Abstract
1. This report focuses on the Federally-Threatened Roseate Tern, with subsequent comments on the Federally-Threatened Piping Plover, as well as two species of breeding terns that are of State-concern (Common and Least), and the abundant seaducks (of five principal species) which are present in winter. Other reports provide Population Viability Analyses for the two Federally-listed species that address possible significance of the mortalities estimated here.
2. To estimate the numbers of Roseate Terns killed requires information on the numbers expected to be flying past the turbines at rotor height; the number of turbine-rotors encountered and the probability of collision. Input data included field measurements (from aerial surveys reported in the first DEIS and data subsequently obtained) of the numbers and altitudes of terns at the project site through the season (May - mid-September) and their behavior (traveling/fishing). Flight speeds were obtained from the literature. Geometrical modeling was used to estimate encounter rates with rotors in the course of crossings of the project area. Collision-probabilities were also estimated geometrically using, in part, the reported collisions at a windfarm in Zeebrugge, Belgium, where terns nest immediately beside turbines and the numbers killed have been noteworthy.
3. Two sets of mortality estimates have been prepared (see 8 & 9, below): expected values calculated from "simple averages" of the pertinent variabilities, and uncertainty estimates based on Monte Carlo and bootstrap methods.
4. The project area on Horseshoe Shoal (HSS) was less-used by terns for foraging that other parts of Nantucket Sound, or areas closer to the breeding colonies, and is not crossed b major travel routes. The numbers of terns present on HSS were very low during the nesting season with larger numbers present in May (before) and in August-September (after).
5. The great majority of the terns using the area (95%) flew below rotor-height, although the numbers at risk have not been established because of the constraints on observing set by weather conditions.
6. For terns traveling at rotor height, no preferred directions have been established: number of crossings was modeled geometrically (Bolker model). This model showed that, while crossing the proposed windfarm circle (the smallest circle that circumscribes the windfarm) at rotor-height, a tern would be expected to encounter, on average, one turbine in 2.3 crossings: this low number principally results from the turbines being widely-spaced. For terns that do encounter a turbine, there is only a small chance of collision; this is for two principal reasons the turbines have only three blades and the rotors have relatively long rotation-periods, and because birds generally avoid these moving objects as well as the stationary turbine towers. The "no avoidance" collision rate with the HSS turbines is estimated from the Band model to be 0.027.
7. From the findings reported from the Zeebrugge windfarm, an avoidance rate of 0.91 was derived for Common Terns: this value is inappropriate for application to HSS because the proximity of the nesting area affects behavior of the birds near the turbines. The avoidance rates used in estimating expected collisions at HSS were 0.953 and 0.983.
8. The expected mortality of Roseate Terns is 0.8 individuals per year. Four expected values estimated for collisions ranged from 0.3 to 2.3 Roseates/year, based on 2 values for the number of Roseates present and 2 values for probability of collision.
9. The uncertainty analysis incorporated the documented variability in 8 variables, using a combination of Monte Carlo and data-resampling methods. The results of 5000 simulations indicated a median mortality of 0.83 Roseates/year with a large uncertainty (5 to 95% probabilities: 0.01 to 8.2 Roseates/year). Sensitivity analysis compared 8 scenarios which explore the effects on the estimate of kill rate of the lack of knowledge in parameter values needed to obtain that estimate. This analysis shows that median is robust (i.e. insensitive) to parameter variations, but that the edges of the mortality distribution are very sensitive to assumptions about bird behavior near turbines, and to field count variance.
10. Other species considered briefly include two State-listed terns. The Common Tern is more numerous than the Roseate Tern; it uses the project area in generally similar ways and may incur 12 collision-mortalities/year. The Least Tern was seen so rarely in the project area that risk for this species is minimal.
11. The Federally Threatened Piping Plover nests around Nantucket Sound and may cross the project area in the course of migratory movements or during dispersal but the numbers, height and course of such flights are unknown, so that no precise collision estimates are possible. However, the number of plovers nesting from New England to Atlantic Canada are known. Using estimates from MassWildlife for numbers of annual crossings of the project area and available information on collision probability, plover collisions are estimated at far fewer than one per year.
12. The winter seaducks are important components of the avifauna. Local movements in Nantucket Sound generally occur below rotor height. Very recent evidence from operating windfarms off the shores of Denmark and Sweden shows that these waterbirds avoid turbines very effectively in the course of migratory flights, so that the risk to these birds at HSS is expected to be very small but no estimates of mortality have been prepared.