Ecological Succession And Fragmentation In A Reservoir .

2y ago
34 Views
2 Downloads
1.96 MB
22 Pages
Last View : 27d ago
Last Download : 3m ago
Upload by : Victor Nelms
Transcription

American Fisheries Society Symposium 62:147–167, 2008 2008 by the American Fisheries SocietyEcological Succession and Fragmentation in a Reservoir:Effects of Sedimentation on Habitats and Fish CommunitiesTim Patton* and Cris LydayDepartment of Biological Sciences, Southeastern Oklahoma State UniversityDurant, Oklahoma 74701, USAAbstract.—While processes of depositional filling and ecological succession innatural lakes have been well described, these concepts are relatively new and seldomapplied to reservoirs, especially at the landscape scale. However, ecological timehas been sufficient to allow us to see successional processes in many reservoir systems. Illustrative of such processes, Lake Texoma is a 36,000-ha reservoir locatedin southern Oklahoma and northern Texas, and patterns of depositional filling andsubsequent processes are apparent in the up-lake ends (there are two large-rivertributaries) of this system. Completed in 1944, Lake Texoma has a drainage areaof more than100,000 km2, most of which is highly erodable agricultural lands. Weused historic aerial photographs, geographic information systems technology, andfield measurements to examine a variety of surface and habitat features and analyzed experimental gill-net samples using ordination techniques to characterize thefish communities in portions of the reservoir most affected by sedimentation. Extensive sedimentation and accretion of sediments above water level has effectivelyresulted in surface area reduction, cove isolation, fragmentation of lacustrine habitats, morphometric changes, and establishment of terrestrial vegetation on newlydeposited lands. Most notably, sedimentation has led to the development of linearbars of deposition above normal pool elevation that have blocked mouths of coves,bisected large areas of the reservoir, and fragmented several pools. In our study sitealone, 332 ha (surface area) of reservoir has experienced accretion of land above thewater level. Reservoir fragments had lower shoreline development values (mean 2.21) than comparable control sites (mean 3.39). Depositional shorelines associated with sedimentation exhibited lower gradients than nondepositional shorelines(mean 2.0% versus 4.2%, respectively), and habitat heterogeneity was lower alongdepositional shorelines than along nondepositional shorelines. Fish communitiesin isolated reservoir fragments appeared to be distinct from fish communities innonfragmented habitats. This change in community structure may be driven by anappreciable reduction of pelagic species from fragmented sites, as these sites havelimited or no connectivity to the main body of the reservoir. With respect to thenewly deposited lands, ecological succession of vegetation followed a progressionfrom mud flats to dense, nearly monotypic stands of black willow Salix nigra forests within a few years. These habitat changes had strong implications to the fishcommunities as well as to adjacent terrestrial wildlife communities and will likelypose many challenges, and perhaps opportunities, for natural resource managers.* Corresponding author: tpatton@sosu.edu147

148patton and lydayIntroductionNumerous processes lead to the formationof naturally occurring lakes in North America (Hutchinson 1957; Wetzel 2001) but, atnorthern latitudes, most exist as a resultof glacial activities during the most recentPleistocene event (e.g., glacial moraine lakes)and, at southern latitudes, most exist as a result of fluvial processes (e.g., oxbow lakes).Accordingly, these lakes are generally hundreds to thousands of years old. Over suchecological and geological time spans, numerous ontogenic changes occur, including biological, chemical, and physical effects. Ontogenic changes have been described in trophicstate (Greeson 1969; Carpenter 1981, Whiteside 1983; Wetzel 2001), food webs and foodchain lengths (Kaunzinger and Morin 1998)and various physicochemical parameters(Wetzel 2001). With sufficient time, allochthonous input, and autochthonous production, many lakes experience sedimentationand filling. As bed accretion occurs, lakesmay make the transition from a lentic systemto more swamp-like or marsh-like habitatsand, with sufficient time, to hydric and possibly mesic terrestrial habitats (Whittaker1975; Klinger 1996; Wetzel 2001).Compared to naturally occurring lakes,man-made reservoirs are a relatively newfeature on the landscape. Large man-madeimpoundments only became commonplacein the United States within about the pastcentury, and the number of dams worldwideis estimated to exceed 40,000 (Nilsson andBerggren 2000). Nevertheless, as with naturally occurring lakes, reservoirs also experience accretion of substrates due to sedimentation. While appreciable accretion may notbe occurring reservoir-wide in most cases, itis likely occurring in at least the headwaterregions of most reservoirs, as is apparent insuch processes as delta formation at reservoir inlets (James and Barko 1990; Williams1991) and isolation of backwaters that wereformerly connected (Slipke and Maceina2005; Slipke et al. 2005). However, in manysystems, total reservoir volume has been reduced appreciably (Julien 1998), and this process has been reported for reservoir systemsworldwide (Tejwani 1984; Annandale 1987;White 1988; Fan and Morris 1992; Hassanzadeh 1995; Tundisi et al. 1998). Sedimentation rates have exceeded those originallypredicted in several reported instances; ina study of 21 reservoirs in India, Tejwani(1984) noted that annual sedimentation rateswere 40–2,166% greater than was assumedat the time of project design. Einsele andHinderer (1997) reported that human activities have increased sediment yield by a factor of 2 to more than 100 among lakes theystudied in Europe. In a study of several Australian reservoirs, Chanson and James (1998,1999) reported that numerous dams becamefully silted, though it should be noted thatthe reservoirs they studied were relativelysmall. In the United States, Thornton et al.(1990) stated that among reservoirs constructed prior to 1953 in the Midwest, GreatPlains, and southeastern and southwesternstates, 33% have lost from one-half to threequarters of their original volume, and about10% have had all usable storage depleted bysediment deposition.A review of the scientific literature reveals that the majority of the research associated with reservoir sedimentation lies within the fields of hydrology, water resources,civil engineering, and geology, as indicatedby the journals in which the informationis published. Less common are studies thataddress the ecological effects of reservoirsedimentation. Long-term patterns of ecological succession in natural lakes has beendescribed with respect to several parameters,and the field of paleolimnology has revealedmuch about historic conditions, processes,and communities (Smol 1992; Wetzel 2001;

ecological succession and fragmentation in a reservoirCohen 2003). Ecological effects of reservoirsuccession may be similar to those of natural lake succession, but research is needed toascertain this assertion.It is clear that sedimentation is havinga profound impact on reservoirs worldwide,but our knowledge of the landscape-scale effects of sedimentation on reservoirs is primarily restricted to the mechanics of deposition. There is a paucity of knowledge andinformation on ecological effects of this process. In this paper, we address issues relatedto sedimentation in Lake Texoma, a largereservoir on the eastern edge of the GreatPlains. Our objectives were to describe (1)broad patterns of sediment deposition, (2)effects of deposition on selected littoralhabitat characteristics, and (3) effects ofsedimentation on the large-bodied fish community structure.Study AreaLake Texoma is a 36,000-ha reservoir onthe Oklahoma–Texas border, impoundingthe Red and Washita rivers (Figure 1). Thereservoir was formed by completion of theDenison Dam in 1944, and the reservoirwas full by the late 1940s. The drainage areaof the reservoir is 101,362 km2, including81,999 km2 in the Red River drainage and19,363 km2 in the Washita River drainage(Matthews et al. 2004; Matthews and MarshMatthews 2007). The drainage area includesmuch of southwestern Oklahoma and northcentral Texas, as well as much of the Texaspanhandle and parts of eastern New Mexico.There is a strong aridity gradient within thedrainage, with precipitation ranging from ca45–94 cm in the east–west gradient. Vegetation patterns reflect the precipitation gradient, with aridity-tolerant short-grass andshrub ecosystems to the west and tall-grassprairie and cross timber ecosystems to theeast. The vast majority of the drainage area149is used for agricultural purposes, includingprimarily livestock grazing and hay androw-crop production.The drainages of the Washita and Redrivers have a long history of extensive erosion, resulting in high turbidity and heavysilt loads in these rivers (Matthews 1988;Matthews et al. 2005). The upstream ends ofboth the Washita and Red River arms of LakeTexoma are visibly experiencing substantialsedimentation. Sedimentation is apparentfrom aerial photographs but is also easily detected while boating in the area. Many localfisherman, residents, marina operators, andbiologists have provided anecdotal evidenceof extensive sedimentation, and it is apparent that portions of the reservoir are rapidlyfilling and becoming impassable by boatsduring low-water periods. More notably, it isalso apparent from aerial photographs andground visitation that sedimentation hasoccurred to the point of accretion of newlands above the water level and that suchdepositions are occurring in a pattern thatresulted in fragmentation of large areas ofwater from the main body of the reservoir.In much of the area, accretion of these newlands above the normal water level had takenplace over a long enough time span that theyhad become forested by the onset of our research in 2004 (this paper).While extensive deposition is visible inboth of the two major arms of the reservoir,we selected a study area within the WashitaRiver arm, as this area appeared to be experiencing a greater extent of sedimentationthan the Red River arm. Prior to construction of the reservoir, the Washita River channel was dammed via the construction of twodikes, and the river was rerouted by the U.S.Army Corps of Engineers to avoid floodingan existing oil field. The old river channelwas effectively contained between the twodikes, and the new river channel bypassed theoil field and became the inlet to the reservoir.

150patton and lydaya.NMOKb.TXDamFigure 1. Lake Texoma, Oklahoma and Texas, including the 101,362 km2 drainage area (inset a)in Oklahoma (OK), Texas, (TX), and New Mexico (NM), and the portion of the reservoir (insetb) for which we conducted geographic information systems analyses, habitat assessment, and fishsampling.Our study site extended from the inlet of thenew river channel to a distance approximately10 km downstream (Figure 1). We selectedthis area because it appeared to contain themajority of the sedimentation and accretionof new lands above the normal water level.MethodsSpatial AnalysesTo assess patterns of sediment deposition,we obtained a time series of aerial photographs of the study site. We used photo-

ecological succession and fragmentation in a reservoirgraphs from 1969, 1983, 1991, and 2003because these are the years for which suchimages were available. Information on exactdates, flight altitudes, and water levels at thetimes that photographs were taken was notavailable. Photographs were scanned and imported for spatial analyses using geographicinformation systems (GIS); all GIS analyseswere completed using ArcView (ESRI 2001).All images were digitized to the 1-m pixellevel, georeferenced, and orthorectified. Oneach of the four time-series photographs,we conducted queries for total area of waterand total area of land that was previouslywater. Four relatively large ( 100-ha) areaswithin the study site had become fragmentedby sediment deposition. From the 2003 image, we determined surface area, total shoreline length, and proportion of depositionalshoreline for each reservoir fragment (fragmented due to sediment deposition). Surfacearea and shoreline length values were usedto calculate the shoreline development indexof reservoir fragments. To serve as a basis ofcomparison, we conducted spatial analyseson two control sites; these sites were largecoves that were nearby and of approximately the same surface area as the four reservoirfragments. We queried the two control sitesfor total area of water and shoreline lengthand calculated their shoreline developmentindex. Because control sites were not fragmented from the main body of the reservoir,we drew a straight line across the mouth ofeach cove to make them polygons suitable forqueries.HabitatWe examined three littoral habitat attributesas they relate to sediment deposition: shoreline development, proportion of shorelinesthat are depositional, and bank slope valuesin depositional versus erosional shorelines.The shoreline development index providesan indication of shoreline irregularity and151has implications towards littoral versus pelagic habitats (Wetzel 2001) and was calculated with values determined using GISanalyses. Shoreline development values weredetermined for each of four reservoir fragments that had formed by 2003 and for eachof the two control sites.Bank slope measurements were donein May 2006. To examine the influence ofdeposition on bank slopes, we used GIS imagery to delineate depositional and nondepositional, or erosional, shoreline and compared the two. Depositional shorelines weredefined as shorelines that exist as a result ofsediment deposition and included those areasthat were represented by new lands (landsand shorelines that were not present in the1969 image). Erosional shorelines were defined as those that had not changed, or thathad possibly receded, since 1969. We choseto call them erosional shorelines because virtually all of them showed patterns of erosion that are typical of reservoirs in whichwater levels fluctuate widely: banks lackedvegetation, active erosion was evident, andsubstrate particle sizes were large. Withineach of three reservoir fragments that weidentified via the spatial analyses, we selectedfive locations along depositional shorelinesand five locations along erosional shorelines.The fourth reservoir fragment site was notexamined due to difficulties with boat accessto the area. Sites were selected using a systematic approach; once depositional and erosional shorelines were delineated and measured using GIS, the total length of eachcategory within each of the three fragmentswas divided by 5. The first site was selectedrandomly between zero and the total length(m) of the shoreline within respective categories. For example, if there were 860 mof depositional shoreline, 860/5 172; thefirst site was randomly selected between 0and 172, and each subsequent site was at172-m intervals. We used these locations as

152patton and lydaybeginning points for transects. Transectsbegan at water’s edge and extended 110 minto the body of water perpendicular to theshoreline, and we measured depth, to thenearest cm, at 10-m intervals (with the firstpoint being 0 cm depth). Depth was measured using a depth rod where water wasshallow enough to wade and using a depthfinder mounted on a boat where water wasexcessively deep to wade. The depth finderwas checked for accuracy at the beginningand end of each sampling period over a widerange of depths using a sounding weight attached to a line; the depth finder was alwayswithin 2% of the measured depths. Withineach of three reservoir fragments, we compared the mean slope of depositional shoreline measurements to the mean slope of theerosional shoreline measurements. All slopecomparisons were made from shoreline to 50m and again from shoreline to 110 m. Finally,we compared the mean slope of all 15 depositional shoreline measurements to the meanslope of all 15 erosional shoreline measurements. All slope comparisons were done using single-tailed paired t-tests (Zar 1998),and we selected an alpha value of 0.05.Fish Community CompositionWe used experimental gill nets to assess thecommunity structure of large-bodied fishes(large enough to be captured in a 1.27-cmmesh, or greater, experimental gill net). Gillnets were 61 3 1.83 m and had a lead-corebottom line and foam-core float line; each nethad eight panels, with mesh sizes rangingfrom 1.27 to 10.16 cm, at 1.27 cm-increments,among the eight panels. Three nets were setin each of the three reservoir fragments(as identified via GIS analyses), and in eachof three control sites, for 3 nights (9 netnights/site; total 54 net nights). Controlsites were selected as a basis for comparisonof community structure, and all three control sites were attached to the main body ofthe reservoir. The control sites consisted oftwo large coves (the same two coves used ascontrol sites for GIS analyses) and one mainreservoir pelagic site adjacent to the reservoirfragment sites. Each net-night included onelittoral set, one surface pelagic set, and onebenthic pelagic set. All gill netting was donein fall (September–November) 2005. A reservoir fragment site and a control site weresampled at adjacent times before sampling additional sites to ameliorate possible seasonaleffects. Thus, if a seasonal effect occurred, weassumed that it would affect fragment sitesand control sites approximately equally. Gillnets were checked after approximately each24 h; all fish were removed, identified, enumerated, weighed, and measured, and any livefishes were released away from nets.To address community structure, weconstructed relative abundance tables, determined species richness, and calculatedShannon diversity (Pielou 1975). We usedtwo-way analysis of variance (Zar 1998) tocompare mean fish abundance (log10-transformed) among fragment and control sites.We used a one-tailed t-test (ZAR 1998) tolook for differences in mean species richnessand Shannon diversity between fragmentand control sites. Last, we conducted correspondence analyses (CA), a multivariateordination approach used to reveal similarities between communities based on speciescomposition, and graphically display theextent of similarities on multidimensionalspace (Everitt and Dunn 2001). All CA analyses were conducted using NTSys software(Applied Biostatistics, Inc. 2001) and wereconducted for sites and for species; site CAsindicate similarities among sites and speciesCAs provide insight on the species that havethe greatest influence on the site scores. Forall CAs, we log-transformed the abundancedata and omitted extremely rare species(those occurring as 1% of the total catch).This prevented rare species from driving the

ecological succession and fragmentation in a reservoirordination, allowing us to base comparisonson the more abundant species comprisingeach community.ResultsSpatial AnalysesTotal surface area of water in the designatedstudy area was 3,710 ha (Figure 1, inset b.).Within this area, sedimentation resulted inextensive deposition and accretion of newlands above the surface of the water, and isolation or partial isolation of several reservoirfragments, since 1969 (Figure 2, Figure 3a).Total surface area of water within the areafor which we were able to obtain the time series of photographs (1969, 1983, 1991, and2003) was 810 ha, and ontogenic sedimentation patterns were apparent between eachtime period (Figure 3b.). From 1969 to 2003,total surface area of water in the entire studysite was reduced by 332 ha, with this sameamount of area now represented by depositional lands that had accreted above the surface of the water (Figures 2 and 3; Table 1).Loss of surface area of water and increase innew depositional lands was exponential overthis time period (Figure 4; Table 1.)HabitatBy 2003, four relatively large fragments(mean 234 ha; range

to sedimentation in Lake Texoma, a large reservoir on the eastern edge of the Great Plains. Our objectives were to describe (1) broad patterns of sediment deposition, (2) effects of deposition on selected littoral habitat characteristics, and (3) effects of sedimentation on the large-bodied fish

Related Documents:

3. Topic: Succession a. Activity: Succession Notes, Succession Packet b. Textbook: Section 4.3 c. Practice: i. Complete the Venn Diagram comparing Primary and Secondary Succession Primary Succession Secondary Succession which is _ (longer/shorter) than _ succession. Look at Year 1-3, what letter does the graph look like?

What is primary succession? 3. When a disturbance changes a community without removing the soil, what type of succession follows? 4. Describe the process of succession in an ecosystem. 5. Why does secondary succession typically proceed faster than primary succession? 6. Use the Venn diagram to compare the two types of ecological succession. Primary

What Is Succession Planning, and Why Is It Needed? Succession planning, sometimes called succession management to emphasize the active and continuous nature of the effort, is a process for preparing people to meet an organization's needs for talent over time. As I defined it in my book Effective succession planning (2005, p. 10), succession .

run. It is an ideal way of planning for succession across several levels in an organization: Leadership succession planning: Non-leadership succession planning: Connect with an expert Types of succession planning Each business structure will require a different type of succession plan. For example, sole proprietorships and partnerships, the

4.3.1 Age and the Ecological Footprint 53 4.3.2 Gender and the Ecological Footprint 53 4.3.3 Travelling Unit and the Ecological Footprint 54 4.3.4 Country of Origin and Ecological Footprint 54 4.3.5 Occupation, Education, Income and the EF 55 4.3.6 Length of Stay and Ecological Footprint 55 4.4 Themes of Ecological Resource Use 56

succession plans to ensure proper staffing for this increasingly important communications role. Succession Planning Toolkit . A succession plan is a process for keeping t alent in the pipeline. It requires continual assessment of current personnel and future agency needs. Creating a succession plan for public safety communications

Emergency Succession A sample template to provide guidance on writing an . Emergency Succession Plan Emergency Succession Planning Template . Other . Bonner Consulting's Blog , contains several recent articles on Succession Planning and Leadership Development Daring To lead 2011

Succession – may take place in one of 3 ways: 1. Testamentary succession - in accordance with a will which the testator regulates the succession. 2. Intestate succession – through the operation of the law of intestate succession, where the deceased did not leave a will – estate is intes