Since the dams: Historical ecology of the Colorado Delta

wpe1.jpg (67582 bytes)

Karl W. Flessa, David L. Dettman, Bernd R. Schöne, David H. Goodwin, Carlie A. Rodriguez, Sarah K. Noggle

Department of Geosciences, University of Arizona, Tucson, AZ 85721 USA

Miguel A. Téllez-Duarte, Guillermo E. Avila-Serrano

Facultad de Ciencias Marinas, Universidad Autónoma de Baja California, Ensenada, B.C., Mxico

Michael Kowalewski

Department of Geological Sciences, Virginia Tech, Blacksburg, VA  24061 USA

Glenn A. Goodfriend

Department of Geology, George Washington University, Washington, D.C. 20052 USA

Poster presented at United States -Mexico Colorado River Delta Symposium, International Boundary and Water Commission, Department of Interior and the Mexican Secretariat of the Environment and Natural Resources, Mexicali, Baja California, Mexico, September 11-12, 2001.


            In most years, the Colorado River no longer reaches the sea. Major upstream dams now regulate downstream river flow and allow the diversion of approximately 90% of the Colorado's freshwater for cites and crops in Arizona, Nevada, and southern California.  By international treaty, Mexico is entitled to approximately 10% of the river's annual flow.  In Mexico, the water is used for irrigation and domestic consumption in Sonora and Baja California.  With the exception of some unusually wet years, little freshwater has reached the Gulf of California since 1960 (Fig. 1).

How has this drastic reduction of freshwater flow affected the marine fauna of the Colorado Delta?  Because no systematic surveys of the delta's fauna were made until 1998, we use the shelly remains of bivalve mollusks to reconstruct environmental conditions, species composition, and productivity in the era before the dams.  We are putting the dead to work to

wpe4.gif (15828 bytes)

  Figure 1. Colorado River flow at Yuma, AZ, 1878-1997.

Environmental Change On the Colorado Delta            

A direct result of the decline in freshwater influx has been an increase in the salinity of the water in the estuary and upper Gulf. Early observations (Townsend, 1901) and measurements during controlled releases of river water (Lavín and Sánchez, 1999) indicate that salinities in the 32 to 35‰ range were common before the dams. This is in sharp contrast to measurements made since the construction of upstream water diversions. Now, salinities are typically in the 35-45‰ range (Alvarez-Borrego et al., 1975; Flessa, personal observations).

            The increase in the salinity of the water in the river’s estuary profoundly changed the circulation in the upper Gulf of California (Lavín and Sánchez, 1999; Lavín, et al., 1998; Carbajal et al., 1997). When the less dense riverwater entered the estuary, it tended to flow into the Gulf at the surface, causing a landward bottom flow of more saline,  denser, marine water. Carbajal et al. (1997), Lavín and Sánchez, (1999) and (Rodriguez et al., 2001) estimate that the zone of freshwater mixing extended as far as 60 km from the river’s mouth.

            Since the diversion of much of the river’s fresh water, the estuarine circulation is now driven by the evaporation of Gulf water in the river’s mouth. High evaporation rates generate dense, saline water that sinks and flows along the bottom into the upper Gulf, while relatively less dense seawater flows toward the estuary near the surface (Lavín et al., 1998).

            Upstream dams and diversion projects have also trapped and diverted much of the Colorado’s sediment load. The river once delivered approximately 160 million metric tons of sediment to the delta every year (van Andel, 1964). The sediment load today is almost zero.  Waves and tidal currents are now eroding the previously deposited fine-grained sediments (Carriquiry and Sánchez, 1999). This sediment reworking is responsible for the high turbidity of the upper Gulf’s waters.  

     wpe8.jpg (24719 bytes)          wpe2.jpg (44096 bytes)

Figure 2. Cheniers (beach ridges) of the Colorado River delta plain.  Cheniers are composed of shells washed out from the intertidal zone


Shells and cheniers

            Although waves and tidal currents are strong enough to remove the previously deposited mud and silt, coarse-grained material, such as shells, are left behind and accumulate in beach deposits known as cheniers (Augustinus, 1989). These shell-rich deposits (Fig. 2) line the Baja California side of the delta for a distance of more than 40 km (Kowalewski and Flessa, 1995).  These active cheniers began to form when river sediment became trapped in Lake Mead, following the completion of Hoover Dam in 1935 (Thompson, 1968). The active cheniers migrate to the west during storms and extreme high tides, marking the retreat of the sediment-starved delta

          The more landward cheniers in Figure 2 mark earlier episodes of sediment starvation -- times when the Colorado River was naturally diverted into the Imperial Valley, trapping the river's sediments in Holocene Lake Cahuilla (Waters, 1983).  Radiocarbon dating of charcoal and peat in beach deposits indicate that Lake Cahuilla's last highstand was between 1640 and 1690 A.D. (Gurrola and Rockwell, 1996). 

wpeF.gif (15751 bytes)

Figure 3.  Calibration and the age of shells of the active chenier.  A. Age-frequency distribution of 125  shells of Chione.  B.  Calibration of change in amino acid ratio (alloisoleucine/isoleucine) with radiocarbon age.(after Kowalewski et al. 1998).

            Shells in the active chenier range in age from 1000 AD to the present (Fig. 3). We used a radiocarbon-calibrated, amino-acid geochronology to date 125 shells. The chenier is a time-averaged deposit that contains shells from every 50-year interval during the past 1,000 years (Kowalewski et al., 1998).

  Faunal Change On the Colorado Delta           

             The shell-rich cheniers of the delta are a record of the composition and abundance of the shelly mollusks that lived in the era before upstream dams and river management.  Comparison with the living mollusk fauna documents the effects of freshwater diversion on the shallow marine fauna of the delta.

            Before the dams, the shelly fauna was dominated by the Colorado delta clam: Mulinia coloradoensis (Fig. 4A).  Specimens of this bivalve are 91% of the shells in the active chenier, with the venerid bivalve Chione (Fig. 4B) only 4% of the fauna (Fig. 5).  Today, 87% of the shells  are Chione; Mulinia coloradoensis is only 5% of the live fauna (Fig 5).  The diversion of freshwater has changed the composition of the delta's shelly fauna.  

                                       wpe12.jpg (22611 bytes)    wpe10.jpg (19090 bytes)

Figure 4. A.  The Colorado delta clam, Mulinia coloradoensis, the most abundant mollusk species on the delta before the dams.  B.  Shells of Chione, the most abundant mollusk living on the delta today.

wpe1.gif (10542 bytes)

Figure 5.  Since dams and water diversions, Chione has replaced Mulinia as the dominant mollusk on the delta and abundance has decreased 94%.  

Decreased productivity

            Before the dams, mollusks were much more abundant than at present.  Approximately two trillion shells larger than 12.5 mm are in the active cheniers.  These shells are a record of the delta's former benthic productivity during the past 1,000 years. 

            We can use these shells to estimate population densities before the dams (Kowalewski et al., 2000).  Growth lines in shells of the Colorado delta clam indicate that the average-sized individual was three years old. This indicates that 333 generations of clams lived during the past millennium, for an average standing population of six billion bivalve mollusks. These clams lived either within the extensive intertidal zone or within both the intertidal and shallow subtidal zones and reached densities of 25 to 50 individuals per m2 in the time before the dams.

            We sampled the live fauna along seven transects in 1999 and 2000. We measured shelly faunal density in the upper 20 cm of the sediment in each of 258 sample stations. The average density of living mollusks today is 3 individuals per m2 -- as much as 94% lower than densities before the dams (Fig. 9).

Clams as ecosystem indicators       

            The high density of benthic mollusks before the dams suggests that crabs, polychaete worms, and other benthic invertebrates were also abundant then.  Mollusks and other organisms were probably sustained by nutrients supplied by the river.  The high density of benthic invertebrates likely also nourished a greater abundance of predatory birds and fish than present in the region today.  Today, shrimp catches in the upper Gulf of California increase following controlled releases of Colorado River water (Galindo-Bect et al., 2000), also suggesting that the river had a major role in supporting marine life in the upper Gulf of California.  The diversion of freshwater has decreased the abundance of the delta's fauna.


Species At Risk On The Colorado Delta

             Three Colorado delta species are at risk of extinction: the Colorado delta clam, Mulinia coloradoensis, the totoaba fish (Totoaba macdonaldi), and the Gulf of California harbor porpoise, or vaquita (Phocoena sinus). The totoaba and the vaquita are listed as endangered under the U.S. Endangered Species Act and both are on the I.U.C.N. (International Union for Conservation of Nature) Red List of Threatened Species.


The Colorado Delta Clam

            The bivalve Mulinia coloradoensis  (Fig. 4A) is known only from the upper Gulf of California.  It was once the most abundant mollusk living on the Colorado Delta.  Trillions of this species' shells have washed out of the tidal flats and now form the shelly beaches, cheniers and shoals of the delta.  Before the dams, the Colorado delta clam ranged as far as 60 km from the river's mouth and densities reached 46 individuals per m2.  Today, the species typically occurs within 30 km of the river's mouth and at densities of only 0.15 individuals per m2.

            The dramatic decline of the Colorado delta clam since upstream diversion of freshwater is most likely the result of the increased salinity of its habitat (Rodriguez, Flessa and Dettman, 2001).  Evidence for the importance of freshwater mixing in the clam's habitat comes from the isotopic geochemistry of the clam’s shell.

            More of the lighter isotope of oxygen (16O) occurs in river water, so shells that are grown in a mixture of seawater and riverwater have oxygen isotope ratios in their calcium carbonate (CaCO3) that are lower than in shells grown in seawater.  The oxygen isotope composition of water or shell is reported as the deviation of the ratio of 18O to 16O from a standard, in typical notation del18O, where units are per mil (‰).  Seawater is close to 0‰ and pre-dam Colorado River water ranged from –6 to –17 ‰ (Dettman et al., 1999).


Figure 6.  Comparison of oxygen isotopes in pre-dam shells of Mulinia coloradoensis and live-collected Chione sp. indicate that Mulinia coloradoensis lived in lower-salinity water than present today (Rodrgiuez et al., 2001).    

wpe4.gif (22973 bytes)  

Figure 7.  A changed river.  Seasonal variation of river flow at Yuma, Az. in 1911, before the dams (blue) and seasonal variation in river flow below Morelos Dam, Mexico, below the last diversion (yellow).

  wpe7.gif (17338 bytes)

Figure 8.  Isotopic record of spring snowmelt in a pre-dam shell of Mulinia coloradoensis.

            We compared isotopic variation in shells of live-collected Chione to variation in shells of pre-dam Mulinia coloradoensis (Fig. 6).  The range in del18O values in Chione is largely caused by the seasonal fluctuation in temperature during the growth period of the bivalve. 


            Del18O values in the Colorado delta clam have a much broader range, sometimes reaching as low as -6.4‰.  Values below the -2.5‰ minimum found in live Chione are the effect of freshwater formerly delivered by the Colorado River.  The Colorado delta clam was abundant when salinity in the area was lower.

            Before upstream river management, the flow of the Colorado River was strongly seasonal.  Approximately 70% of the annual flow arrived at the delta in the months of May, June and July (Fig.  7, see Harding et al., 1995), the result of spring snowmelt in the Rocky Mountains.  The annual spring snowmelt is evident in the oxygen isotope variation in the shell of a pre-dam Mulinia coloradoensis (Fig. 8).  Del18O values below the summertime minimum of -2.5‰ signal the arrival of a large pulse of freshwater into the river's estuary.  The timing of freshwater influx may have been as important to the biology of the Colorado delta clam as the lowered salinity that resulted.  Pulses of freshwater may have regulated growth and/or triggered spawning in this species.


An endangered species

            Unless some freshwater flow to the delta is resumed, the small population size, limited geographic range, and altered habitat of Mulinia coloradoensis places this species at risk of extinction.  


The totoaba

  wpe1B.jpg (41432 bytes)

Figure 9. Totoaba, endangered fish in Colorado Delta area. (Image from

            Totoaba, Totoaba macdonaldi, (Fig. 9) are large schooling fish that were once the basis of a thriving commercial and sport fishery in the upper Gulf of California.  These are large fish, with a maximum reported weight of ~100 kg and a length near 2 m (Flanagan and Hendrickson, 1976).  Adults spawn in the Colorado delta region in the winter, remaining in the area for several weeks; juveniles remain the nursery grounds for two years before migrating to the south (Cisneros-Mata et al., 1995).  Sexual maturity is reached after six or seven years, and growth banding in otoliths ("ear bones"), indicate that some individuals may reach 25 years in age (Román-Rodriguez and Hammann, 1997).

            The totoaba’s decline is usually attributed to overfishing, bycatch in shrimp nets, and poaching. In addition, increased salinity in the river’s estuary may have degraded the fish’s spawning and nursery grounds (Cisneros-Mata, et al., 1995).  


The vaquita        

  wpe1D.gif (24687 bytes)

Figure 10. Vaquita, or Gulf of California harbor porpoise. (Image from


            The vaquita, Phocoena sinus, (Fig. 10) is a small porpoise limited in its distribution to the upper Gulf of California; all confirmed sightings are within 150 km of the mouth of the Colorado River (Vidal, 1995). The vaquita is thought to be the most endangered marine cetacean, with a population size estimated at between 106 and 470 individuals (Barlow et al., 1997).  The biology of the species is known principally from the study of individuals accidentally caught in nets, beach strandings, and museum specimens. Hohn et al. (1996) examined 56 individuals, finding a bimodal age structure and a maximum age of 21 years. Sexual maturity occurs between three and six years of age, reproduction is in late winter - early spring, and the calving interval is greater than one year.  Adults tend to reach 140 cm (females) and 135 cm (males) in length, slightly smaller than its cool-water relative, Phocoena phocoena.  Stomach contents include squid beaks and fish remains (Vidal, 1995).

            The principal current source of mortality of the vaquita is its incidental capture in fishing nets (Hohn, et al., 1996; D’Agrosa et al., 2000) but the role of increased salinity in its key habitat is unknown.



Looking to the future

            The environment and fauna of the Colorado River Delta and upper Gulf of California have been changed by the diversion of freshwater for human uses.  The relative abundance of mollusk species has changed, benthic productivity has greatly decreased, and at least one species is threatened with extinction because of increased salinity in its habitat. Although the resumption of "excess" Colorado River flow to the Gulf has resulted in a partial revival of the delta's riparian habitats (Leucke et al., 1999), the amount of flow needed to partially restore marine habitats is unknown.

            Our future work will (1) examine the role of freshwater in the biology of the Colorado delta clam; (2) estimate the minimum river flow needed to sustain a viable population of the Colorado delta clam; (3) examine isotopic variation in otoliths of the totoaba to determine if river flow is important to the biology of the fish; and (4) examine isotopic variation in skeletal remains of the vaquita to determine if river flow is important to the biology of the porpoise.

            The delta is not dead, but its partial restoration depends on international collaboration in conservation science, river management and resource utilization.

  wpe9.jpg (84348 bytes)

Figure 11. Trillions of shells of the Colorado delta clam, Mulinia coloradoensis form the beaches and cheniers of the delta.  Today, this species is rare.


           We thank the following organizations for their support of our research: the National Science Foundation (U.S.), Consejo Nacional de Ciencia y Tecnología (México), the Eppley Foundation for Research, the Western Water Division of the U.S. Geological Survey, the Center for Biological Diversity, Defenders of Wildlife, the University of Arizona, and the Universidad Autónoma de Baja California.  We thank José Campoy-Favela (Director, Reserva de la Biosfera Alto Golfo de California y delta del Río Colorado) for his support and assistance.  We also thank the many pangueros of the upper Gulf of California for skillfully navigating their boats in the shallow waters of the delta, with special thanks to Ramón Soberanes and the late Martín Pescador.


Álvarez-Borrego, S., Flores-Báez, B.P., Galindo-Bect, L.A., 1975. Hidrología del Alto Golfo de California II.   Condiciones  durante inveierno, primavera y verano.  Ciencias Marinas, v. 2, p.21-36.

Augustinus, P.G.E.F., 1989. Cheniers and chenier plains: A general introduction. Marine Geology, v. 90, p. 219-229.

Barlow, J., Gerrodette, T., and Silber, G., 1997.  First estimates of vaquita abundance.  Marine Mammal Science, v. 13, p. 44-58.

Carbajal, N., Sousa, A., and Durazo, R., 1997, A numerical model of the ex-ROFI of the Colorado River.  Journal of Marine Systems, v. 12, p. 17-33.

Carriquiry, J.D. and Sánchez, A., 1999. Sedimentation in the Colorado River delta and upper Gulf of California after nearly a century of discharge loss. Marine Geology, v. 158, p. 125- 145.

Cisneros-Mata, M.A., Montemayor-López, G., and Román-Rodríguez, M.J., 1995. Life history and conservation of Totoaba macdonaldi. Conservation Biology, v. 9, p.806-814.

D'Agrosa, C., Lennert-Cody, C.E. and Vidal, O., 2000. Vaquita bycatch in Mexico's artisanal gillnet fisheries: Driving a small population to extinction.  Conservation Biology, v. 14, p. 1110-1119.

Dettman, D.L., Kendall, C., Lohmann, K.C., and Flessa, K.W., 1999.  A new method for isotope paleohydrology:  Molluscan carbonate, North American rivers, and the seasonal oxygen isotope cycle of the pre-dam Colorado River.  EOS (Supplement), Transactions of the American Geophysical Union v. 80 (46), p. F480.

Flanagan, C.A. and Hendrickson, J.R., 1976.  Observations on the commercial fishery and reproductive biology of the totoaba, Cyncoscion macdonaldi, in the northern Gulf of California. Fishery Bulletin, v. 74, p. 531-544.

Galindo-Bect, M.S., Glenn, E. P., Page, H. M., Galindo-Bect, L. A., Hernandez-Ayun, J. M.,Petty, R. L., and Garcia-Hernandez, J., 2000, Analysis of the penaeid shrimp catch in the northern Gulf of California in relation to Colorado River discharge. Fishery Bulletin, v. 98, p. 222-225.

Gurrola, L.D. and Rockwell, T.K., 1996.  Timing and slip for prehistoric earthquakes on the Superstition Mountain fault, Imperial Valley, southern California.  Journal of Geophysical Research, v. 101, p.5977-5985.

Harding, B.L., Sangoyomi, T.B., and Payton, E.A., 1995.  Impacts of a severe sustained drought on Colorado River water resources.  Water Resources Bulletin, v. 31, p. 815-824.

Hohn, A.A., Read, A.J., Fernández, S., Vidál, O., and Findley, L.T., 1996. Life history of the vaquita, Phocoena sinus, (Phocoenidae, Cetacea). Journal of Zoology (London), v. 239, p. 235-251.

Kowalewski, M., and Flessa, K. W., 1995, Comparative taphonomy and faunal composition of shelly cheniers from  northeastern Baja California, Mexico.  Ciencias Marinas, v. 21, p. 155-177.

Kowalewski, M., Goodfriend, G.A. and Flessa, K.W., 1998.  High-resolution estimates of temporal mixing within shell beds: The evils and virtues of time-averaging. Paleobiology, v. 24, p. 287-304.

Kowalewski, M., Ávila-Serrano, G.E., Flessa, K.W., and Goodfriend, G.A., 2000, Dead delta’s former productivity: Two trillion shells at the mouth of the Colorado River.  Geology, v. 28, p. 1059-1062.

Lavín, M.F., Godínez, V.M., and Alvarez, L.G., 1998. Inverse–estuarine features of the upper Gulf of California.  Estuarine, Coastal and Shelf Science, v. 47, p. 769-795.

Lavín, M.F., and Sánchez, S., 1999, On how the Colorado River affected the hydrography of the upper Gulf of California. Continental Shelf Research, v. 19, p. 1545-1560.

Luecke, D.F., Pitt, J., Congdon, C., Glenn, E., Valdés-Casillas, C., and Briggs, M., 1999.  A delta once more.  Environmental Defense Fund Publications, Washington, D.C., p. 51.

Rodriguez, C.A., Flessa, K.W., and Dettman, D.L., 2001, Effects of upstream diversion of Colorado River water on the estuarine bivalve mollusc Mulinia coloradoensis.  Conservation Biology, v. 15, p. 249-258.

Rodriguez, C.A., Flessa, K.W., Tellez-Duarte, M.A., Dettman, D.L. and Avila-Serrano, G.A., 2001.  Macrofaunal and isotopic estimates of the former extent of the Colorado River estuary, upper Gulf of California, Mexico. Journal of Arid Environments, v. 49, p. 183-193.

Román-Rodríguez, M. J. and Hammann, M. F.,1997. Age and growth of totoaba, Totoaba macdonaldi, (Sciaenidae), in the upper Gulf of California. Fishery Bulletin, v 95, p. 620-628.

Thompson, R. W., 1968, Tidal Flat Sedimentation on the Colorado River Delta, Northwestern Gulf of California.  Geological Society of America Memoir 107, Boulder, Co. p. 133.

Townsend, C.H., 1901. Dredging and other records of the U.S. Fish Commission Steamer"Albatross" with bibliography relative to the work of the vessel. US Fish Commission Report for1900, p. 387-562.

Vidal, O., 1995. Population biology and incidental mortality of the vaquita, Phocoena sinus. Report of the International Whaling Commission Special Issue, v. 16, p. 247-272. 

 Waters, M.R., 1983.  Late Holocene lacustrine stratigraphy and archaeology of ancient Lake Cahuilla, California.  Quaternary Research, v. 19, p. 373-387.



Centro de Estudios de Almejas Muertas at the University of Arizona -

San Diego State University's Salton Basin-Colorado Delta Mothersite -
Reserva de la Biosfera Alto Golfo de California y Delta del Río Colorado –
Estación de Campo Golfo de Santa Clara -
International Boundary and Water Commission -
Yuma Area Office, U.S. Bureau of Reclamation -

Environmental Defense's report on the Colorado River Delta -