Wires L.R., F.J. Cuthbert, D.R. Trexel & Joshi A.R . 2001
Status of the Double-crested Cormorant (Phalacrocorax auritus) in North America.
Final Report to USFWS

- EXECUTIVE SUMMARY -

Introduction:
Since the late-1970s, numbers of Double-crested Cormorants (Phalacrocorax auritus) (DCCO) have increased significantly in many regions of North America. A variety of problems, both real and perceived, have been associated with these increases, including impacts to aquaculture, sport and commercial fisheries, natural habitats, and other avian species. Concern is especially strong over impacts to sport and commercial fishes and aquaculture. Because of increasing public pressure on U.S. government agencies to reduce DCCO conflicts, the USFWS is preparing an Environmental Impact Statement (EIS), and in conjunction with the U.S. Department of Agriculture/Wildlife Services (USDA/WS) and state resource management agencies, will develop a national management plan for the DCCO. This assessment will be used to prepare the EIS and management plan.

Populations and trends:
The DCCO breeding range in North America is divided into five geographic areas. Since at least 1980, numbers have clearly increased in three of the breeding areas: Canadian and U.S. interior, Northeast Atlantic Coast and Southern U.S. In these populations, much of the growth occurred between the late 1970s – early 1990s; from the early 1990s – 2000 growth rates have slowed or appeared to stabilize in many states and provinces. For the Pacific Coast and Alaskan breeding populations it was not possible to summarize trends overall because recent data for birds breeding in significant portions of these regions (e.g., Alaska, Mexico) are not available, or have not been collected in a coordinated and timely fashion for the populations as a whole. Along some parts of the Pacific Coast, breeding numbers declined in the 1990s (e.g., British Columbia, species is listed as Vulnerable and is being considered for Threatened status). In other areas significant increases occurred. Concurrently, numbers also increased on the wintering grounds, particularly in the Mississippi River Delta region, an area of high human-cormorant conflict over catfish resources.

Many historical records from across the continent indicate that the species was or may have been more abundant and widespread than is currently presumed. While most of these early accounts are largely qualitative, many report huge numbers of cormorants, suggesting that recent population increases may represent recovery towards historical (presettlement) levels in certain regions. In some areas where the DCCO has been documented as a recent breeder, the species is actually re-colonizing after an absence of 50 – 300 years.

Reasons for population increases: There appear to be five major factors that led to dramatic increases in DCCOs in North America since about 1970. These include:

1) Ban on DDT (1972) and other pesticide reduction regulation. Prior to this time (but post WWII) widespread use of DDT occurred. Cormorants accumulated high levels of DDT through their food supply, which interfered with reproduction. Depressed populations began to increase after DDT was banned.

2) In 1972 the DCCO was added to the Migratory Bird Treaty Act protected bird list. Before 1972, federal legislation did not prevent killing or harassment of cormorants during their annual cycle. Some states also provided special protection for DCCOs around this time.

3) Human induced changes (e.g. accidental and intentional introduction of exotics; over fishing; changes in water quality) in aquatic communities in the breeding range.

4) Development of aquaculture (e.g. catfish farms) in the south (especially Mississippi Delta region) that provided a new food source.

5) Creation of additional breeding and foraging habitat (e.g. reservoirs; dredge spoil islands).

Diet and native fish populations:
DCCO diet is characterized by great temporal and spatial variation. The DCCO is known to feed on > 250 species of fresh and saltwater fishes. Cormorants are generalists and eat abundant fish in the size range 3 – 40 cm; < 15 cm is preferred. Review of diet studies (> 40) indicates most sport and commercially valuable fish species do not contribute substantially to DCCO diet. Though there are exceptions, most studies conclude that sport and commercially valuable fish species are not negatively impacted by DCCOs, and that DCCOs have minor effects on fish populations relative to human harvest and other mortality factors. The most common claim against DCCOs is that they reduce sport or commercial catches, but the actual relationship between cormorant predation, fish population size and human harvest is poorly understood. This lack of information contributes to the complexity of cormorant-fish-fishery interactions.

Rigorous quantification of cormorant predation on fish populations or on subsequent sport or commercial catches requires more precise estimates of several key parameters, including: prey fish population sizes; prey fish mortality sources and rates; age class distribution of fish consumed. Additionally, a better understanding of compensatory processes within prey fish populations is essential (e.g., predation may reduce competition so that remaining fish survive longer or younger fish grow faster). However, no study conducted so far has obtained robust estimates for all of these parameters. Therefore, while DCCOs may cause fish populations to decline, none of the studies reviewed provided data rigorous enough to demonstrate that they do so. The effect of cormorant predation can be either compensatory (if the cormorants do not eat them, the same proportion may be removed by other factors) or additive (mortality due to cormorant predation is not replaced by another factor). However, investigators have rarely examined cormorant predation in the context of other mortality or limiting factors.

Because of great spatial variation in DCCO diet and unique complexities of individual aquatic ecosystems, DCCO predation impacts need to be assessed locally. To do this biologists need a more comprehensive understanding of local fish population dynamics and standardized methods for assessing cormorant diet.

Diet and aquaculture facilities:
Studies show DCCOs may eat large numbers of catfish locally and temporally. However, no study has quantified the economic impact on net harvest. Only one study has examined the issue of additive and compensatory mortality and concluded that mortality due to DCCO predation impacts were additive under certain circumstances, but insignificant in others.

Impacts on vegetation:
Most colonial waterbirds destroy vegetation at breeding and / or roosting sites to some extent, and cormorants cause some of the most dramatic change. Cormorants impact vegetation through deposition of guano (excrement) that kills underlying vegetation and eventually trees, and through nest building behavior when they strip leaves and small branches. In the short term these changes are of greatest concern if they affect rare plant communities or private property. From a long-term perspective these changes may be insignificant on an ecosystem scale. Few studies have been conducted to characterize and quantify vegetation change due to cormorant nesting and roosting habits.

Impacts on other bird species:
DCCOs are hypothesized to have two potential effects on other colonial waterbird species: competition for nest sites and habitat degradation. Direct interspecific competition for nests and nest sites may occur but has not been documented through careful study. Most impacts appear to occur indirectly through habitat degradation (e.g. defoliation, tree die-off). While there is some evidence that DCCOs may displace other species, no studies have clearly established DCCO impact on other birds at even a colony level scale.

Management options:
Humans have attempted to manage cormorant numbers in the western hemisphere for at least 400 years. Currently in the U.S. all lethal take requires permits from the USFWS, except at aquaculture facilities in those states under the 1998 Federal Depredation Order. Depredation permits can be obtained to prevent economic impacts or impacts to endangered, threatened or species of conservation concern. Non-lethal harassment of birds depredating or about to depredate does not require permits. To reduce cormorant impacts primarily to fisheries, aquaculture, vegetation and other colonial waterbirds, a large number of techniques has been developed or proposed. These techniques utilize lethal and non-lethal measures and may be used at local, regional or population levels. The effectiveness of these measures is difficult to assess because in many cases impacts have been poorly quantified.

Most techniques used at the local level are non-lethal. Lethal control may help reinforce local non-lethal control techniques. However, because cormorants are highly mobile, lethal control at the local level may be ineffective at decreasing local populations. Although economic effectiveness cannot be assessed for individual control techniques, some appear more effective than others; future research should focus on reducing the costs of the most promising techniques. Many techniques have been poorly investigated; therefore conclusions about their economic and numerical effectiveness may be premature. Because aquaculture ponds are high quality foraging sites (high fish densities; lack of escape cover), control of cormorants on the breeding grounds is unlikely to eliminate the need to practice local control. To make aquaculture ponds less desirable foraging sites, some form of control at the local level (e.g. exclosures, harassment) will likely still be needed. Previous efforts indicate that population control in general must be large scale and will require sizable human and capital inputs to be effective. Additionally, potential density dependent effects that compensate for control related mortality are poorly understood. Addressing these and numerous other uncertainties will enhance the development of a scientifically based, large-scale population control effort.

Finally, no control is a management option that is economically justified if the costs of control are greater than the losses associated with cormorant impacts.

Population Models:
Models have identified data gaps critical for understanding population dynamics and predicting control effectiveness; modeling is potentially a very strong tool for gaining insights into cormorant management. Prediction of future DCCO population trends and analysis of control methods is hampered by lack of age-specific data for this species. More effort needs to be put into obtaining data needed to strengthen model predictions, and increased effort should focus on predicting management outcomes and follow progress. Until better data are available, however, such modeling efforts should include rigorous sensitivity analyses to investigate uncertainties in parameters used and assumptions made in the model.

Current research and monitoring efforts:
Of 33 U.S. states and nine Canadian provinces to which surveys were sent, nine reported research in progress and 19 have monitoring programs. Research addresses: cormorant diet, bioenergetics, impacts to aquaculture, sport and commercial fisheries, foraging range and foraging behavior. Additional studies are attempting to determine effectiveness of harassment at day and night roosts, effectiveness of barriers at aquaculture ponds, and nutrient enrichment in aquatic and terrestrial habitats. A satellite telemetry study will determine migration patterns, breeding locations and winter movements of cormorants at catfish farms. All monitoring efforts are used to determine population distribution and trends.

Future research priorities:
The assessment identified many research needs. Highest priority studies on DCCOs fall within the following broad topics: (1) demography, (2) impacts on fisheries and aquaculture, (3) management techniques, (4) impacts on flora and fauna and (5) distribution.