Can River Otters Naturally Recolonize the Grand Canyon?
By Merav Ben-David, Ph.D.

River otters (Lontra canadensis) are piscivorous predators, which forage near the apex of the trophic pyramid and readily accumulate high levels of pollutants. Indeed, river otters in North America were reduced throughout much of their historic range by the early 1900s because of pollution, urbanization, and overharvest. Consequently, numerous projects were initiated to reintroduce river otters to areas from which they were extirpated. Many of the reintroduction programs were highly successful (see article in our Autumn 2001 issue on the river otter in Missouri), but also very costly and sometimes came at the expense of individual animals. Recently, a proposal to reintroduce river otters to the Grand Canyon was put forward. Although this proposal has large appeal, reintroducing river otters to the Grand Canyon has several limitations.

The population status and current distribution of the subspecies that historically inhabited the Grand Canyon, L. c. sonora, is unknown. The source populations for reintroduction will likely be of different subspecies trapped in places such as Alaska, Missouri, Louisiana, and Florida, where otter populations are stable and trapping will likely not affect population status. Thus, a successful reintroduction in the Grand Canyon may lead to genetic swamping of the few remaining L. c. sonora in adjacent watersheds. Genetic swamping occurs when a large number of individuals with new genetic material breeds with the few original inhabitants of the area producing offspring that carry little of the genetic information of the original population.

In addition, such a reintroduction effort will likely be costly and require numerous permits, health evaluations, veterinary care, and complicated logistics. In every reintroduction program newly trapped animals require health monitoring to ensure their subsequent post-release survival, as well as disease screening to ensure that the newly released animals are not transporting infectious diseases into new areas. Such monitoring may require detention of otters in captivity for long periods of time (rabies testing for example may require several months of observations). Maintaining wild caught otters in captivity for long periods of time may compromise their subsequent post-release survival.

Therefore, before such a project is initiated, the feasibility of natural recolonization of river otters from the upper reaches of the Colorado River and tributaries needs to be assessed. We (Nathan P. Nibbelink, Gail M. Blundell, and myself) are developing a spatially explicit, individual-based model to evaluate the potential for natural recolonization of the Grand Canyon by river otters that may disperse from known populations in the upper Colorado River Basin. This spatially explicit model creates virtual otters that survive year to year, reproduce, and some of the young disperse into new available habitats along the virtual Colorado River and tributaries. At the beginning year (or rather first iteration of the model) the computer plots on the map a number of individuals in Rocky Mountain National Park (RMNP) and the Wyoming Green River (WGR) where surveys identified viable otter populations.

Photo by Eric Peterson©

The number of animals on the landscape (density) is based on results of surveys conducted by the University of Wyoming Student Chapter of the Wildlife Society in RMNP this year (see article in this issue). The number of males and females in our model is assumed to be equal (50:50 sex ratio). Each year the computer selects an individual and then based on the survival rate in the population "decides" whether the animal will live or die. If the animal dies its home range becomes empty and is available for dispersing animals. If the animal lives it reproduces. In our model (as in real life) only females produce young. Each of the young is then assigned a gender and survivorship. For surviving young the computer then "decides" whether the animal will disperse or not and how far it will go. If the dispersing animal encounters a vacant home range it will settle there. If not it will move once more. Again, the computer assigns the animal a dispersal distance and if the animal does not find a vacant home range after its second dispersal attempt it will die. The computer cycles with this "decision" process through each and every individual in the population and at the end of the year (iteration) updates the location of each virtual animal on the virtual landscape. The final iteration of the model is when otters reach the Grand Canyon.

How does the computer "decide" whether an animal will survive or die? Or how many young it will have? We create a distribution of survival rate and "ask" the computer to randomly draw from it.

These distributions are derived from data collected in other studies and published in the literature. Similarly we use distributions of litter sizes, number of dispersers per generation, and dispersal distances. We use different survival rates, reproductive rates, dispersal rates, and dispersal distances in each simulation to explore the effects of the level of these variables on population expansion.

One can imagine that as the population increases (more individuals) and expands, the time it takes the computer to run through each iteration increases, because the same calculations are done for each individual animal. Therefore such spatially explicit, individual based models require high computer power and appropriate computing facilities. We thank the Wyoming Geographic Information Science Center for providing us with these capabilities.

In our initial simulations, assuming no barriers to dispersal, survival rates were by far the strongest limitation to river otter population expansion. Those simulations resulted in range expansion of 2,000 to 6,000 square kilometers per year. At that rate it would take between 45 and 134 years for otters to naturally recolonize the Grand Canyon from RMNP and the WGR. These values suggest that natural recolonization under the best circumstances will not occur in our lifetimes. One needs to remember, however, that the model assumes that no otters exist outside RMNP and WGR, which is probably incorrect. If otters are already common further down stream then natural recolonization of the Grand Canyon will be faster. Therefore, our task will be to survey larger areas in Wyoming, Colorado, New Mexico and Arizona to determine the current population status of river otters along tributaries of the Colorado River.

On the other hand, in our initial simulations we ignored barriers to dispersal all together. We did not account for the potential effects of dams, roads, urban areas, water loss, pollution, and lack of suitable habitat on both population dynamics (survival and reproduction rates) and dispersal. Future modeling work, therefore, will attempt to assess relative effects of movement barriers to river otter dispersal. This exercise, we hope, will assist us in making management recommendations for improving dispersal corridors for river otters in the Colorado River Basin.

By improving habitat quality for river otters along the Colorado River we are likely to not only help river otters naturally recolonize the Grand Canyon but also improve habitats for multitudes of other species including ourselves.