER FLOODPLAIN ECOSYSTEM By Kelly James Wessell A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Fisheries and Wildlife 1 999 ABSTRACT THE DISTRIBUTION, DIEL MIGRATION, AND GROWTH OF THE GRASS SHRIMP PALAEMONETES PALUDOSUS IN THE KISSIMMEE RIVER- FLOODPLAIN ECOSYSTEM By Kelly James Wessell Historically, the Kissimmee River meandered over an extensive floodplain wetland. In response to catastrophic flooding and settlement pressures in Central Florida, the Kissimmee was channelized, converting the complex, braided channel into a straightened canal. The result has been a sharp decrease in flinging wetland habitat and the associated biota. Soon after channelization was completed, environmental concerns prompted the State of Florida to start examining options for restoration to reestablish the river's natural hydrology and restore lost wetland habitat. The grass shrimp, Palaemonetes paludosus has been identified as a keystone invertebrate species in this system, and this study was designed to examine its distribution, diel migration, and growth within the two dominant macrophyte communities of the Kissimmee River riparian marsh: Nuphar and Polygonum. Results indicated that grass shrimp were more abundant in Polygonum beds. This species also showed no well-defined diel migration, although I have found a significant vertical pattern in some instances. P. paludosus growth is highest on periphyton and Polygonum leaves. Grass Shrimp distribution may be explained by their decreased susceptibility to predation because of the higher stem ' densities inherent in Polygonum beds. ACKNOWLEDGMENTS I wish to first thank my major professor, Dr. Richard W. Merritt for the incredible opportunity to work on the Kissirmnee River and for all of his guidance on this project. I am truly lucky to have such a supportive mentor and advisor. Special thanks go to my committee members, Dr. Thomas Burton and Dr. Thomas Coon, for their leadership and support throughout this research and for providing valuable suggestions in reviewing this manuscript. I am indebted to Ngoc Kieu for her tireless help in the field and with sample processing. I am also grateful to many other people for help with my field and lab work: Eric Naguski, Tom Burton, Don Uzarski, Michael Higgins, Becky Blasius, John Wallace, and Osvaldo Hernandez. Thanks also go to Dr. Carlos de la Rosa, Dr. David Anderson, and the reSt of the crew at Riverwoods Field Laboratory for the use of lab space and living quarters and to the South Florida Water Management District for providing travel funds. Bridgette VandenEeden did biomass calculations, and Jim Brock provided data loggers and sondes for measuring dissolved oxygen. I also wish to extend my thanks to Dr. Ken Cummins, who provided invaluable help in every stage of this study. Without his guidance and knowledge, this study would not have been possible. iii TABLE OF CONTENTS List of Tables ...................................................................................................................... v List of Figures .................................................................................................................... vi History and Management of the Kissimmee River .............................................................. 1 The Pre-Channelized Kissimmee River ................................................................... 1 The Kissimmee River Channelization ..................................................................... 3 The Pool B Demonstration Project .......................................................................... 5 The Kissimmee River Restoration Project ............................................................... 7 Grass Shrimp Ecology ....................................................................................................... 10 Study Objectives ................................................................................................................ 14 Methods and Materials ....................................................................................................... 15 Study Site ............................................................................................................... 15 Distribution and Abundance .................................................................................. 17 Die] Migration ........................................................................................................ 18 Growth Studies ...................................................................................................... 20 Results and Discussion ...................................................................................................... 22 Distribution and Abundance .................................................................................. 22 Die] Migration ........................................................................................................ 30 Growth Studies ...................................................................................................... 35 General Discussion ................................................................................................ 37 Conclusions ........................................................................................................................ 41 Literature Cited .................................................................................................................. 42 iv Table 1. Table 2. Table 3. Table 4. Table 5. Table 6. Table 7. Table 8. Table 9. Table 10. Table 11. LIST OF TABLES ANOVA table for distribution data. Dependant variable: Log(# P. paludosus+l ) ............................................................................................ ......... 23 Number of grass shrimp per sample calculated by season and plant type. Log-transformed values were used in the statistical analysis ................... Results from Fisher's LSD tests on season*plant combinations when considering average number of shrimp per sample. Tests were done on log-transformed values. Comparisons were not done between seasons. AN OVA table for distribution data. Dependant variable: Log(P. paludosus biomass+l). ............................................................................. Mean grass shrimp biomass (mg) per sample calculated by season and plant type. Log-transformed values were used in the statistical analysis Results from F isher's LSD tests on season*plant combinations when considering average shrimp biomass per sample. Tests were done on log-transformed values. Comparisons were not done between seasons. ANOVA table for the test comparing average shrimp biomass across seasons and plant types. Dependent variable: Mean Individual P. paludosus biomass sample". Data points with biomass=0 were ......... 23 ......... 23 ......... 24 ......... 24 ......... 24 removed for this analysis and biomass values were left untransformed. .......... 27 Seasonal effect on grass shrimp density (number per sample) and average biomass (mg) ........................................................................................ 27 ANOVA table comparing P. paludosus and H. azteca log transformed biomass values. Dependent variable: Log(Mass+l) ......................................... 29 Results from the Kruskal-Wallis AN OVA testing differences in grass shrimp movement in and out of the floodplain during the day and at night. .................................................................................................................. 3 1 Results form the Kruskal-Wallis AN OVA testing differences in grass shrimp vertical movement during the day and at night. .................................... 31 LIST OF FIGURES Figure 1. The Kissimmee River basin from the 4229 km2 upper basin headwaters in the Kissimmee Lakes through the 1200 lqn2 lower basin and into Lake Okeechobee. Location of the basin in south-central Florida is indicated in the inset. (From Merritt et al. 1996) .............................................. 2 Figure 2. Pre and post-channelization distributions of dominant wetland plant communities on the Kissimmee River floodplain (modified from Toth et al. 1995). ........................................................................................................ 4 Figure 3. Diagram of the Kissimmee River lower basin showing Pools A-E and water control structures. This work was conducted in the lower Pool B remnant channel, the area affected by the Pool B Demonstration Project, denoted by the star. ............................................................................... 6 Figure 4. Current (a) and predicted post-restoration (b) vegetation in the Kissimmee River-floodplain ecosystem (Pools B, C, and D) (Modified from Toth et al. 1995). ...................................................................... 8 Figure 5. Conceptual model relating the autotrophic and heterotrophic pathways, the associated keystone invertebrates, and suggested links to higher trophic levels in the Kissimmee River riparian marsh ecosystem (modified from Merritt et al. 1999) .................................................................. 11 Figure 6. Relative abundance of selected invertebrate taxa (percentages based on numbers of animals) in the two dominant littoral fringe habitats of the Kissimmee River riparian marsh: Nuphar luteum and Polygonum densiflorum. Data from Merritt et al. 1999. .................................................... 12 Figure 7. Diel dissolved oxygen curve comparing 2 riparian marsh habitats (Nuphar and Polygonum) and Open water. (From sonde data collected on 5-6 May 1998) ........................................................................................... 16 Figure 8. Diagram showing Breder trap placement in the Kissimmee River riparian littoral fringe macrophyte habitats. (a) Top view; (b) Side view (Shaded to Show traps opening in opposite directions.) ......................... 19 Figure 9. Mean number (a) and mean biomass (b) (:tSE) of P. paludosus in the dominant macrophyte habitats of the Kissimmee River riparian marsh. ......... 25 Figure 10. Comparison of mean P. paludosus vs. H. azteca biomass per sample in the two dominant littoral fringe macrophyte habitats of the Kissimmee River riparian marsh. ....................................................................................... 29 vi Figure 11. Mean number :t SE of P. paludosus captured moving from the channel to the floodplain (CH>FP) and from the floodplain to the channel (F P>CH) during the day (a) and at night (b). .................................... 32 Figure 12. Mean number £813 of P. paludosus captured in top and bottom Breder traps (a) during the day and (b) at night. Significant differences at p<0.05 denoted by double asterisks. ................................................................ 33 Figure 13. Net dry weight of P. paludosus in each of the three food treatments. Data derived by subtracting the mean initial weight from each dried shrimp that survived the growth experiment. .................................................. 36 vii THE DISTRIBUTION, DIEL MIGRATION, AND GROWTH OF THE GRASS SHRINIP PALAEMONE T ES PALUDOSUS IN THE KISSIMMEE RIVER- FLOODPLAIN ECOSYSTEM HISTORY AND MANAGEMENT OF THE KISSIMMEE RIVER THE PRE—CHAMVELIZED KISSIMMEE RIVER The Kissimmee River originates in the Kissimmee Lakes region of central Florida just south of Orlando and makes up the northern portion of the Kissimmee-Lake Okeechobee-Everglades watershed (Figure 1). Historically, the river was a complex braided channel that meandered approximately 166 km within a 1.5-3 km wide floodplain. In its historical condition, water levels fluctuated on a seasonal basis and discharge exceeded 11 m3 / second during 90%-95% of the period of record, with highest discharges typically occurring at the end of the wet season (September-November) (Koebel 1995). Prior to channelization, 94% of the floodplain (16,920 ha) was inundated over 50% of the time (Shen et a1. 1994). When inundated, water depths on the floodplain were generally 0.3-0.7 m, with depths over 1 m occurring on over 40% of the floodplain for at least one-third of the period of record (Toth 1990). This system was unique to North American river-floodplain ecosystems in that it had an extremely well-developed fringing floodplain wetland that occurred along most of the river's length (Koebel 1995). This wetland habitat along with the seasonal fluctuation in water levels and nearly constant river-floodplain connectivity sustained a highly diverse invertebrate community including caddisflies, dragonflies, darnselflies, water bugs, water beetles, isopods, amphipods, decapods, midges, and mollusks (Koebel 1995). Upper Basin . g“- - , , I . lower Basin Lake Okeechobee Figure 1. The Kissimmee River basin from the 4229 km2 upper basin headwaters in the Kissimmee Lakes through the 1200 lcrn2 lower basin and into Lake Okeechobee, Location of the basin in south-central Florida is indicated in the inset. From Merritt et a1 1996. In addition, the Kissimmee River supported as many as 35 species of fish, including a world class largemouth bass fishery (Trexler 1995), 16 species of wading birds, 16 species of waterfowl, and six other water bird species (Weller 1995). Often, the life cycles of invertebrates, fish (Junk et a1. 1989), and water birds (Weller 1995) are closely tied with seasonal flooding. Greater fish recruitment occurs in years with smooth increases in water levels and floods of high amplitude and long duration (Payne 1986). In other words, the natural fluctuations in water levels were necessary to sustain much of the Kissimmee River's fauna. THE KISSIMMEE RIVER CHANNELIZA TION In response to catastrophic flooding and settlement pressures in Central Florida, the US. Army Corps of Engineers began a project in 1962 that channelized the river. The once meandering river was converted to a straightened, 9 m deep by 100 m wide canal and impounded into a series of five relatively stagnant storage reservoirs, so that water levels no longer fluctuated on a seasonal basis. The project affected approximately 161 km of river and resulted in the conversion of 14,000 ha of floodplain wetland to pasture (Toth et al. 1995) (Figure 2). The elimination of the seasonal water level fluctuations and the extensive loss of wetland plant communities have had significant effects on both invertebrate and vertebrate communities (Merritt et al. 1996). Additionally, low flow through remnant channels resulted in a build up of senescent plant material that covered the sand substrate with large amounts of organic matter, greatly increasing the biological oxygen demand of the system (Toth 1990). Specific effects of channelization on the biological communities include a 90% decrease in the number of fl wrrrow [:3 surroususn WET PRAIRIE N on m n SPILLWAYILOCK Figure 2. Pre and post-channelization distributions of dominant wetland plant communities on the Kissimmee River floodplain (modified from Toth et a1. 1995). waterfowl and wading birds (Weller 1995), a decline in the proportion of game fish captured in surveys (Trexler 1995), and a shift in invertebrates to those more common in lentic systems (Harris et al. 1995). Soon after channelization was completed in the early 1970's, environmental concerns were raised, and this prompted the State of Florida and the US. Army Corps of Engineers to begin looking at options for restoration. Numerous structural and nonstructural plans were considered, including modification of upper basin lake regulation schedules, pool stage manipulations, earthen plugs, and backfilling (Koebel 1995). Analysis of these alternatives revealed that many of the plans were not feasible or did not meet restoration objectives. In late 1983, the U. S. Army Corps of Engineers narrowed its restoration focus down to two alternatives, including partial backfilling of the C-38 canal (the channelized river) and a combined wetlands approach that included pool stage manipulation and impounded wetlands. THE POOL B DEMONSTRA TION PROJECT In 1984, the South Florida Water Management District (SF WMD) initiated a demonstration project to evaluate the effects of increased flow and floodplain inundation within the channelized river. This was accomplished with a series of three weirs that directed additional flow through three remnant channels along the section of the river known as Pool B (Figure 3). The results of the Demonstration Project was to reestablish prechannelization floodplain inundation patterns through small portions of three remnant channels in Pool B, resulting in the prolonged flooding of nearly 20% and periodic flooding of approximately 75% of the historical floodplain in this area. ‘ Ljr I!" if {571,9} {1:75}: 1 - A Welt *SWPIO Slto 9. . «m- . . ' ' '=:.’.’r‘::r'c’:(cigfgl;—r' . Figure 3. Diagram of the Kissimmee River lower basin Showing Pools A-E and water control structures. This work was conducted in the lower Pool B remnant channel, the area affected by the Pool B Demonstration Project, denoted by the star. Almost immediately upon completion of the Demonstration Project, wetland plant communities began to revert to those more characteristic of the historical system (Figure 2). Additionally, accumulations of dead organic matter were washed from the remnant channel into the C-3 8, restoring the natural sand substrate and reducing the biological oxygen demand (Koebel 1995). Reintroduction of flow through this area also has resulted in the colonization of invertebrate taxa more characteristic of lotic systems (Toth 1993) and increases in game fish (W ullschleger et al. 1990) and waterfowl (Toland 1990) relative to the unrestored portions of the river. The success of the demonstration project prompted the State of Florida to approve the SFWMD plan to backfill approximately 35 km of channelized river, eventually resulting in the restoration of about 11,000 ha of the historical floodplain wetland. THE KISSIMMEE RIVER RESTORA HON PROJECT The restoration of the Kissimmee River-floodplain ecosystem in central Florida is the largest project of its kind ever attempted. Major objectives of the restoration are to return the system to its historical condition in which it supported large populations of wading birds and waterfowl along with an outstanding sport fishery. This will be accomplished by returning the river to a state in which its flow, seasonal discharge patterns, floodplain inundation frequencies, and stage recession rates are comparable to the prechannelization conditions (Koebel 1995). As the floodplain and remnant channels are inundated with flowing water, the system is expected to shift from one dominated by pasture to one of riparian marsh habitat (Figure 4) (Merritt et al. 1999). Aquatic 1 80% \ Wetland Shrub Broadleaf Marsh ‘ T 8.0% (a) 10.0% , WW Wet Prairie Human Influenced '- W 15.0% o . ii” I 13'0 /0 , Upland Shrub 8.0% Pasturej 38.0% Wetl:n(<)i(yShrub_ Wet Prairie (b) ' ° 29.0% Broadleaf Marsh 62.0% Figure 4. Current (a) and predicted post-restoration (b) vegetation in the Kissimmee River-floodplain ecosystem (Pools B, C, and D) (Modified from Toth et a1. 1995). During the restoration process, ecological conditions will be monitored to evaluate the project’s success at restoring the habitats and populations of river and floodplain species. Invertebrates are an integral part of aquatic ecosystems and can serve as a useful indication of the extent to which this system is responding to restoration efforts. In the primarily lentic environment of the channelized Kissimmee River, invertebrate production has been restricted to the littoral zone with little floodplain habitat available. Since most productivity in large, undisturbed rivers occurs in the floodplain (Harris et al. 1995; Junk et al. 1989), the increase in floodplain habitat and reestablishment of the natural hydrological regime will undoubtedly have significant impacts on invertebrate communities, and these changes can help predict the effects of restoration on higher trophic levels. Harris et a1. (1995) suggested several key elements of the Kissimmee's invertebrate community for study prior to and during the Kissimmee River restoration. Secondary productivity must be studied in order to get an idea of overall community structure and the availability of food to higher trophic levels. Examining drift dynamics and determining the movement of organisms and organic matter in and out of the floodplain would aid in understanding the functional linkages between the river and its floodplain. They also suggested that diet studies of waterfowl, wading birds, and fish would clarify the specific trophic linkages between invertebrates and their potential predators and enable predictions to be made regarding the colonization of the restored habitat by higher trophic levels. In addition, special attention should be paid to invertebrate indicator groups and keystone species such as the grass shrimp, Palaemonetes paludosus. A conceptual model has been developed by the SFWMD that identifies two key littoral fringe plant communities: Nuphar Iuteum and Polygonum denszflorum (Figure 5). The grass Shrimp Palaemonetes paludosus (Gibbes) is particularly abundant in each of these plant communities (Figure 6) (Merritt et al. 1996, 1999). Because of its relative abundance and large size, P. paludosus is a significant link between primary producers and higher trophic levels in the Kissimmee River-floodplain system (Merritt et a1. 1999), including vertebrate predators that detect their prey visually, such as many large fish, wading birds, and waterfowl (Figure 5). GRASS SHRIMP ECOLOGY The North American species of Palaemonetes represent what are probably the most poorly known of our freshwater decapoda. This is partly due to their limited commercial value and lack of extensive fieldwork (Strenth 1976). The existing knowledge of the group consists of work done on primarily marine and estuarine species such as P. pugio Holthuis (Welsh 1975; Kneib 1985; Greg and Fleeger 1995, 1996; Eggleston et al. 1998; Vemberg and Piyatiratitivorakul 1998; Cross et al. 1996) and P. vulgaris (Miller et a1. 1995; Sogard and Able 1994; Coen et al. 1981), with a relatively scant ecological knowledge base of freshwater species such as P. paludosus. Data from studies of marine and estuarine Palaemonetes suggest that interactions among grass shrimp, benthic predators, and nektonic omnivores have strong direct and indirect effects upon benthic faunal densities and community composition (Posey and Hines 1991). Additionally, Pringle and others (1993) have determined that omnivorous shrimp are important organizers of lotic community structure and play a key role in 10 .I-fl q. _ . .- - - .. h 'a- no, N x"! I-w Mr“. \ "T" ""9“ "" 1:;q,;,f;}f...l"1%.;31391434;415F533 II i: I I iii A?! E 1 RI s“; . 1.41% '&D::l.':i-Illl‘ll'in _ .=- ".-"l'|‘l -.r - 1. I 91% 'n‘u.‘ a" ' 'i ‘ ‘ ' '1]. -- '-. . " ' Autotrophic Heterotrophic Path Path "'4:- I \ ., Hyalella azteca Palaemonetes paludosus l Tactile Feeding Birds & Juvenile Game Fish Visual-feeding Wading Predaceous Birds & Large Game Fish Invertebrates Figure 5. Conceptual model relating the autotrophic and heterotrOphic pathways, the associated keystone invertebrates, and suggested links to higher trophic levels in the Kissimmee River riparian marsh ecosystem (modified from Merritt et al. 1999) 11 .33 .3 8 £52 89a Sam .EsxoSe—Sn Enxewbok 28 E385 Bafizz ”€38 giant REM ova—Ema 05 no 3853 Oman“ ~88:— Bdfiaoe 95 2: E @1255 me 338:: cc woman mom8aoou03 SE 0358.52: Bonus—um «0 823.8% 9320M .w onE 2.283820 / 12.25:; .L $323.1 .1 Eauefiwueow Eaguéem EBB»: REES/N 12 reducing sediment cover on rock substrata and enhancing algal populations. Grass shrimp have also been shown to be important predators of phytoplankton grazers such as cladocerans, and indirectly contribute to higher turbidity and eutrophication (Samuels and Mason 1998). Palaemonetes paludosus is widespread in the eastern United States, and is found as far west as eastern Texas (Strenth 1976). This grass shrimp Species is especially abundant in the extensive marshes and swamps of Southern Florida (Kushlan and Kushlan 1980), and can be found among emergent vegetation, snags, or clinging to the undersides of vegetation mats. Its body is transparent and ranges in size from 3-25 mm long (Meehean 1936). Although algae are its major food, P. paludosus is omnivorous and may feed on dead leaves, insects and other benthic coarse particulate organic matter (CPOM) (Beck and Cowell 1976). Some authors (e.g., Meehan 1936) describe P. paludosus as an intolerant Species that is usually found in clean water with high dissolved oxygen. Other investigators (Kushlan and Kushlan 1980) maintain that they are extremely tolerant to low dissolved oxygen (DO) levels and have noticed them swimming at the surface in areas of low DO, presumably taking advantage of the oxygen diffusing across the surface layer. In Florida, ovigorous females have been collected throughout the year (Dobkin 1963). However, the percentage of ovigorous females peaks when the water levels rise in the summer, and during early fall, when water levels are usually the highest (Kushlan and Kushlan 1980). This suggests that restoration of the natural hydrological regime in the Kissimmee River-floodplain ecosystem could have a positive effect on grass shrimp production. 13 Females typically produce broods of 8-85 eggs during their one year life cycle and carry eggs and zoeae for up to 2 months (Beck and Cowell 1976). Larval shrimp (zoeae) hatch from eggs after an incubation period of approximately 12-14 days at 26- 28°C (Beck and Cowell 1976). The larvae molt three times before they reach sexual maturity. Hatching to maturity takes 2-3 months when water temperatures exceed 26°C, though cooler temperatures delay maturation (Beck and Cowell 1976). The frequency of molting and duration of larval life varies with the quantity and quality of food available (Broad 1957). STUDY OBJECTIVES I studied the distribution, diel migration, and growth of Palaemonetes paludosus to gain a better understanding of the functional linkage between the river and its floodplain for this species. This was aimed at providing information to predict how the restoration process and the resulting increase in floodplain habitat might affect the distribution and abundance of this species in the Kissimmee River-floodplain ecosystem. Four specific aspects of grass Shrimp ecology in the restored Kissimmee River-floodplain were examined to evaluate the importance and success of this species. These include: 1) A comparison of its seasonal distribution and abundance in the riparian marsh as a firnction macrophyte type; 2) A comparison of its total biomass in the Kissimmee River riparian marsh in relation to the other keystone invertebrate, H. azteca; 3) Determining its diel migration patterns between the river and floodplain; and 4) Determining its growth rate as a function of the type of food consumed. 14 METHODS AND MATERIALS STUDY SITE The study site is located in the lower Pool B remnant channel of the Kissimmee River near Lorida, Florida in the partially restored area affected by the Demonstration Project (Figure 3). Flow through this remnant channel is generally low compared to historical conditions. During 1988, this remnant channel had discharges between 0-11 m3/sec approximately 56% of the time and exceeded 26 m3/sec 31% of the time (Toth 1991). Water temperature ranged from about 25°C in the winter to 35°C in the summer. Generally, dissolved oxygen levels are considered poor and were highest during the spring and lowest during the late summer and early fall (Wullschleger et al. 1990). The two most dominant littoral fringe plant communities in this area are Nuphar Iuteum and Polygonum densiflorum, and are structured very differently from each other. Nuphar communities are characterized by relatively low stem densities (15/ m2) (Cummins et al. 1999), high light penetration, and are generally autotrophic (Merritt et al. 1999). Much of the primary productivity is due to the dense periphyton communities that colonize Nuphar stems, and this is potentially a rich food source for invertebrates (Cummins and Klug 1979). Polygonum beds, on the other hand, are characterized by very high stem densities (46/m2), with very little subsurface light penetration (Cummins et a1. 1999). Because of this, Polygonum communities are generally heterotrophic. Dissolved oxygen levels are relatively low in both plant bed types, but tend to be slightly higher in the Nuphar beds than in the Polygonum beds, differing by between 1-2 mg/L (Figure 7). Water depth in both macrophyte beds ranged between 45 cm and 110 cm during the study period. 15 :33 ~82 9m co 380:8 3% ocean 58.6 .83? come 98 Afiseembem Ea $23.35 $335 smug: outfit a mfiamfioo 255 combs 838m? .05 .N. 033 eat em NN cm 3 or 3. NF 9..w o v o _ _ _ e _ _ F _ _ _ 00.0 2323on I? .225 :30 II. 223:2 1? 00. F ooN c; O 00 (116w) on 00:.» - ooh 00.0 16 DISTRIBUTION AND AB UNDANCE The two dominant littoral fiinge habitats (Nuphar luteum and Polygonum densiflorum) of the Kissimmee River riparian marsh along the lower remnant channel of Pool B were sampled twice a year under different flow and water level conditions for two years. Vegetation was sampled by positioning a D-frame aquatic dip net (0.8 mm mesh) several centimeters into the sediments and vigorously moving it along the stems of the plants for 30 seconds. I moved the net from side to side and front to back and brought it through the plant bed to the surface to insure dislodgrnent of the associated invertebrates. On each sampling date, 18 samples were taken from both Nuphar and Polygonum beds. For each sample, the net and its contents were washed into an enamel tray and the coarse material was washed off by hand into a whirlpac bag and preserved with 70% EtOH. Samples were returned to the lab for sorting and measuring of invertebrates under a dissecting microscope. The number of P. paludosus and H. azteca were recorded for each Sample and biomass was determined fiom a length-biomass regression. Data from both summer samples (Aug 97 and June 98) were pooled and compared to data from pooled winter samples (Feb 98 and Feb 99). Mean number of individuals per sample by plant type and season were calculated before log-transforming data in order to adhere to the AN OVA assumptions of normality and homogeneity of variance. Differences in grass shrimp distribution by plant type and season were analyzed with log-transformed data (log(y+1)) in a 2-way analysis of variance with season as the first factor (2 levels: summer and winter) and plant type as the second factor (2 levels: Nuphar and Polygonum) (SAS Institute 1996). Interactions were further dissected with Fisher's LSD tests. 17 To determine the seasonal differences in average shrimp size, mean biomass per individual was calculated for each sample, and a 2-way AN OVA tested the effect of season and plant type on average shrimp size (SAS Institute 1996). In this analysis, the mean individual shrimp size was calculated only from samples containing at least one individual. Zero values were not included, and values were left untransformed. In order to evaluate the relative importance of the two keystone species in the two different macrophyte habitats of the Kissimmee River riparian marsh and test the SFWMD'S conceptual model (Figure 5), a two-way ANOVA was conducted with species (2 levels: P. paludosus and H. azteca) and plant (2 levels: Nuphar and Polygonum) as the treatment factors. In order to take into account the extreme differences in these two Species' sizes, mean biomass values per sample were used in the analysis. For all tests performed, effects were considered significant when p<0.05. Although analyses were done on log-transformed values, figures display the more biologically significant untransformed values. DIEL MIGRA HON Diel migration was examined by placing Breder traps at the margins of both Polygonum and Nuphar beds. Traps were placed in such a way as to capture grass shrimp as they move between the river channel and the floodplain and vertically within the water column (Figure 8). Traps were emptied just after sunrise and just before sunset to determine when shrimp were most likely to move between the floodplain and channel. The traps were set up in such a way as to preclude larger fish from gaining entry to 18 (3) Shrimp moving from floodplain to channel. Floodplain/Macrophyte Bed Remnant River Channel Shrimp moving from channel to floodplain. 0)) Water Level . . . 1'1: Figure 8. Diagram Showing Breder trap placement in the Kissimmee River riparian littoral fringe macrophyte habitats. (a) Top view; (b) Side view (Shaded to Show traps opening in opposite directions.) 19 minimize predation on captured shrimp. Previous studies show that Breder traps are an efficient means of capturing small fishes, and that the variety of invertebrates caught in these traps was equivalent to that of the fishes (Breder 1960). Data were pooled according to whether they came from day or night nms, and mean numbers of grass shrimp caught in tOp versus bottom traps and traps opening toward the floodplain versus traps opening toward the channel were calculated. Differences in both vertical and horizontal grass Shrimp movement were inferred from a Kruskal-Wallis non-parametric analysis of variance (SAS Institute 1996). GROWTH STUDIES To determine grass shrimp growth on different food types, I started with 180 shrimp of similar size. A subsample of 30 Shrimp were dried and weighed to get an estimate of the initial dry weight. The remainder of the shrimp were used in the food treatments. I examined 3 different food types commonly available to P. paludosus in the dominant macrophyte communities in the lower Pool B remnant channel: 1) Nuphar stems coated with dense periphyton 2) Conditioned Polygonum leaves 3) Fine detritus F Season 1 15.5454 15.5454 42.25 0.0001 Plant 1 6.7974 6.7974 18.47 0.0001 Season*Plant 1 1.8979 1.8979 5.16 0.0247 Error 140 51.5095 0.3679 Total 143 75.7502 Table 2. Number of grass shrimp per sample calculated by season and plant type. Log- transformed values were used in the statistical analysis. Marginal Mean Mean # Grass Log(Mean+l) (SE) Shrimp/Sample (SE) Summer 49.8 (10.4) 0.97 (0.10) Winter 2.2 (0.5) 0.32 (0.04) Nuphar 5.0 (1.2) 0.43 (0.06) Polygonum 47.1 (10.5) 0.88 (0.10) Table 3. Results from Fisher's LSD tests on season*plant combinations when considering average number of shrimp per sample. Tests were done on log-transformed values. Comparisons were not done between seasons. Season*Plant Mean # Grass Log(Mean+l) (SE) LSD Combination Shrimp/Sample LSE) SummeflNuphar 8.1 (2.2) 0.65 (0.09) B SummeflPolygonum 91.6 (18.3) 1.31 (0.16) A Winter*Nuphar 1.9 (0.8) 0.21 (0.06) A Winter*Polygonum 2.6 (0.5) 0.42 (0.06) A 23 Table 4. AN OVA table for distribution data. Dependant variable: Log(P. paludosus biomass+l). Source Df SS MS F Value Pr>F Season 1 24.5759 24.5759 25.50 0.0001 Plant 1 12.2978 12.2978 12.76 0.0005 Season*Plant 1 0.5271 0.5271 0.55 0.4608 Error 140 134.9361 0.9638 Total 143 172.3368 Table 5. Mean grass shrimp biomass (mg) per sample calculated by season and plant type. Log-transformed values were used in the Statistical analysis. Marginal Mean Mean Log(Mean+l) (SE) Biomass/Sample (SE) (23) Summer 313.6 (64.3) 1.58 (0.13) Winter 42.1 (11.4) 0.75 (0.12) Nuphar 57.2 (13.8) 0.88 (0.11) Polygonum 298.5(647) 1.46 (0.19 Table 6. Results from Fisher's LSD tests on season*plant combinations when considering average shrimp biomass per sample. Tests were done on log-transformed values. Comparisons were not done between seasons. Season*Plant Mean Log(Mean+1) (SE) LSD Combination Biomass/Sample (SE) (mg) SummefiNuphar 75.1 (19.3) 1.22 (0.15) B SummeflPOlygonum 552.1 (114.8) 1.93 (0.20) A Winter*Nuphar 39.3 (19.5) 0.52 (0.15) B Winter*Poiygonum 44.9 (12.0) 0.98 (0.15) A 24 120 g (a) 4,. 100 ~ 1 El Nuphar 8 1 i 5 80 ~ I Polygonum { 8 a In D. 40 - a: 3'1: 5 c 20 E o 4 Fit—— Summer Winter A 700 2’ l V (b) ! a 3% 600 - 5 El Nuphar l 1 fl ‘l : 0 ; E ‘5. 50° ’ I IPOlygonum E 1 I6 2 400 L '5’ In .9 300 . E: a O: 200 ~ In 3 100 ~ E .2 m o 1"?- Summer Winter Figure 9. Mean number (a) and mean biomass (b) (:hSE) of P. paludosus in the dominant macrophyte habitats of the Kissimmee River riparian marsh. 25 only 57.2 mg per Nuphar sample (df=1; F=12.76; p<0.0001) (Table 4). There was also a significant seasonal effect, with mean biomass per sample equal to 313.6 mg in the summer and only 42.1 mg in the winter (df=1; F=25.50; p=0.0005) (Table 5; Figure 9b). However, there was no significant season by plant interaction when considering mean biomass instead of numbers (df=1; F=0.55; p=0.4608) (Table 4). Results from Fisher's LSD test confirmed the AN OVA results and indicated that the plant effect was consistent throughout the year when considering log-transformed values for biomass instead of numbers (Table 6; Figure 9b). S l D . E r . l r 2 $11 . 51' There was no significant difference in size of shrimp between Polygonum and Nuphar plant communities (df=1; F=0.24; p=0.6241) (Table 7). However, shrimp sampled in the winter were significantly larger than those sampled in the summer (df=1; F=16.16; p<0.0001) (Table 7). The average grass shrimp captured in the summer samples weighted 7.97 mg, while the average shrimp captured in winter weighed 18.63 mg (Table 8). Higher numbers of grass shrimp were found in the summer samples as opposed to the winter samples (Table 8). These differences in average Shrimp size help explain the disagreement between the tests looking at numbers and the tests looking at biomass. The increase in average shrimp biomass in the winter samples makes up for the lower numbers of shrimp per sample and results in an insignificant interaction term when considering log-transformed values for biomass instead of numbers. Since summer samples contained a relatively large number of small shrimp while winter samples contained a smaller number of large shrimp, a reasonable inference is that 26 Table 7. AN OVA table for the test comparing average shrimp biomass across seasons and plant types. Dependent variable: Mean Individual P. paludosus biomass sample". Data points with biomass=0 were removed for this analysis and biomass values were left untransformed. Source Df SS MS F Value Pr>F Season 1 2484.54 2484.54 16.16 0.0001 Plant 1 37.20 37.20 0.24 0.6241 Error 90 13840.80 153.79 Total 92 16326.08 Table 8. Seasonal effect on grass shrimp density (number per sample) and average biomass (mg). Marginal Mean # Shrimp per Mean Biomass per Mean Sample Individual (mg) Summer 49.8 7.97 Winter 2 .2 1 8.63 27 P. paludosus reproduces in early Spring. Additional sampling in early March confirmed that extremely high densities of larval shrimp (150-500 shrimp/sample) were being recruited into the populations. Summer samples contained shrimp in their early adult life, and winter samples appeared to be composed of this same cohort near the end of their life cycle. W P. paludosus made up a much higher proportion of overall biomass than H. azteca in the two dominant macrophyte communities of the Kissimmee River riparian marsh (p=0.0001) (Table 9; Figure 10). Even when biomass values for both Species were combined, Polygonum still supported more biomass than Nuphar (p=0.0001) (Table 9; Figure 10). Comparing values for mean biomass of grass shrimp and amphipods separately for each macrophyte habitat, P. paludosus had significantly higher biomass per sample in both habitats than H. azteca did (Nuphar: p=0.0005; Polygonum: p=0.0302) (FigurelO). This is contrary to the SFWMD’s conceptual model, which predicts that H. azteca is more abundant in Nuphar beds than in Polygonum because it is a scraper and can more efficiently harvest periphyton growing on Nuphar stems (Figure 5). These results emphasize the importance of P. paludosus in this system. Considering its relative abundance and large size compared to the other common macroinvertebrates in the Kissimmee River littoral fringe habitat, it is apparent that this species overwhelmingly dominates the macroinvertebrate fauna in the Pool B remnant channel of the Kissimmee 28 Table 9. ANOVA table comparing P. paludosus and H. azteca log transformed biomass values. Dependent variable: Log(Mass+1). Source Df SS MS F Value Pr>F Plant 1 33.02 33.02 47.95 0.0001 Species 1 11.17 11.17 16.23 0.0001 Plant*Species 1 0.62 0.62 0.90 0.3439 Error 284 195.57 0.69 Total 287 240.39 400 5‘, 350 ~ I P. paludosus; +I f g 300 P DH. azteca 2 250 ~ g E 3 200 ~ E m 150 - E .2 m 100 - : III g 50 — o . l Nuphar Polygonum Figure 10. Comparison of mean P. paludosus vs H. azteca biomass per sample in the two dominant littoral fiinge macrophyte habitats of the Kissimmee River riparian marsh. 29 River, and so must be a very important food source for higher trophic levels such as large fish, waterfowl, and wading birds. DIEL MIGRA TION Results of the migration trials suggested that there were no significant diel horizontal movement patterns into or out of the floodplain of P. paludosus in the Kissimmee River riparian marsh (Table 10; Figure 11). However, in one case, there was a Significant difference in grass shrimp captured in top versus bottom traps (Table 11; Figure 12). W In Polygonum beds, shrimp capture rates were low, suggesting little shrimp movement overall. An average of 0.6 shrimp were captured moving from the channel into the floodplain, and the same number was captured moving out of the floodplain into the channel during the day. At night, these numbers differed only slightly, with an average of 0.9 shrimp moving from the channel into the floodplain and 1.0 shrimp moving in the opposite direction (Figure 11). In Nuphar beds, an average of 1.0 shrimp per trap was caught moving from the channel to the floodplain, while 3.0 shrimp on average were captured moving into channel during the day. However, these numbers are not significantly different (p=0.1147) (Table 10). At night, differences in horizontal movement were similar to the daytime trials and were not significant (p=0. 1941) (Table 10). An average of 1.7 shrimp per trap were captured moving from the channel to the 30 Table 10. Results from the Kruskal-Wallis ANOVA testing differences in grass Shrimp movement in and out of the floodplain during the day and at night. Tests for Horizontal Movement P Value Da Nuphar 0.1 147 y Polygonum 0.1683 . Nuphar 0.1941 nght Polygonum 0.5722 Table 11. Results form the Kruskal-Wallis ANOVA testing differences in grass shrimp vertical movement during the day and at night. Tests for Vertical Movement P Value Da Nuphar 0.0343 y Polygonum 0.5343 . Nuphar 0.1239 nght Polygonum 0. 1219 31 4.0 3.5 3.0 2.5 2.0 H m 1.5 .H E.“ 1.0 i- E-t r... 0.5 Q) a. a. 0.0 E h .: m g; 8.0 a E = 7.0 Z 5 q, 6.0 2 5.0 4.0 3.0 r 2.0 1.0 0.0 (a) (b) Nuphar [1.8. Nuphar 1:] CH>FP I F P>CH Polygonum 1:1 CH>FP IFP>CH l Polygonum Figure 11. Mean number t SE of P. paludosus captured moving from the channel to the floodplain (CH>FP) and from the floodplain to the channel (FP>CH) during the day (a) and at night (b). 32 4.5 ,. (a) .. E] Top 3.5 ~ I Bottom 2.5 ~ 1.5 ~ 0.5 - 0.0 Polygonum 4.5 (b) 4.0 — [3 Top I Bottom Mean Number Shrimp per Trap 3: SE 3.0 ~ 2.0 _ 1.0 ~ 0.0 Nuphar Polygonum Figure 12. Mean number i SE of P. paludosus captured in top and bottom Breder traps (a) during the day and (b) at night. Significant differences at p<0.05 denoted by double asterisks. 33 floodplain, while 4.5 shrimp per trap were captured on average moving in the opposite direction. This section of the Kissimmee River, although partially restored, was still subject to stage management regimes, and this would confound any inherent diel horizontal pattern. When water is being drawn down, shrimp will be more likely to move out of the floodplain. Conversely, when water levels are increasing, shrimp will have a greater Opportunity to use the riparian marsh habitat. These experiments should be repeated after the removal of the dam and lock structures when stage will be allowed to fluctuate more naturally. Kem'caLMoxemau An average of 0.6 shrimp were caught in both top and bottom traps in Polygonum beds during the day. At night, slightly more shrimp were captured in bottom versus top traps in Polygonum, with an average of 1.3 shrimp per trap caught moving along the bottom, and only 0.6 shrimp per trap were caught in the top traps (Figure 12). The results of the Kruskal-Wallis test indicated no significant difference in these numbers (p=0.2246) (Table 11). In Nuphar beds, an interesting pattern emerges. The difference in mean number of shrimp caught in top versus bottom traps is significant during the day (p=0.0162) (Table 11). On average, 3.5 shrimp were captured moving along the bottom, while only 0.5 were captured in top traps (Figure 12). At night, however, 3.4 shrimp were caught in bottom traps and 2.9 were caught in top traps, which were not statistically significant (p=0.1299) (Table 11; Figure 12). 34 The difference in numbers of shrimp captured in top vs. bottom traps in Nuphar beds during the day could reflect the lack of habitat complexity inherent in this habitat. Shrimp may be sticking close to the bottom during the day as a means of avoiding predation from visual predators such as wading birds and waterfowl. This is not necessary in Polygonum beds since they offer a much denser, more complex habitat. The difference in diurnal versus nocturnal vertical distribution patterns suggests that generally, more shrimp were captured during the night than during the day (Figures 11 and 12). Other grass shrimp species show similar behavior. Sogard and Able (1994) observed higher nocturnal movement in P. vulgaris, and attribute this to diel variability to predation risk. Grass shrimp are relatively large, and it would be to their advantage to restrict daytime movement in order to avoid being spotted by visual predators. However, at night, this becomes less of an issue. Grass shrimp are transparent, and any movement associated with foraging would be much less likely to result in predator attacks at night. GROWN! STUDIES Results from the growth studies show at least some growth in all food treatments (Figure 13). However, no significant net growth occurred in the fine detritus treatment (p=0.3072), with a mean increase of only 1.99 i 1.90 mg per shrimp after 150 degree days. Polygonum leaves resulted in significant net growth after 150 degree days (p=0.0012), with a mean increase of 4.03 :1: 1.18 mg per shrimp. Periphyton on Nuphar stems resulted in the highest net growth overall, with a mean increase of 6.69 :t 3.04 mg per shrimp. However, there was no significant difference in net growth between shrimp 35 aces—tomes 538m 05 vozgm $5 95.3w vote :08 88m Emma? BEE :38 05 9:59.53 .3 Barrow 8.5 9:253: coo.“ 025 05 .«o :03 E aowécm ~20 Emma? be 62 .m_ 85me ausazz Escouzom ‘ was“. cad oqr oQN o?» :91 ch cQo 095 :9» (6w) as : lllfilaM Mo 13M 36 grown on Polygonum leaves and shrimp grown on periphyton from Nuphar stems (Figure 13). From the data on distribution and abundance, it was shown that P. paludosus overwhelmingly preferred Polygonum beds, and the available food in these beds is senescent Polygonum leaves or fine detritus that collects on the bottom. There is little or no periphyton grth in these habitats due to Polygonum's high stem density and low subsurface light. P. paludosus is a benthic species that very seldomly ventures into the water column. The structure of the two macrophyte beds is such that the most available food for a benthic species living in Nuphar beds is fine detritus. Only by moving up into the water column would periphyton become available to such a species. In Polygonum beds, a benthic species would have plenty of fine detritus in addition to the steady supply of decaying Polygonum leaves. Despite the fact that grass shrimp are physically able to harvest periphyton from Nuphar stems, their preference for Polygonum beds could be a response to the food type available given the fact that they are primarily benthic and spend most of their time foraging along the bottom. GENERAL DISCUSSION P. paludosus was much more abundant in Polygonum beds according to data based on both numbers and biomass. Given these data, the question arises, “Why do grass shrimp prefer Polygonum habitat over Nuphar?” P. paludosus showed significant grth on both Polygonum leaves and periphyton associated with Nuphar stems (Figure 13). Their preference for Polygonum habitat, as mentioned above, could be a result of their benthic nature and the food type available to benthic species in each macrophyte 37 community, but overall, food availability is probably not an important factor in grass shrimp habitat selection in the Pool B remnant channel of the Kissimmee River. In the summer, dissolved oxygen sometimes dropped to near-zero levels in Polygonum beds, while the drop was not nearly as severe in Nuphar beds (Figure 7). Thus, dissolved oxygen levels are also probably not significant in driving grass shrimp habitat selection. Indeed, shrimp were more abundant in habitat with lower DO than is usually available in Nuphar (Figure 7). Thus, there must be some sort of trade-off for P. paludosus in choosing Polygonum habitat over Nuphar. Trap data suggested that P. paludosus moved considerably less in Polygonum beds than they did in Nuphar beds with no well-defined horizontal or vertical diel migration pattern (Tables 10, 11; Figures 11, 12). In Nuphar beds, while overall density was less, movement seemed to be more pronounced, with a significant diel vertical distributional pattern (Figure 12). Shrimp appeared to remain close to the bottom during the day, while they move around in the water colurrm more at night. Why do we see a vertical migration pattern in Nuphar, but not in Polygonum? One explanation is that the higher habitat complexity inherent in Polygonum as Opposed to Nuphar communities provides better refuge areas for P. paludosus. Several studies looking at the relationship between habitat complexity and macroinvertebrate habitat choice confirm this hypothesis (Crowder and Cooper 1982; Stoner and Lewis 1985). Vulnerability to predators is often inversely related to habitat complexity (Coen et a1. 1981), and this is well documented for the decapoda. Crayfish density in lakes increased with the degree of macrophyte cover, and this relationship was modified by decreased vulnerability (i.e., increased size) to predators (Stein and Magnuson 1976). 38 The distribution of some marine species of Palaemonetes has been positively correlated with increased habitat complexity (Khan et al. 1997), and habitat complexity also has been shown to reduce predatory efficiency by reducing prey capture rates (Crowder and Cooper 1982). Khan et al. (1997) suggested that characteristics of the macrophytes (physical complexity) and the shrimp (residual predator conditioning) were important factors in observed grass shrimp distributions. It seems reasonable that P. paludosus capitalizes on this increased protection from predators provided by the complex habitat of Polygonum communities. Another possible explanation for this is that the relatively low D0 in Polygonum confers an advantage to grass shrimp by reducing the number of fish predators present in these beds, thereby increasing its survival. Furse et al. (1996) found that changes in D0, particularly declines below stressful levels, were the primary influence in largemouth bass habitat use and overall movement patterns. They found that largemouth bass use both Nuphar and Polygonum macrophyte communities almost equally overall, but were more likely to be found in areas where DO>2 mg/L throughout the year. Whitrnore et al. ( 1960) showed that largemouth bass showed strong avoidance of habitats with DO levels <1.5 mg/L, while Petit (1973) reported that largemouth bass stopped feeding at D0 of 2 mg/L, and at 1 mg/L, all died within 11 hours. Each of these studies found increasing avoidance in vegetation as temperatures increased, suggesting that the high temperatures in the Kissimmee River Pool B remnant channel would contribute even more to fish stress related to low DO. As metabolic needs increase due to high temperatures, their tolerance to low DO would decrease even more. The fact that D0 is often lower in Polygonum beds and sometimes falls below 2.5 mg/L, due to the low subsurface light and 39 heterotrOphic nature of these communities (Figure 7), suggests that largemouth bass are more likely to choose Nuphar beds over Polygonum. Therefore, grass shrimp inhabiting Polygonum beds would be less likely to have contact with their fish predators than those living in Nuphar. Whether grass shrimp inherently prefer denser, more complex habitats such as Polygonum because they are provided with more refuge areas, or the uneven distribution of grass shrimp in the Pool B remnant river channel of the Kissimmee River is a result of decreased predator efficiency in Polygonum beds, it seems reasonable that the innate complexity of Polygonum communities could help explain why grass shrimp are more abundant in these areas. 40 CONCLUSIONS A goal of the Kissimmee River restoration project is to increase littoral fringe macrophyte communities such as Nuphar and Polygonum, which will increase the overall abundance of P. paludosus. Understanding the distribution of grass shrimp with respect to plant type enables predictions to be made regarding biological interactions within the Kissimmee River-floodplain ecosystem and how they will respond to restoration efforts. Since P. paludosus is a keystone invertebrate species in this system (Merritt et al. 1996) due to its relatively large size and abundance, knowledge of its distribution and abundance will help locate and quantify the potential food base for visual feeding bird predators and large game fish. This information will be useful when evaluating the success of Kissimmee River Restoration Project. The expected increase in wetland plant communities, including Polygonum and Nuphar, combined with an overall increase in dissolved oxygen, will have significant effects on biological communities. By quantifying the distribution and abundance of P. paludosus, and using information obtained fiom the growth studies, it will be possible to calculate estimates of P. paludosus production in the Kissimmee River riparian marsh. 41 LITERATURE CITED Beck, J .T. and BC. Cowell. 1976. Life history and ecology of the freshwater caridean shrimp, Palaemonetes paludosus (Gibbes). American Midland Naturalist 96: 52- 65. Breder, CM. 1960. Design for a fry trap. Zoologica 45: 155-160. Broad, AC. 1957. The relationship between diet and larval development of Palaemonetes. Biological Bulletin (Woods Hole, MA) 112: 162-170. Coen, L.D., K.J. Heck, and LG. Abele. 1981. Experiments on competition and predation among shrimps of seagrass meadows. Ecology 62: 1484-1493. Cross, R. E., R.T. Kneib, and AF. Stiven.l996 Size-dependent interactions in fish and shrimp. Twenty Fourth Annual Benthic Ecology Meeting, Columbia, SC. p. 94. Crowder, LB. and WE. Cooper. 1982. Habitat structural complexity and the interaction between bluegills and their prey. Ecology 63: 1802-1813. Cummins, K.W. and M.J. Klug. 1979. Feeding ecology of stream invertebrates. Ann. Rev. Ecol. Syst. 10: 147-172. Dobkin, S. 1963. The larval development of Palaemonetes paludosus (Gibbes 1850) (DecapodazPalaemonidae) reared in the laboratory. Crustaceana 6: 2-61. Eggleston, D.B., L.L. Etherington, and WE Elis. 1998. Organism response to habitat patchiness: species and habitat-dependent recruitment of decapod crustaceans. Journal of Experimental Marine Biology and Ecology 223: 111-132. Furse, J .B., L]. Davis, and LA. Bull. 1996. Habitat use and movements of largemouth bass associated with changes in dissolved oxygen and hydrology in Kissimmee River, Florida. Proceeding of the Annual Conference of the Southeastern Association of Fish and Wildlife Agencies 12-25. Gregg, C. and J. F leeger. 1995. Why are they called grass shrimp? Experimental tests of the effects Palaemonetes pugio has on benthic and phytal dwelling communities. Twenty-third Benthic Ecology Meeting, New Brunswick, NJ. 42 Gregg, C. and J. Fleeger.l996. Effects of the grass shrimp, Palaemonetes pugio, on sediment and stem-dwelling saltrnarsh fauna. Twenty-fourth Annual Benthic Ecolgoy Meeting, Columbia, SC. p. 43. Harris, S.C., T.H. Martin, and K.W. Curnmins. 1995. A model for aquatic invertebrate response to Kissimmee River restoration. Restoration Ecology 3: 181-194. Junk, W. J ., Bayley, P. B., and Sparks, R. E.l989. The flood pulse concept in river- floodplain systems. Proceedings of the International Large River Symposium, Ottawa, Canada. Pp. 11-27. Khan, R.N., H.C. Merchant, and RE. Knowlton. 1997. Effects of macrophytic cover on the distribution of grass shrimps, Palaemonetes pugio and P. vulgaris. Invertebrate Biology 116: 243-247. Kneib, RT. 1985. Predation and disturbance by grass shrimp, Palaemonetes pugio Holthuis, in soft-substratum benthic invertebrate assemblages. Journal of Experimental Marine Biology and Ecology 93: 91-102. Koebel, J .W. 1995. An Historical Perspective on the Kissimmee River Restoration Project. Restoration Ecology 3: 149-159. Kushlan, J .A. and MS. Kushlan. 1980. Population Fluctuations of the Prawn, Palaemonetes paludosus, in the Everglades. American Midland Naturalist 103: 401-403. Meehean, CL. 1936. Notes on the freshwater shrimp Palaemonetes paludosa (Gibbes). Trans. Am. Microsc. Soc. 55: 433-441. Merritt, R.W., M.J. Higgins, K.W. Cummins, and B. Vandeneeden, 1999. Seasonal Differences in Invertebrate Functional Feeding Group Relationships of the Kissimmee River-floodplain Ecosystem, Florida. In: D. Batzer, R.B. Rader, and SA. Wissinger (eds), Invertebrates in Freshwater Wetlands of North America, John Wiley & Sons, Inc., pp. 55-79. Merritt, R.W., J .R. Wallace, M.J. Higgins, MK. Alexander, M.B. Berg, W.T. Morgan, K.W. Cummins, and B. Vandeneeden. 1996. Procedures for the functional analysis of invertebrate communities of the Kissimmee River-floodplain ecosystem. Florida Scientist 59: 216-274. 43 Miller, D.C., S.L. Poucher, L. Coiro, S. Rego, and W. Munns. 1995. Effects of hypoxia on growth and survival of crustaceans and fishes of Long Island Sound. Proceedings of the Long Island Brook, NY (U. S. EPA) p. 92. Payne, A.J., 1986. The ecology of tropical lakes and rivers, John Wiley and Sons, New York. Petit, GD. 1973. Effects of dissolved oxygen on survival and behavior of selected fishes of western Lake Erie. Ohio Biological Survey Bulletin 4: 1-76. Posey, M.H. and AH. Hines. 1991. Complex predator-prey interactions within an estuarine benthic community. Ecology 72: 2155-2169. Pringle, C.M., G.A. Blake, A.P. Covich, K.M. Buzby, and A. Finley. 1993. Effects of omnivorous shrimp in a montane tropical stream: sediment removal, disturbance of sessile invertebrates and enhancement of understory algal biomass. Oecologia 93: 1-11. Samuels, A.J. and CF. Mason. 1998. Shrimps and eutrophication in the waterbodies of a coastal grazing marsh. Hydrobiologia 377: 195-199. SAS Institute. 1996. SAS user’s guide: statistics, version 6.12. SAS Institute, Cary, North Carolina, USA. Shen, H.W., G. Tabios, and J .A. Harder. 1994. Kissimmee River Restoration Study. Journal of Water Resources Planning and Management 120: 330-349. Sogard, SM. and K.W. Able. 1994. Diel variation in immigration of fishes and decapod crustaceans to artificial seagrass habitat. Estuaries 17: 622-630. Stein, RA. and J .J . Magnuson, 1976. Behavioral response of crayfish to a fish predator. Ecology 57: 751-761. Stoner, AW. and PI. Lewis. 1985. The influence of quantitative and qualitative aspects of habitat complexity in tropical sea-grass meadows. Journal of Experimental Marine Biology and Ecology 94: 19-40. Strenth, NE. 1976. A Review of the Systematics and Zoogeography of the Freshwater Species of Palaemonetes Heller of North America (Crustacea: Decapoda). Smithsonian Contributions to Zoology 228: 1-27. Toland, BR. 1990. Effects of the Kissirmnee River Pool B Restoration Demonstration Project on Ciconiiformes and Aseriforrnes. Proceedings of the Kissimmee River Restoration Symposium pp. 83-91. Toth, LA. 1990. Impacts of channelization on the Kissimmee River ecosystem. Proceedings of the Kissimmee River Restoration Symposium pp. 47-56. Toth, LA. 1993. The ecolgical basis for the Kissimmee River restoration plan. Florida Scientist 56: 25-51. Toth, L.A., D.A. Arrington, M.A. Brady, and DA. Muszick. 1995. Conceptual evaluation of factors potentially affecting restoration of habitat structure within the channelized Kissimmee River ecosystem. Restoration Ecology 3: 160-180. Trexler, J. C. 1995. Restoration of the Kissimmee River: A conceptual model of past and present fish communities and its consequences for evaluating restoration success. Restoration Ecology 3: 195-210. Vemberg, F .J . and S. Piyatiratitivorakul. 1998. Effects of salinity and temperature on the bioenergetics of adult stages of the grass shrimp (Palaemonetes pugio Holthuis) from the North Inlet Estuary, South Carolina. Estuaries 21: 176-193. Weller, M.W. 1995. Use of two waterbird guilds as evaluation tools for the Kissimmee River restoration. Restoration Ecology 3: 211-224. Welsh, BL. 1975. The role of grass shrimp, Palaemonetes pugio, in a tidal marsh ecosystem. Ecology 56: 513-530. Whitrnore, C.M., C.E. Warren, and P. Doudoroff. 1960. Avoidance reactions of salrnonid and centrarchid fishes to low oxygen concentrations. Transactions of the American Fisheries Society 106: 323-330. Wullschleger, J .G., S.J. Miller, and L]. Davis. 1990. an evaluation of the effects of the restoration demonstration project on Kissimmee River fishes. Proceedings of the Kissimmee River Restoration Symposium pp. 67-81. 45 "illllllllllllllll