Pwrchd3u.gif (2091 bytes)

Landscape Changes and Colony Site Dynamics: How Gull-billed Terns Cope at the Sea’s Edge

R. Michael Erwin, Bill Williams, Bryan Watts,
Barry Truitt, Dan Stotts and Brian Eyler


Survival Threats to Gull-Billed Terns


Gull-billed Terns (Sterna nilotica) (hereafter, GBT) have experienced population declines in coastal Virginia and elsewhere along the Atlantic Coast, with numbers in Virginia dropping from 2000+ breeding individuals in the mid-1970s to between 600-800 the past three years (1993-96). Low breeding productivity has been suspected over the past decade (B. Truitt, pers. obs.). GBTs nest on two types of habitats along coasts - on barrier island sandy beaches, often associated with other terns and Black Skimmers (Rynchops niger) , and on shellpiles along the fringes of saltmarsh islands in coastal lagoons. Populations using these two habitats are subjected to quite different selective pressures as a result of physical and biological differences. Storm events (average of 38/yr in coastal Virginia) are much more dramatic in changing barrier island profiles than shellpiles. Also, predators, including gulls , raccoons and foxes, are more abundant (but often unpredictable in location) on barrier islands than on shellpile habitats. On the other hand, barrier islands often attract large numbers of other sand-nesting species which may be rare or absent on the size-limited shellpiles.

Here we report on one landscape question concerning the role of habitat stability on colony site fidelity (“colony turnover”) : how do physical (e.g. storms) vs. biological factors (predation, social attraction) influence the probability of colony sites being reoccupied between years?

Two competing hypotheses are tested:

1. Physical and/or predator disturbances on barrier islands result in higher colony site turnover between years than on shellpile habitats in marshes.

2. The attraction of larger numbers of both conspecifics and related species on barrier islands (relative to shellpiles) results in lower colony site turnover on barrier islands vs. shellpiles.  

Shellpile Habitat

Photo by Robert Baker


Barrier Island Habitat

Photo by Robert Baker


We investigated the nesting ecology of this species by taking a two-tier approach: the first was an extensive one in which we estimated the nesting populations of all GBT colonies at all locations in coastal Virginia from 1993 to 1996; the second was an intensive one (1994-96) in which from four to six colonies were monitored 1-3 times weekly for nest success. At each site, a sample of from 8 to 50 nests was marked and followed during the nesting period. Sites were chosen in both habitats, barrier island and shellpile; however, the number of barrier island sites monitored was limited to only 1-3 whereas 3-4 sites on shellpiles were monitored.

Colony site turnover (in fact, the percentage of colonies present in only one of any two years compared) was calculated after Erwin et al. 1981 (Auk 98). Turnover was compared for each of the consecutive years from 1993 to 1996 separately for barrier vs. shellpile habitats. A site was considered “occupied” by a colony if at least 5 nests were found. Also, a colony site was considered distinct if it was located a minimum of 200 m from another site.


Colony site turnover was high (>30%) in four of the six yearly comparisons made (Table 1). Of the total of 25 sites used over the four-year period, only three were used all four years. Turnover was higher on average on shellpile sites than on barrier islands (Table 1). Most colony sites (16 of 25) were used only one or two of the four years (Fig. 1).

Related to colony site shifts, we found flooding to be more frequent on shellpiles than on the barrier islands examined (Table 2). However, birds tended to renest in the same year at the same sites even after a complete washover.

Colony sizes of GBTs on barrier islands tended to be larger than those on shellpiles, although the variances and limited samples resulted in no significant differences (P>0.05) (Table 3).

Table 1. Colony Site Turnover by Gull-billed Terns (1993-1996)

Habitat 1993-94 1994-95 1995-96 Mean
Barrier islands 0.162 (6/4) 0.41 (4/7) 0.32 (7/5) 0.30
Shellpiles 0.71 (8/6) 0.72 (6/9) 0.22 (9/9) 0.55
1Turnover, t, expressed as percent colonies used only 1 of 2 years (Erwin et al. 1981 Auk). N sites in the 2 years (in parentheses)

Table 2. Flooding Effects: 1994-1995

Habitat N None Partial Complete
Barriers 7 6 1 0
Shellpiles 15 2 6 7

Table 3. Colony Sizes of Gull-billed Terns

(N Adults) 1993-1996

Habitat Mean S.D. Range
Barrier islands (n=9) 56 56 5-157
Shellpiles (n=16) 29 20 5-77
    t=1.41, P>0.1


Because the largest GBT colonies were associated with some of the largest Common Tern-Black Skimmer colonies in the state, we suggest that the second hypothesis is more likely supported than the first. Further, two of the three colony sites occupied all four years of the study were on these barrier islands. Although flooding probabilities were higher on shellpiles, birds tended to renest later in the season at the same sites rather than relocating. If flooding were the critical factor, one would expect large-scale shifts from shellpiles to barrier islands following flooding. Others (Burger and Gochfeld 1991) have suggested that, for skimmers, flooding does not often lead to colony relocation.

Previous nesting success is often the best indicator of habitat quality for a species. Our preliminary data suggest very poor success in all three years in both habitats. Often, less than half the nests that hatch young produce at least one chick. We do not feel this factor would lead to a habitat difference in fidelity.

Colony site turnover is higher in GBTs than for other related species. Further work is needed on the species in other parts of the coastal U.S.


GBT colonies nesting on shellpiles are frequently flooded during 'new moon' high tides. Colonies were flooded at least once each year during the three years of our study. To reduce flooding, we recommend augmentation using shells or sand to elevate the shellpiles a minimum of 15 cm. This procedure could be effective in areas as small as 250 square meters. In Virginia, at present, clam and oyster shells are used only for oyster reef restoration.

GBT colonies, especially those nesting on barrier islands, are also subject to disturbance by recreational boaters and beachcombers. More effective protection of critical nesting areas is required. Warning signs should be installed around colonies by 10 May of each year and left in place through the end of the nesting season (August or September). Areas with very high densities of recreationists may require the use of wardens to patrol the islands and prevent colony disturbances.

Predator populations have increased on the Virginia barrier islands over the past 25 years (B. Truitt, pers. obs.), causing many GBT colonies to relocate to other predator-free sites. Islands that currently maintain viable predator populations have no GBT colonies nesting on them. Predators also appear to be expanding into nearby areas that do contain GBT colonies. Removing predators (i.e. foxes, raccoons, and in some instances gulls) from the islands could restore valuable nesting habitat.


1. Gull-billed Terns in coastal Virginia exhibit colony site changes (between-years) that are similar to (barriers) or higher than (marsh shellpiles) those reported for similar species in the region. Only 3 of 25 colony sites were occupied in all 4 years of the study.

2. Contrary to expectation, colony-site turnover was higher on marsh shellpiles (0.55) than on barrier islands (0.30).

3. Flooding is more problematic on shellpiles, but mammalian predators are more abundant on barrier islands. Production was very low in all years.

4. Because of the presence of other nesting species (Common Terns, Black Skimmers), and more available habitat, colony size may be a stronger attractant on barrier islands than on shellpiles.

5. Long-term sea level rise could prove deleterious to a number of beach and marsh nesting species already “on the edge.”