USGS Home
Contact USGS
Search USGS

U.S. GEOLOGICAL SURVEY
Scientific Investigations Report 2009–5025

Back to Table of Contents

Introduction

On June 7, 2006, Caledonia Marsh, a previously drained wetland, was flooded unintentionally due to a break in the levee on its western border with Upper Klamath Lake, Oregon (fig. 1). Caledonia Marsh was separated from Upper Klamath Lake in 1916 by the construction of a levee around the northern and western boundary of the marsh. Until June 7, Caledonia Marsh was used primarily for agriculture (Snyder and Morace, 1997). In the southeastern corner of Caledonia Marsh, two parcels of land were restored to wetlands in 1997 and 2000 as mitigation for impacts to wetlands from resort development. The name Caledonia Marsh will be referred to as “flooded Caledonia Marsh” for the remainder of this report, where appropriate.

Aerial photographs taken 2 days after the breach reveal the extent of flooding in the marsh (fig. 2). A sediment plume extending out of the marsh and into the lake indicates sediment disturbance at or near the site of the breach (fig. 3). Waters near the inside of the external levee were turbid, indicating sediment disturbance and suspension into the water column. Field crews measured a depth of 8 m at the site of the breach, which indicates deep scouring when floodwaters flowed into the marsh.

Hydrologic Description of Upper Klamath and Agency Lakes

Upper Klamath and Agency Lakes are located in south-central Oregon, in a graben structural valley; the lake basin has a drainage area of 9,842 km². The climate in this valley is semiarid due to the rain-shadow effect of the flanking mountains, which reach an elevation of 2,743 m, and the warm easterly currents moving off the Pacific Ocean from the west (Johnson, 1985). Upper Klamath Lake is a large, shallow lake with a surface area of 232 km². The average and maximum lake depths are 2.8 and 15 m, respectively. Agency Lake, which has a surface area of 38 km², is connected to the northern part of Upper Klamath Lake through Agency Straits. Inflows to the lakes include the Wood River to Agency Lake, the Williamson River to Upper Klamath Lake, and spring sources at various locations, specifically from Pelican Bay and from the eastern shore near Modoc Rim (fig. 1). Upper Klamath Lake outflows to the Link River, which was dammed in 1921. The Link River Diversion Dam controls the lake to a maximum elevation of 1,263.4 m and provides water for the Bureau of Reclamation’s Klamath irrigation project by diverting flow from the Link River into the “A” canal.

Upper Klamath Lake is home to two species of endangered fish, the Lost River sucker (Deltistes luxatus) and the shortnose sucker (Chasmistes brevirostris). Historically, Upper Klamath Lake was eutrophic, but an increase in monocultural blooms of Aphanizomenon flos-aquae, a blue-green alga (Phinney and Peek, 1960; Kann, 1998), has been linked to changes in the basin. Human activities, such as the severance and drainage of wetlands adjacent to the lake and water-control and diversion practices (Bradbury and others, 2004; Eilers and others, 2004), resulted in hypereutrophic conditions that contributed directly and indirectly to the decline of endangered suckers (Perkins and others, 2000; Banish and others, 2007).

Wetlands Around Upper Klamath and Agency Lakes

Wetlands have been identified as key to nutrient cycling dynamics, in addition to providing food and habitat for birds and fish. They can act as nutrient sinks before water is released downstream to rivers, lakes, and oceans (Carter, 1996; Novitzki and others, 1996). Constructing wetlands, in particular swamps and marshes, may be a viable way to reduce nitrogen and phosphorus loads to a hydrologic system (Reuter and others, 1992; Guardo and others, 1995), although the efficiency of nutrient reduction is debatable depending on the hydrodynamics of the system and wetland type (Bullock and Acreman, 2003; Fisher and Acreman, 2004; McCormick and Campbell, 2007). Wetlands that are drained tend to undergo accelerated decomposition and land subsidence (Kentula, 1996). The exposed soils are subject to aerobic microbial activity accelerating the decomposition of organic matter and mobilizing nutrients previously held by peat soils. This decomposition is greater at higher temperatures and in agricultural lands that undergo tillage, which breaks the soil surface and increases air and oxygenated water penetration into the soils (Maciak, 1972). Subsidence occurs as pore water leaves the pore spaces and allows air to replace water in the pores. Because the compressibility of air is greater than the compressibility of water, sediments become compacted by the overlying weight and subsidence occurs. Decompositional processes that oxidize organic materials in these soils also result in subsidence (Stephens, 1956; Mulqueen, 1986).

The draining of wetlands around Upper Klamath Lake likely had an affect on nutrient dynamics in the system (Miller and Tash, 1967; Bortleson and Fretwell, 1993; Snyder and Morace, 1997; Rykbost and Charlton, 2001). Historically, 162 km² of wetland habitat surrounded Upper Klamath and Agency Lakes (fig. 4). Between 1889 and 1971, 57 percent of those wetlands were drained for cattle grazing or crop cultivation (Snyder and Morace, 1997). One approach being implemented to restore the functional ecology of the lake is to breach the levees separating these drained wetlands from the lake. By 2001, the 12 km² Wood River property was restored to wetland habitat, thereby increasing the area of undrained wetland to 49 percent of the historical total. In 1998, 29 km² at Agency Lake Ranch was converted from a previously drained wetland for cattle grazing to a water-storage area, where water is impounded between October and July and pumped to low water levels from July to October depending on the climate in a particular year (Damien Ciotti, Bureau of Reclamation, oral commun., 2007). Agency Lake Ranch was expanded in 2005 to include an additional 11 km² at Barnes Ranch. Although Agency Lake Ranch and Barnes Ranch were not restored to a typical vegetated wetland habitat, the area that was flooded is more likely to attract wetland plants and animals as water storage than as a cattle ranch.

The Williamson River Delta area, which is owned by The Nature Conservancy, is another large wetland restoration project (fig. 4). On October 30, 2007, about 19 km² of previously reclaimed wetland west of Williamson River was reconnected to Upper Klamath and Agency Lakes by breaching several levees at the northern end of Upper Klamath Lake and the eastern shore of Agency Lake. In the future, about 11 km² on the eastern side of the Williamson River will be flooded to complete the restoration of the Williamson River Delta to historical conditions. In the long term, the reconnection of these wetlands is expected to provide benefit in terms of restored habitat for larval and juvenile suckers, as well as the eventual restored function of nutrient sequestration. Reflooding previously drained wetland areas also should slow the process of nutrient release that occurs when the peat soils exposed to the air by draining are oxidized (Snyder and Morace, 1997). In the short term, however, reflooding these reclaimed wetlands that have been in agricultural use may act as a source of nutrients to the lake (Aldous and others, 2005).

The unexpected flooding of Caledonia Marsh provided an opportunity to collect data that will help to quantify the load of nutrients that might be expected when other reclaimed wetlands are flooded. The nutrient loads from Caledonia Marsh are particularly relevant to the reflooding of previously drained wetlands on which the land use was similar to that on the flooded Caledonia Marsh—largely grain and vegetable crop cultivation. Other reclaimed wetlands surrounding UKL that were used for grain and vegetable crops include the recently reflooded Williamson River Delta. The reflooded wetlands of the Wood River, Barnes Ranch, and Agency Lake Ranch were historically used for cattle grazing, and the nutrient loads from Caledonia Marsh may be less comparable to the nutrients loads from those wetlands.

Purpose and Scope

The purpose of this report is to characterize and compare nutrient and chlorophyll a concentrations, and water quality parameters in the flooded Caledonia Marsh for several months immediately after flooding. Results of the water quality sampling among the three sites in the marsh and between the marsh and the lake are compared. The results of this study will help to understand the water quality dynamics of flooded, previously drained wetlands and to facilitate the restoration of wetlands.

Flooded Caledonia Marsh and Upper Klamath Lake Water Quality Sampling Sites

Sampling at three sites in flooded Caledonia Marsh (fig. 1) started on June 21, 2006, 2 weeks after the breach, and continued until September 11, 2006, when water levels dropped too low to allow boat access to sites. Water column depths were 2.5, 2.1, and 2.7 m at sites FCM1, FCM2, and FCM3, respectively at the start of sampling, and were lower by approximately 1 m by September 11, 2006. Site locations were chosen to cover an even spatial distribution of the flooded Caledonia Marsh area, and to provide samples from within each of the submerged internal levees. Site FCM2, farthest to the east, was near an internal levee covered by apparent noncultivated vegetation on the periphery of cropland, while sites FCM1 and FCM3 were on cropland with no apparent vegetation growth at that time. All three sites were on or near areas farmed for grain (Jon Barkee and Jeff Woolworth, Running Y Ranch., Klamath Falls, oral commun., 2007), although it is not likely that there was much growth at the time of the levee breach. Sites FCM1 and FCM3 were closest to the breach, at a distance of about 1.3 and 0.9 km, respectively. Site FCM1 was separated from the breach by one submerged internal levee. Site FCM2 was farthest from the breach, at a distance of about 2.9 km, and was separated from the breach by three submerged internal levees (fig. 5). Site FCM2 also was closest to an area of restored wetlands in the southeastern corner of the marsh.

Through a long-term monitoring program, water samples were collected concurrently at six sites (MDN, WMR, RPT, HDB, MDT, and EPT) in Upper Klamath Lake (fig. 1). Although these sites are referred to herein, the data from the lakewide sites are discussed in Lindenberg and others (2009).

Back to Table of Contents