Persistence of Daphnia in the epilimnion of Lake Mead, Arizona–Nevada, during extreme drought and expansion of invasive quagga mussels (2000–2009)

Abstract

Lake Mead has experienced drought conditions since 2000 and the introduction of invasive quagga mussels since about 2004. Empirical evidence from the Great Lakes region suggests a potential for energy transfer to the benthos and away from planktivorous fish through competition between quagga mussels and filter-feeding, herbivorous zooplankton. We describe Daphnia dynamics from 2043 analyses performed on the 3 basins of subtropical Lake Mead during 2000–2009. Daphnia were generally most numerous in shallower areas strongly influenced by nutrient inputs and with decreased clarity and higher chlorophyll a (chl-a) concentrations. Daphnia populations increased in mid- to late winter and early spring prior to the spawning of gamefish populations and rapidly decreased with increasing predation pressure, water temperature and cyanobacteria populations. Concurrent with the reduction in Daphnia populations in late spring, quagga veliger abundance rapidly increased, declined in summer, and peaked again in the fall. Quagga veligers in Lake Mead were generally most numerous in open water areas with maximum depth, greater clarity and low chl-a. Although mean Daphnia abundance since the beginning of drought conditions and the establishment of quagga mussels is lower in Boulder Basin, the temporal and spatial dynamics of the zooplankton community remain comparable to previously described historic patterns in all 3 basins of Lake Mead.

Key words:

• Daphnia
• drought
• invasive species
• Lake Mead
• quagga mussels
• southwestern reservoirs

Lake Mead is experiencing conditions that typically stress aquatic ecosystems (drought and introduction of dreissenid mussels). Reservoir levels began to drop dramatically in 2000 in this subtropical reservoir and are now near a 40-year low. Little is known about the impacts of climatic changes and resulting periods of drought to top-down, bottom-up, and abiotic controls that structure qualitative and quantitative characteristics of plankton communities. Studies describing the zooplankton response to drought are particularly limited (McGowan et al. 2005, Havens et al. 2007). Quagga mussels (Dreissena rostriformis bugensis) became established in Boulder Basin subsequent to drought conditions. Quagga mussels were first confirmed in January 2007, but likely colonized around 2004 (McMahon 2007). Since becoming established, quagga mussels have expanded throughout all basins of Lake Mead. Given their ability to attain high biomass, combined with intrinsically high clearance rates, dreissenid mussels can modify established trophic linkages by grazing phytoplankton and altering patterns in nutrient recycling (Vanderploeg et al. 2002, 2009, Wilson and Sarnelle 2002). Extensive research on the ecological impacts of invasive dreissenid mussels exists for north temperate lakes (i.e., Great Lakes region, Europe), but the ultimate ecological impact of invasive dreissenid mussels on subtropical reservoirs with differing thermal and hydrological regimes is unknown.

Our objective was to examine a long-term database (2000–2009) for evidence that Daphnia spp. have been significantly affected by drought and subsequent invasion of quagga mussels. Previous studies of the zooplankton community of Lake Mead have been limited in scope and duration (Burke 1977, Paulson and Baker 1983, Wilde 1984, Axler et al. 1988, Wilde and Paulson 1988, Sollberger and Paulson 1991, Loomis 2009).

Methods

Sample locations

Lake Mead was at near full capacity at the beginning of the study, but water elevation progressively dropped over the subsequent 10 years. The end-of-month elevation and content as of January 2000 was 1214.26 ft, 25.0 million acre-feet (97% capacity). The end-of-month elevation and content as of December 2009 was 1096.30 ft, 11.2 million acre-feet (43% capacity). Major surface inflows to Lake Mead include the Colorado River (∼97%), the Virgin and Muddy rivers, and the Las Vegas Wash (Holdren et al. 2006). Boulder Basin is the most downstream portion of the reservoir and collects the combined flows of Overton Arm and the Colorado River and receives nonpoint surface discharges as well as the treated effluent from the City of Las Vegas. The hydrodynamics and thermal regime of Lake Mead are complex (LaBounty and Burns 2005). Monitoring of epilimnetic zooplankton populations of the pelagic region in the 3 basins of Lake Mead was conducted from 2000 through 2009. Samples were collected from established sampling stations in the 3 basins, Boulder Basin (n = 8), Colorado River (n = 8) and Overton Arm (n = 6; Fig. 1). Sampling transects for each basin included gradients from stations near inflows to open, deeper water. The US Bureau of Reclamation (Denver Technical Services Center) has collected approximately 20 samples monthly from Boulder Basin and Overton Arm since 2000 and the Colorado River since 2001. The US Bureau of Reclamation (Lower Colorado Office) has collected approximately 20 samples quarterly from all 3 basins of Lake Mead since November 2006. The Southern Nevada Water Authority has collected samples weekly from approximately 6 stations in Boulder Basin since 2000.

Figure 1 Location of Lake Mead zooplankton sampling stations (CR = Colorado River, LVB = Las Vegas Bay, VR = Virgin River). ▴ = Colorado River Stations; • = Overton Arm Stations; • ▪ Boulder Basin Stations.

Limnological characteristics of Boulder Basin, Overton Arm and the Colorado River and details of collection and analyses of chlorophyll a (chl-a) and physical parameters are described elsewhere (e.g., LaBounty and Burns 2005, 2007, Holdren et al. 2006). The trophic state of Lake Mead is mostly mesotrophic with the exception of Boulder Basin, which is classified as supertrophic in the inner basin (located immediately downstream from the Las Vegas Wash) and mesotrophic in the middle and outer basins (LaBounty and Burns 2005). Lake Mead is stratified throughout the majority of the year. Mixing of the epilimnion occurs during a 2 month period (Jan–Feb) to a depth of at least 60–70 m (LaBounty and Horn 1997, LaBounty and Burns 2005, 2007). Average water temperature in the epilimnion ranges from approximately 12 C in early February to approximately 27 C in early August (LaBounty and Burns 2005).

Zooplankton collection and analyses

The epilimnion was operationally defined based on collection methodology. Zooplankton samples were limited to either the top 30 m for stations deeper than 30 m or from the top of the water column to the bottom for stations <30 m deep (Southern Nevada Water Authority's collections were from the top 40 m or to the bottom). Mueller and Horn (2004) documented that the vast majority of zooplankton in Lake Mead are found within the top 40 m of the water column in the open water stations. They also documented that selection of this range for sampling negated any artifacts related to vertical migration of zooplankton because most would be within this range regardless of sampling time. Zooplankton samples were collected with a Wisconsin-style plankton net with 64 or 80 μm mesh, which would underestimate the rotifer population (Chick et al. 2010), but given the importance of daphnids in the ecosystem it is unlikely that epilimnetic rotifer populations were significant (Vanni 1986). Samples were immediately preserved with Lugol's solution and placed on ice after collection.

Appropriate aliquots were placed in an Utermohl chamber, and zooplankton were counted at 100× magnification on a customized Wilovert inverted microscope until at least 200 individuals were tallied (cladocerans, copepods, rotifers, veligers). Identification followed Edmundson (1959), Ruttner-Kolisko (1974), Stemberger (1979), and Pennak (1989). The Zebra Mussel Information System (ZMIS) developed and distributed by the US Army Corp of Engineers was used for identification of Dreissena veligers, and enumeration and identification was conducted at 100× (USACE 2002).

Biomass estimates were based on established length–width relationships (Dumont et al. 1975, McCauley 1984, Lawrence et al. 1987). The lengths or the lengths and widths of each species were measured equal to 10 for common species and less for more rare taxa. More detailed descriptions of biomass estimation methodology are given in Havens et al. (2007) and Havens and Beaver (2010). In accordance with McCauley (1984), biomass was computed for the appropriate number of individuals for each sample location, and the arithmetic mean biomass was multiplied by the species abundance to produce a species biomass for each sample.

Data analyses

The study included 2043 zooplankton analyses over the 10 year sampling period. Because most stations were sampled multiple times in a single month, individual station values for a given month were averaged (22 station means). A grand mean was computed for each of the individual flows (Boulder Basin, Overton Arm, Colorado River) based on station means. Chl-a values (top 0–5 m of the water column) and Secchi disk transparency were obtained from the Southern Nevada Water Authority Members’ database and were computed as described for zooplankton.

Results

Chl-a concentrations and Daphnia abundance and biomass

Basin-wide means for Daphnia abundance (D. pulex and D. galeata) and chl-a concentrations from 2000 through 2009 each demonstrate seasonal trends (Fig. 2). Peaks in chl-a concentrations consistently occurred in Boulder Basin, Overton Arm and the Colorado River during summer and were lowest during winter. Daphnia densities consistently peaked in winter and early spring in each basin. Boulder Basin demonstrated declining trends in both chl-a concentrations and Daphnia abundance during the 10 year study. The less productive Overton Arm and Colorado River stations did not demonstrate a declining pattern in chl-a concentrations. Basin-wide mean Daphnia biomass estimates for the period November 2006 through December 2009 reflect the winter to early spring peak seen in the abundance data (Fig. 3). Daphnia accounted for approximately 70% of the mean total zooplankton biomass for the Lake Mead system (Boulder Basin = 75%, Overton Arm = 69%, Colorado River = 75%). Boulder Basin demonstrated higher mean Daphnia biomass in comparison with Overton Arm and the Colorado River (Boulder Basin = 28.0 μg d.w./L, Overton Arm = 18.4 μg d.w./L, Colorado River =13.8 μg d.w./L). All basins experienced larger Daphnia biomass in 2009 compared to values in 2007 and 2008. An evaluation of the importance of proximity of nutrient inflows to Daphnia biomass and abundance (Fig. 4) showed that Boulder Basin (LVB 3.5) and Overton Arm (VR25.1) stations located immediately downstream from the inflows displayed more Daphnia biomass and abundance for the period January through May for all 3 years under study. The Colorado River stations did not exhibit this relationship, and in 2009 the downstream stations were actually higher than stations near the inflow.

Figure 2 Chl-a (μg/L) and Daphnia abundance (#/L; n = 1872) for (a) Boulder Basin (n = 981); (b) Overton Arm (n = 323); and (c) Colorado River (n = 568) from Jan 2000–Dec 2009.

= Chl-a (μg/L);

= Daphnia abundance (#/L).

Figure 3 Daphnia biomass (μg d.w./L) for (a) Boulder Basin (n = 515); (b) Overton Arm (n = 183); and (c) Colorado River (n = 369) during Oct 2006–Dec 2009. ▪ = Daphnia biomass (μg d.w./L); □ = total zooplankton biomass (μg d.w./L).

Figure 4 Daphnia biomass (μg d.w./L; a = Boulder Basin, b = Overton Arm, c = Colorado River) and abundance (#/L; d = Boulder Basin, e = Overton Arm, f = Colorado River) for the sampling stations within Boulder Basin (n = 255), Overton Arm (n = 80) and Colorado River (n = 148) during the period Jan–May for 2007, 2008, and 2009. □ = 2007;

= 2008; ▪ = 2009.

Quagga veliger abundance and distribution

Quagga veligers were first detected in zooplankton samples collected at CR 355.75 (the Narrows) in November 2006, and adults were confirmed at the Las Vegas Boat Harbor and at Lake Havasu, a downstream reservoir, in January 2007. Boulder Basin and Colorado River stations experienced peaks in fall 2007 (Fig. 5). Quagga veliger abundance in Overton Arm in fall 2007 was modest; however, by fall of 2008 the mean abundance exceeded the other basins (>30/L). A midsummer decrease in mean quagga veliger abundance occurred in Boulder Basin and Colorado River 2007 and 2008 and was followed by increased abundance until October. In 2008 and 2009, quagga veligers displayed a clear bimodal seasonality with spring and fall pulses that corresponded to water temperatures of 18–20 C in the top 5 m of the water column in this study and other studies of Lake Mead (LaBounty and Burns 2005). By 2009 the densities of veligers decreased in Boulder Basin and Overton Arm but maintained the bimodal seasonality observed in 2007 and 2008. Colorado River veliger abundance in 2009 was similar to 2007 and 2008. Quagga veligers were most numerous in open water areas with maximum depth, greater clarity, low chl-a concentrations (Fig. 6; Table 1) and the largest amount of submerged geological features (Twichell et al. 1999, Twichell and Cross 2004).

Discussion

Comparison with previous studies on Lake Mead

Previous Lake Mead research indicates that the response of the phytoplankton community to nutrient loadings has been the single most important factor regulating the seasonal, spatial and temporal abundance of zooplankton in the phosphorus (P)-limited reservoir. Established temporal and spatial patterns in primary productivity and phytoplankton populations include maxima near point source nutrient inputs (e.g., Las Vegas Wash [Boulder Basin], Virgin and Muddy rivers [Overton Arm]), decreasing with distance from the inflows (e.g., LaBounty and Horn 1997, Holdren et al. 2006). Phytoplankton populations typically peak from July through September. Similar temporal and spatial patterns in zooplankton and fish populations in Lake Mead have been ascribed to variations in nutrient inputs (nutrient abatement–nutrient inflows) and associated gradients in phytoplankton populations (e.g., LaBounty et al. 2004, Mueller and Horn 2004).

Consistent with previous studies, Daphnia populations in this study were low during summer, increased in fall, reached maximum in winter–early spring and then decreased again during spring. The highest Daphnia densities occurred in the inner Las Vegas Bay followed by inflow stations in the Overton Arm. Daphnia densities consistently decreased down-reservoir from the inflows for these 2 basins, and densities were lowest in the main basin stations. In contrast, the more rapidly moving Colorado River inflow stations did not follow this pattern. The importance of nutrient inputs to the trophic linkages in Lake Mead is underscored by the results of the in situ fertilization of Overton Arm in the late 1980s which resulted in a greater than 4-fold increase in Daphnia pulex abundances and improved lipid storage and egg production rates (Axler et al. 1988). Recently, Loomis (2009) documented the trophic linkage between the forage fish base (threadfin shad) and zooplankton in Las Vegas Bay and Overton Arm of Lake Mead (also see Wilde and Paulson 1988).

North temperate zone studies

Ecological impacts of dreissenid mussels in north temperate systems include reduced phytoplankton populations and less efficient energy transfer to higher trophic levels (Conroy and Culver 2005, Garton et al. 2005). Filter-feeding competition between dreissenid mussels and both macrozooplankton and microzooplankton for food resources (phytoplankton and bacteria) has adversely impacted fisheries in the Great Lakes and resulted in energy transfer to the benthos and away from planktivorous fish (Bridgeman et al. 1995, MacIsaac et al. 1995, Nalepa et al. 2008, Fahnenstiel et al. 2010). Empirical evidence from the Great Lakes region indicates that dreissenid grazing activities can also have a positive impact on cyanobacteria (i.e., Microcystis aeruginosa) in low-nutrient, P-limited lakes (P < 25 μg/L) despite reductions in P-loadings to the systems (Vanderploeg et al. 2002, Raikow et al. 2004, Knoll et al. 2008). Dreissenid mussels exhibit differential grazing on algal species (e.g., Lavrentyev et al. 1995) and prefer hard, deep and colder habitats, but with increasing depth they will colonize sands and silty sands (Mills et al. 1996).

Figure 5 Quagga veliger abundance (#/L) for (a) Boulder Basin (n = 515); (b) Overton Arm (n = 183); and (c) Colorado River (n = 369) during Oct 2006–Dec 2009.

= quagga veliger abundance (#/L).

Figure 6 Abundance of quagga veligers (organisms/L) at the sampling locations on (a) October 2007, (b) October 2008, and (c) October 2009. Sampling station CR343.2 was not sampled in October 2009. Sampling station VR12.9 was not sampled on the selected dates.

Table 1 Characteristics of zooplankton sampling stations. For each station, values represent means for chlorophyll a (n = 1506) and Secchi depth (n = 1148).

	Station ID	Depth (m) Apr 2009	Chl-a (μg/L)	Secchi Depth (m)
Boulder Basin	LVB 3.5	∼14	7.3	3.5
	LVB 4.95	∼41	2.1	7.3
	LVB 6.7	∼75	1.2	10.0
	LVB 7.3	∼87	1.1	10.3
	Intake	∼49	1.1	10.5
	CR 346.4	∼116	1.0	11.5
	CR 343.2	∼118	1.0	11.5
	CR 342.5	∼118	0.9	12.1
Overton Arm	VR 25.1	∼5	3.2	2.2
	VR 18.0	∼33	1.3	5.1
	VR 13.0	∼50	0.9	8.0
	VR 12.9	∼51	1.2	7.9
	VR 9.4	∼66	0.9	9.3
	VR 6.0	∼86	1.1	9.3
Colorado River	CR 394.0	∼47	2.3	4.3
	CR 390.0	∼63	1.7	5.0
	CR 380.0	∼79	1.1	8.3
	VR 2.0	∼94	1.2	10.1
	CR 360.7	∼105	1.1	11.7
	CR 355.75	∼107	1.0	12.1
	CR 351.7	∼112	0.9	11.6
	CR 348.4 NW0.8	∼93	1.0	11.6
LVB: Las Vegas Bay, CR: Colorado River, VR: Virgin River.

Importance of temperature and season

Life history patterns of phytoplankton populations, Daphnia and the major forage fish in Lake Mead, threadfin shad (Dorosoma petenense), are strongly coupled (Wilde and Paulson 1988, 1989) and at least partially dependent on water temperature. Following thermal overturn of the epilimnion in winter (∼12 C in Jan–Feb), the phytoplankton community of Lake Mead is dominated by high-quality algal food resources such as cryptomonads and chrysophytes, which are replaced by chlorophytes in the spring (18–20 C) and increased cyanobacteria in the summer (>25 C; LaBounty et al. 2004; Beaver et al. unpublished data). Life history traits for Daphnia pulex include a strong relationship between body-size, fecundity and feeding efficiency, with optimal temperatures around 20 C and decreasing population and metabolic status at higher temperatures (Lynch 1977, Rellstab and Spaak 2009). Elevated summer epilimnetic temperatures in Lake Mead favor a phytoplankton community dominated by low-quality food resources (i.e., cyanobacteria), and Daphnia populations are greatly reduced. Threadfin shad spawn concurrently with the spring Daphnia pulse from March through June (Wilde and Paulson 1989), and their annual productivity is strongly related to total phosphorus (TP) and chl-a concentrations (Ney 1996). Summer temperatures in the epilimnion of Lake Mead (>25 C) are also suboptimal for dreissenid mussels (Spidle et al. 1995), and lower food quality (i.e., cyanobacteria) may increase their respiration rates and energy expenditures resulting in reduced reproduction and filter-feeding rates (Stoeckmann 2003).

Conclusions

Because of their contribution to total zooplankton biomass, Daphnia are the key intermediaries in energy transfer to higher trophic levels in Lake Mead. Largest populations of Daphnia were associated with nutrient-rich and higher primary production inflow areas, while quagga veligers were most numerous in open water areas. Impacts of invasive quagga mussels on food web dynamics in Lake Mead are unclear. Conventional thought that dreissenid mussel grazing will result in major ecological change outside the north temperate zone is unsubstantiated by our data. Lake Mead is a monomictic reservoir with contrasting thermal, hydrological and ecological conditions to those of the north temperate ecosystems that form the basis for empirical observations on the impacts of dreissenid mussels. Thermal overturn of the epilimnion is limited to winter months, and grazing impacts of quagga mussels are more intensive in well-mixed systems than thermally stratified systems (Ackerman et al. 2001, Idrisi et al. 2001, Kissman et al. 2010, Fahnenstiel et al. 2010). The success of quagga mussels in the Lake Mead ecosystem could be measured by reduction in epilimnetic phytoplankton and zooplankton populations and increased importance of cyanobacteria. Chl-a concentrations and Daphnia abundance have decreased only in Boulder Basin since drought conditions and subsequent invasion by quagga mussels; however, interpretation of the ecological significance is confounded by simultaneous reductions in nutrient loadings from improved wastewater treatment and hydrologic inflows (LaBounty and Burns 2007, LaBounty 2008). Nevertheless, the well-described trophic level interactions beginning with thermal overturn in winter and subsequent pulses of high-quality phytoplankton, Daphnia and fish populations in the spring will likely persist subject to continued, localized nutrient inflows.

Acknowledgments

This work is a continuation of research initiated by Dr. James LaBounty. His critical insights and forethought on the importance of the functioning of the Lake Mead ecosystem have provided all researchers with a better understanding of the complexities of the ecosystem. The authors are grateful to 3 anonymous reviewers for their constructive comments.