In this Chapter information about bird predation and management is given, if available, for each dam, starting with Bonneville Dam and going upstream. First, however, the possible effect of weather or turbidity on bird foraging success is discussed.
During studies by the U. S. Army Corps of Engineers at Bonneville, The Dalles, and John Day Dams in 1995, Jones et al. (1996:28) observed that there did not appear to be any difference in the number of gulls present during rainy or dry weather. However, they found that the percentage of gull dives into water in which gulls surfaced with fish or with swallowing movements decreased from 42% on days without rain to 25% on days with rain (Jones et al. 1996:28).
At Wanapum Dam, Ruggerone (1986:737-738) found that gull foraging success significantly declined when the foraging area below the dam was in a shadow.
Differences in turbidity, river flows, and spill discharges may have affected the number of gulls present during the Corps' studies during 1995-1998 (Jones et al. 1998:19, 1999:10, 11, 13). Elsewhere, high turbidity has been reported to reduce predator foraging success or decrease juvenile salmonid survival vulnerability to predators (Cezilly 1992, Gregory and Levings 1998).
Bonneville Dam is located at River Mile 145.5, and many juvenile salmonids are transported around dams and released below Bonneville Dam at about River Mile 141-144 (Mundy et al. 1994:2, 15-16; National Marine Fisheries Service 2000c:1). The amount of bird predation at transportation release sites is not known, but, in the summer of 1986, one of the release sites became unusable because of predation by northern pikeminnows and gulls (Koski et al. 1987:14, Athearn 1991:341). In 1990, unspecified "predators" concentrated at a release site when water flows declined and water temperatures increased (Ceballos et al. 1991:12). In 1997, truck drivers used a firehose to keep gulls away while they released transported fish (U.S. Army Corps of Engineers 1998:5).
To try to reduce predation by birds and northern pikeminnows, barge releases are after dark, and the release site is varied, but truck releases are during daylight (Athearn 1991:341, U.S. Army Corps of Engineers 1998:5).
In early 1999, U.S. Army Corps of Engineers began using a $62 million, 2-mile pipe (flume) for juvenile salmonids to bypass the turbines of the second powerhouse at Bonneville Dam; the goal was to perhaps increase juvenile survival by 6-15% by reducing physical injury to fish from the old bypass and by reducing predation by northern pikeminnows (Brinckman 1999, Espenson 1999a). After it began operation, gulls were attracted to the outfall of the pipe and were estimated to take about 900-1,000 juvenile salmon on some days and as many as 2% of the passing fish (Brinckman 1999, Espenson 1999a). However, the accuracy of these estimates is not clear; for example, on a day when gulls were estimated to have taken 900 fish, about 250,000 were estimated to have gone through this bypass (Espenson 1999a)--900 is 0.4% of the passing fish.
To curb this predation, two water cannons (hydrocannons) that could spray water up to 155 ft (47.2 m) commenced operation at the outfall to disperse gulls on 30 April 1999, and wire lines with plastic flags were also placed at the outfall to discourage gulls (Brinckman 1999, Espenson 1999a). In May 1999, a spokeswoman for the Corps said that there were no longer any gulls present at the outfall (Brinckman 1999).
At Bonneville Dam during 1995-1998, California gulls were the most numerous bird, but other fish-eating birds included some western, glaucous- winged, herring, and Bonaparte's gulls; double-crested cormorants, great blue herons, common mergansers, osprey, western grebes, eared grebes, pied-billed grebes, American crows, and bald eagles were also noted (Jones et al. 1996:7- 9, 1998:11, 1999:6). In 1996, more than 10 cormorants were only seen three times, and there was a maximum of 13 herons, 11 mergansers, five osprey, and a few grebes during April-August; additionally, a few crows sometimes flew over the water and caught fish with their feet (Jones et al. 1996:8-9, 16-17, 20, 26). In 1998, California gulls were the only gull noted during April- July (Jones et al. 1999:6). In 1997 and 1998, Bonaparte's gulls were present in late October-November at Bonneville Dam and may have fed on juvenile shad (Jones et al. 1998:11, 1999:6).
Gull abundance was greatest during April-May when the number of juvenile salmonids passing Bonneville Dam was also greatest and was low after mid-July when the number of salmonids had decreased (Jones et al. 1996:13, 1997:6, 1998:13, 1999:9). Gulls were most numerous at the tailrace of the spillway, though some were also seen at the tailraces of the power houses (Table 5.1).
Gull abundance at the dam increased after nearby hatchery releases. The Spring Creek National Hatchery is about 22 miles (35 km) upstream of Bonneville Dam (Jones et al. 1996:12). In 1995, bird predation at Bonneville Dam increased 2-3 days after two releases from the hatchery (Jones et al. 1996:12-13).
Jones et al. (1996:3-5, 1997:3-5, 1998:4) used the frequency of gulls catching fish to calculate the total number of juvenile salmonids caught by gulls. They did not identify these fish and assumed that they were only juvenile salmonids; however, it is not clear if this assumption is true (sections 2-C and 2-D). In 1996, when the salmonid Passage Index was greatest, about 85,000 fish were estimated to have been caught by gulls, and the number declined in 1997-1998, when the Passage Index was less than half that of 1996 (Table 5.1).
Using the estimates of the number of fish caught by gulls and the number of juvenile salmonids passing Bonneville Power House 1, gulls caught about 1.2% or less of the salmonids that were passing (Table 5.2). However, this may overestimate gull predation because the Passage Index is only for those passing Power House 1, but salmonids also passed by the spillway and by the bypass of Power House 2 (section 5-C-2).
--------------------------------------------------------------------------TABLE 5.1. Estimated numbers of gulls and fish caught by gulls at Bonneville Dam tailraces during 1995-1998 from data in Jones et al. (1999:6, 12). There were 54, 41, 28, and 24 observation days, respectively, during 1995-1998, and the observation periods are from Jones et al. (1998:4, 1999:1, 9). Yr=year. Number of Fish Caught=estimated number of fish predated by gulls; Jones et al. assumed that these were all juvenile salmonids; the methods used for these estimates are discussed in Jones et al. (1996:3-5, 1997:4-5). N=number of gulls that Jones et al. (1997:3-4) "equalized" to be comparable because the number of observation days varied among years. Wires=1 mm thick stainless steel array of wires over water to deter flying gulls (Jones et al. 1997:1). PI=Passage Index, which is an estimate of the millions of juvenile salmonids at Bonneville Dam Power House 1; it is a sum of data in Fish Passage Center (1997:66, 1999:48).
----------------------------------------------------------------------------
Bonneville Dam___________________________________________
Power Power
House 1 House 2 Spillway
Observ- Tailrace___ Tailrace___ Tailrace___ Total Number
vation Gulls Gulls Gulls Gulls of Fish
Yr Period (N) Wires (N) Wires (N) Wires (N) Caught PI
---------------------------------------------------------------------------
95 4/11-8/3 262 No* 523 No* 713 No* 1,448 ?a 5.2
96 4/15-7/31 81 No** 371 No** 1,223 No** 1,676 84,694 7.5
97 4/15-7/30 19 Yes 12 Yes 346 Yes 377 11,114 3.3
98 4/13-7/31 31 Yes 58 Yes 583 Yes 672 35,966 3.0
a The method of estimating the number of fish taken by gulls differed
between 1995 and later years (Jones et al. 1997:8), so the 1995
number is not given.
* In 1995, there were a few, sagging dacron fishing lines suspended
25-30 ft (7.6-9.1 m) above the water: four at Power House 1, three at
Power House 2 that were 20 ft (6.1 m) apart, and six at the spillway
(Jones et al. 1996:5, 1997:8). These lines were ineffective because
they were not placed in areas where gulls foraged and because the lines
were often broken and were not consistently maintained (Jones et al.
1996:20, 23; 1997:8).
** In 1996, wires were installed on June 25 at Power House 1, on July 23
at Power House 2, and on September 11-12 at the Spillway tailrace
(Jones et al. 1997:8-9), so wires would not have deterred gulls
during most of the smolt migration season.
--------------------------------------------------------------------------
--------------------------------------------------------------------------
TABLE 5.2. Rough estimates of the proportion of juvenile salmonids passing Bonneville and John Day Dams that were estimated to have been taken by birds during daylight. Birds also took fish at night at John
Day Dam, but the number was not estimated (section 5-E-3). The numbers
of fish estimated to have been taken by birds at Bonneville Dam is from
Table 5.1 and at John Day Dam is from Table 5.4. It is assumed in this
Table that all fish estimated to be taken by birds were juvenile
salmonids, although evidence for this assumption is lacking (see sections
2-C and 2-D).
The Passage Index is only for juvenile salmonids and is a "relative indicator of population magnitude" (Fish Passage Center 1998:52), so it is a crude estimate of the number of fish passing Bonneville Power House 1 or John Day Dam. During the Observation Periods for birds (which are from Table 5.1 and Table 5.4), probably all steelhead, coho, sockeye, and yearling chinook passed these dams, but about 10% of the subyearling chinook may have passed after the Observation Period (Table 1.1). Accordingly, the Adjusted Passage Index is calculated by subtracting 10% of the estimated number of subyearling chinook from the Passage Index given for 1996-1998 in Fish Passage Center (1997:66, 1999:48).
For John Day Dam, only data for 1998 are calculated because the number of fish taken by birds in 1995 was not estimated, the observation period for estimating bird predation in 1996 ended on July 10 so many juvenile salmonids may have passed later (Fish Passage Center 1999:49), and, in 1997, an unknown, significant proportion of subyearling chinook passed after the Observation Period ended on July 30 (Fish Passage Center 1998:C-14).
---------------------------------------------------------------------------
Juvenile Salmonids______________________
Estimated
Number Adjusted
Bird of Fish Fish Proportion of
Observation Caught Passage Adjusted Index
Year Dam Period by Birds Index Caught by Birds
---------------------------------------------------------------------------
1996 Bonneville 4/15-7/31 84,694* 7,305,900 <1.2% *
1997 Bonneville 4/15-7/30 11,114* 2,804,398 <0.4% *
1998 Bonneville 4/13-7/31 35,966* 3,154,266 <1.1% *
1998 John Day 4/13-7/31 94,176 5,274,032 1.8%
* The Bonneville data may overestimate the proportion taken by birds
because the Passage Index is only for Power House 1, but the estimate
for bird predation includes both powerhouses and the spillway
(Table 5.1).
---------------------------------------------------------------------------
In 1995, a few dacron fishing lines hung over part of the tailraces; these lines were ineffective because they did not cover much area and were not maintained (footnote * in Table 5.1). In June-September 1996, stainless steel wire arrays were installed by Wildlife Services (Jones et al. 1997:8-9) and were "virtually 100% effective" in keeping gulls away during 1998-1999 (Jones et al. 1998:20, 1999:13). Consequently, gull abundance and the number of fish estimated to have been caught by gulls declined after the arrays were installed, but some of this decline may also have resulted from fewer juvenile salmonids being present (i.e., a lower Passage Index)(Table 5.1). However, not all areas at tailraces could be covered with wires, so that is why some gulls were still present at Bonneville tailraces and caught fish after wires were installed (Table 5.1).
Where possible, information for bird predation at The Dalles or John Day Dams is separated for each dam in the following sections. Bird management at these dams is discussed in section 5-F.
During 1995-1998, gulls were the most abundant fish-eating bird at The Dalles Dam, and California gulls were the most numerous gull; but some double-crested cormorants, great blue herons, western grebes, and common mergansers were also observed (Jones et al. 1996:7, 9; 1998:14, 1999:6). Only gulls were enumerated, and they were most abundant during late April-early July, with few seen later (Jones et al. 1998:15). Most gulls were recorded at the spillway tailrace, but some were also at the power house and ice/trash sluice tailraces (Table 5.3).
Illumination allowed some gulls to feed at night in the forebay of The Dalles Dam, but the number of fish taken then was not estimated (Jones et al. 1999:14).
--------------------------------------------------------------------------TABLE 5.3. Estimated numbers of gulls and fish caught by gulls at The Dalles Dam tailraces during 1995-1998 from data in Jones et al. (1999:10). The observation periods and number of observation days are from Jones et al. (1998:4, 1999:1). Number of Fish Predated=estimated number of fish predated by gulls; Jones et al. assumed that these were all juvenile salmonids; the methods used for these estimates are discussed in Jones et al. (1996:3-5, 1997:4-5). Yr=year. N=number of gulls that Jones et al. (1997:3-4) "equalized" to be comparable because the number of observation days varied among years. Wires=stainless steel array of wires over water to deter flying gulls. No Passage Index for the numbers of juvenile salmonids reaching The Dalles Dam is available (Fish Passage Center 1999:48).
---------------------------------------------------------------------------
The Dalles Dam__________________________________________
Power Ice/Trash
House Spillway Sluice
Tailrace___ Tailrace____ Tailrace____ Total Number
Observation__ Gulls Gulls Gulls Gulls of Fish
Yr Period Days (N) Wires (N) Wires (N) Wires (N) Caught
---------------------------------------------------------------------------
95 4/25-8/30 19 52 No 5,143 No 11 Yes 5,204 64,787a
96 4/16-7/10 29 201 No 2,034 Yes 12 Yes 2,247 ?a
97 4/15-7/30 31 17 No 473 Yes 0 Yes 490 ?a
98 4/13-7/31 ? 37 No 986 Y/N* 6 Yes 1,029 ?a
a The number of juvenile salmonids was estimated only in 1995 (Jones et
al. 1996:18), but in later years the wire array over the spillway
tailrace moved most gulls beyond where their predation success could be
studied, so the number of juveniles was not estimated (e.g., Jones et
al. 1999:13).
* Wires covering the spillway were installed in 1996 and remained the
same in 1997 (Jones et al. 1998:2), but, in 1998, some of the wires
covering the spillway were missing and were replaced by June 3 (Jones
et al. 1999:2).
--------------------------------------------------------------------------
In 1995, gulls and western grebes were the most abundant fish-eating bird predators, but some double-crested cormorants, common mergansers, ospreys, and Caspian terns were also observed (Jones et al. 1996:7, 11). Western grebes foraged almost exclusively in the forebay of the powerhouse, where an average of 16 grebes was counted during April-May and 6-9 grebes were noted during June-August (Jones et al. 1996:7, 11).
During 1997-1998, most gulls were California gulls, but some ring- billed gulls were also noted (Jones et al. 1998:16, 1999:6). Gull abundance was high from late April to early July when numbers of migrating juvenile salmonids at John Day Dam were also high (Jones et al. 1996:13, 1997:13, 1998:18, 1999:12). Most gulls were seen at the spillway and juvenile bypass tailraces, but some were also at the power house tailrace (Table 5.4). In fall and winter, some gulls were also occasionally present as Jones et al. (1997:15) noted that about 50-100 unidentified gulls were feeding (perhaps on juvenile shad, see section 2-B) in tailwaters of John Day Dam in November 1996, and some remained into December.
Lighting allowed gulls to feed at night in the John Day forebay during 1996-1998 and, in 1997, at the juvenile bypass system outfall (Jones et al. 1997:16, 1998:9, 17, 20; 1999:14).
--------------------------------------------------------------------------TABLE 5.4. Estimated numbers of gulls and fish caught by gulls at John Day Dam tailraces during 1995-1998 from data in Jones et al. (1999:11, 13). There were 18, 28, 31, and 22 observation days, respectively, during 1995-1998, and the observation periods are from Jones et al. (1998:4, 1999:1, 12). Yr=year. Number of Fish Caught=estimated number of fish predated by gulls; Jones et al. assumed that these were all juvenile salmonids; the methods used for these estimates are discussed in Jones et al. (1996:3-5, 1997:4-5). N=number of gulls that Jones et al. (1997:3-4) "equalized" to be comparable because the number of observation days varied among years. Wires=stainless steel array of wires over water to deter flying gulls. PI=Passage Index, which is an estimate of the millions of juvenile salmonids passing John Day Dam; it is a sum of data in Fish Passage Center (1997:66, 1999:48).
---------------------------------------------------------------------------
John Day Dam________________________________________________
Power Juvenile
House Spillway Bypass
Observ- Tailrace___ Tailrace____ Tailrace____ Total Number
ation Gulls Gulls Gulls Gulls of Fish
Yr Period (N) Wires (N) Wires (N) Wires (N) Caught PI
---------------------------------------------------------------------------
95 4/4-8/30 19 Yes 859 No 120 No* 1,164 ? 4.3
96 4/16-7/10 42 Yes 4 Yes 169 No* 215 5,196 3.0
97 4/15-7/30 19 Yes 118 No** 235 No* 372 22,772 1.5
98 4/13-7/31 11 Yes 1,467 No** 265 No* 1,743 94,176 5.5
* A pair of oscillating high-pressure hydrocannons was the only
deterrent for the juvenile bypass outfall (section 5-F-2, Jones et al.
1999:11).
** Contractors removed the spillway wire array in 1997, and it was not
replaced in 1998 because of a lack of funding (Jones et al.
1999:3, 11).
--------------------------------------------------------------------------
For radiotagged juvenile salmonids, Snelling et al. (1997) report that 1% of 154 yearling and subyearling chinook were taken by birds at the juvenile bypass outfall at John Day Dam in 1994. In 1995, 11% of 100 yearling and 6% of 75 subyearling chinook were taken by gulls at the Dalles Dam tailrace (Snelling et al. 1997). In 1996, 5% of 311 yearling and 3% of 219 subyearling chinook were accounted for by birds, though it is not specified if this was for one or both dams (Snelling et al. 1997).
It is not clear if at least some of these radiotagged fish were killed during dam passage and then were scavenged by gulls or other birds (section 2-F-2). Passage mortality may be greater for tagged than nontagged fish because the process of radiotagging can increase the vulnerability of fish to predation or other mortality (Adams et al. 1998, Hockersmith et al. 1999:45). Consequently, it is unclear if results by Snelling et al. are a good estimate of the amount of predation for nonradiotagged fish. Further, it is not known if control measures employed by Wildlife Services since 1996 (see section 5-F) may have reduced predation since the study by Snelling et al. (1997).
During 4 April-12 May 1997, 25% (N=14) of 56 gulls collected at The Dalles and John Day dams by Wildlife Services during its bird control activities (section 5-F-5) had fish in their stomachs, and all identifiable fish remains were juvenile salmonids; their nonfish diet was mostly mollusks (K. Collis, pers. comm.). Consequently, most (75%) gulls lethally controlled by Wildlife Services at these dams did not contain juvenile salmonids. For 24 California gulls collected during 4 April-12 May 1997, 66% of their diet by weight was salmonids; this was greater than for 25 California gulls collected at nesting colonies during 14 April-15 May 1997 for which salmonids were 43% of their diet (Roby et al. 1998:15, 38), so some gulls may have specialized on salmonid feeding at dams.
During the 1997 and 1998 nesting seasons, Collis et al. (2002:543) found that juvenile salmonids were 15% of the diet of California gulls at Little Memaloose and 3% of the diet of California gulls at Miller Rocks, but 4% and 7% of gull diet at these colonies, respectively, were of unidentified fish that may have included salmonids (see Colonies 1 and 2, respectively, in Fig. 1.1).
In 1997, Roby et al. (1998:34) did not find any fish in the diet of ring-billed gulls nesting at Miller Rocks, but the sample size was small (N=8).
Jones et al. (1996:3-5, 1997:3-5, 1998:4) used the frequency of gulls catching fish in calculating the number of fish caught by gulls. They did not identify these fish and assumed that they were only juvenile salmonids; however, it is not clear if this assumption is true (sections 2-C and 2-D). At The Dalles Dam, they estimated that 64,787 fish were taken in 1995, and they did not estimate the amount of predation in other years (Table 5.3 footnote a). Most of this predation was presumably in the spillway tailrace because that is where most gulls were observed rather than at the powerhouse or ice/trash sluice tailraces (Table 5.3). A Passage Index for juvenile salmonids at The Dalles Dam was not estimated (Fish Passage Center 1999:48).
At John Day Dam, 5-94 thousand fish were estimated to have been taken annually during 1996-1998 (Table 5.4). Most gulls were usually at the spillway tailrace, which is where most predation may have occurred (Table 5.4). The greatest number of fish caught by gulls occurred in 1998, which is also when the Passage Index for juvenile salmonids was greatest (Table 5.4). Although it was not possible to estimate the proportion of salmonids taken by gulls during 1995-1997 (see Table 5.2 legend), gulls may have caught about 1.8% of passing juvenile salmonids in 1998 (Table 5.2).
Wildlife Services was contracted (presumably by the U.S. Army Corps of Engineers, who operated these dams) to control birds at The Dalles and John Day Dams (Jones et al. 1998:8, 1999:5). Washington Wildlife Services declined providing any information about its bird control in the Columbia Basin (section 4-B), so the following information is from Jones et al. (1996-1999) of the U.S. Army Corps of Engineers.
Although it is not specified who installed the stainless steel wire arrays at these dams (e.g., see Jones et al. 1997:11, 15), Wildlife Services appeared to be responsible for maintaining these arrays to exclude birds (Jones et al. 1997:17, 1999:2). These wires were effective in keeping gulls away (Jones et al. 1997:16, 1998:20, 1999:13), since gull abundance was low at powerhouse and spillway tailraces with wires and declined at areas after wires were installed (Table 5.3 and Table 5.4). But not all areas at tailraces could be covered with wires, so some gulls remained (Table 5.3 and Table 5.4).
At the John Day Juvenile Bypass System outfall, an avian hydrocannon was also operating by the middle of February 1997 to keep birds away (Jones et al. 1998:2). The avian hydrocannon was composed of two irrigation type impulse sprinklers that under favorable conditions swept a 135 ft (41.1 m) radius, but prevailing winds often reduced the reach to a 105 ft (32.0 m) radius (Jones et al. 1998:2, 7). It excluded gulls within the reach of its spray, but many gulls fed beyond its reach (Jones et al. 1998:7-8, 1999:11, 13).
In 1997, Wildlife Services directed a spotlight at night-time feeding birds at John Day Dam (Jones et al. 1998:9), but it apparently was unsuccessful as Jones et al. (1999:14) continued to recommend that lighting in the forebays of The Dalles and John Day Dams be eliminated to limit night feeding.
Wildlife Services also used pyrotechnic devices with disturbing sounds such as cracker and screamer shells or a propane cannon to disperse birds (Jones et al. 1996:20-22, 24; 1998:8-9, 1999:5).
In 1997, Wildlife Services dispersed 2,679 birds at The Dalles Dam-- about 97% were gulls; at John Day Dam, it hazed 5,797 birds, of which about 80% were gulls (Table 5.5). In 1998, many more gulls were dispersed at both dams, but the number of birds that were not gulls was not specified (Table 5.5). The number of gulls hazed at these two dams in these two years was a small proportion of the total for Washington during 1997-1998 (see Table 4.2).
Nonlethal hazing did not keep all birds away, though it may have reduced predation. In 1995, some birds quickly became accustomed to the propane cannon and ceased responding to it; for all pyrotechnic devices, birds were absent from an area after hazing an average of 1.6 minutes, and all birds only left the area 17% of the time after hazing (Jones et al. 1996:21-22, 24). In 1997, 32% of birds had returned within 15 minutes of being dispersed by pyrotechnic devices, but hazing appeared to have a longer lasting effect in 1998 as less than 20% had returned within 50 minutes (Table 5.6). Elsewhere, hazing also can reduce but not eliminate predation (section 3-B-2).
The Wildlife Services' hazing was evidently extended throughout the year, as Jones et al. (1999:14) recommended that Wildlife Services' efforts
"be concentrated during the month of May to protect salmonids and relaxed or even discontinued during the winter months when gulls, overwintering nonresident grebes and other avian piscivorous species are likely feeding upon juvenile shad."
--------------------------------------------------------------------------TABLE 5.5. Number of birds dispersed or taken by Wildlife Services at The Dalles or John Day Dams in 1997-1998. This Table is from Wildlife Service data given in Jones et al. (1998:9, 1999:5); Washington Wildlife Services would not provide any information (section 4-B). Since some birds that were not gulls were dispersed and Washington Wildlife Services lethally controlled fish-eating birds of many species (Table 4.3), some birds that were not gulls may have also been controlled lethally at dams.
+=at least the indicated number of birds. In 1997, the number of gulls or nongulls that were controlled lethally was not differentiated. In 1998, more birds may have been controlled because the number of birds other than gulls that were dispersed or taken was not specified and the number of gulls controlled was only during the 1998 study period (April 13-July 31) of Jones et al. (1999:1, 5), but Wildlife Services may also have controlled birds outside of this time period (section 5-F-4).
--------------------------------------------------------------------------
Bird Control by Wildlife Services__________________
Removed by
Dispersed______________ Lethal Control_________
Other________ Other________
Dam Year Gulls Birds Total Gulls Birds Total
--------------------------------------------------------------------------
The Dalles 1997 2,599 80 2,679 ? ? 346*
1998 4,405 ? 4,405+ 1,277 ? 1,277+
John Day 1997 4,638 1,159 5,797 ? ? 269*
1998 7,801 ? 7,801+ 1,251 ? 1,251+
Total 19,443 1,239+ 20,682+ 2,528+ ? 3,143+
* In 1997, the total number of birds taken was reported, but it is not
stated if the total only includes gulls or may include other birds
(Jones et al. 1998:9).
--------------------------------------------------------------------------
--------------------------------------------------------------------------
TABLE 5.6. Average percentage of the initial number of gulls in an area that were present up to 60 minutes after pyrotechnic hazing or lethal control by Wildlife Service personnel at The Dalles and John Day Dams in
1997 and 1998. These data are from Jones et al. (1998:8, 1999:5); Jones
et al. (1999:5) also included data at 5 minute intervals after 20
minutes. For 1997, Jones et al. (1998:8) also give data for 20-30
minutes after a disturbance, but these data are not comparable to data
for 5-15 minutes because the initial number of gulls during these
intervals is not the same, so these data are not included. -=data not
available or not comparable.
--------------------------------------------------------------------------
Average % of Initial
Number of Gulls Present___________________________
Interval 1997___________________ 1998___________________
After Pyrotechnic Lethal Pyrotechnic Lethal
Disturbance Hazing Only Control Hazing Only Control
------------------------------------------------------------------
5 min 22 0 2 <1
10 min 27 2 0 1
15 min 32 11 0 1
20 min - - 3 4
30 min - - 9 9
40 min - - 18 15
50 min - - 15 16
60 min - - 37 20
--------------------------------------------------------------------------
Wildlife Services was given a permit to lethally control birds at dams by the U.S. Fish and Wildlife Service (Jones et al. 1998:5). A total of at least 3,143 birds was taken at these two dams in 1997 and 1998 (Table 5.5). The Wildlife Service's lethal control of gulls by shooting did not keep gulls away because a third of a flock of 19 gulls returned within 18 minutes of one being shot in 1995 (Jones et al. 1996:21). In 1997, 11% of the original number of gulls were present 15 minutes after one was shot (Table 5.6). In 1998, lethal control effects were longer lasting as only 1% of birds were present after 15 minutes, but 20% of the original number were present an hour after one was shot (Table 5.6). Thus, lethal control did reduce the number of gulls present for a while, but it is unknown if this resulted in a reduction of predation or if gulls foraged more efficiently after they returned. Elsewhere, lethal control of fish-eating birds also did not eliminate predation (section 3-C-1).
Evidence that lethal control may not reduce predation as much as it may be assumed is that 75% of the gulls collected in 1997 by Wildlife Services while conducting bird control at these dams and whose stomach contents were examined had not eaten juvenile salmonids (section 5-E-5).
In 2002, researchers in Columbia Bird Research (2002) noted cursorily reported western grebes in April and gulls, Caspian terns, and pelicans during April-July in the tailrace of McNary Dam. Gulls and terns were observed taking fish that may have been juvenile salmonids. When a final report about this research is completed, then it will be possible to know more about the abundance and foraging of these birds.
Merrell (1959) writes:
"In May 1955 a large concentration of gulls was observed feeding below McNary Dam on the Columbia River. Since a heavy migration of downstream-moving salmon and steelhead occurs during this season, it was assumed that gulls were feeding on these fingerlings."
Merrell reports that 27 California gulls, 11 ring-billed gulls, two immature western gulls, and one Forster's tern were shot during 11-12 May 1955. The gulls' stomach contents indicated that their principal prey were 158 lampreys 2-8 inches long; there were only five recognizable salmonids, which would be about 3% of the identified prey. The lamprey found in gull stomachs included 101 Pacific lamprey ammocoetes and 57 adult western brook lamprey. He suggests that more of the unidentified fish bones in the gull stomachs may have been salmonids or that the gulls may have regurgitated salmonids after they were shot but before they were collected. He did not report the stomach contents of the Forster's tern.
It is not clear how juvenile Pacific lamprey abundance now compares to 1955, though brook lamprey abundance now appears to be reduced (Kostow 2002:36). Adult Pacific lamprey numbers have fluctuated greatly at McNary Dam and were also low during 1954-1957 (Close et al. 1995:9, Kostow 2002:40), but there are no counts of juvenile Pacific lamprey or adult brook lamprey then and now. If lamprey are not as abundant now as in 1955 or if there are many more hatchery salmonids vulnerable to predation now (see section 3-B-3), then the diet of gulls feeding below McNary Dam may now be different than found by Merrell.
In 1997 and 1998, juvenile salmonids were not found in the diet of ring-billed and California gulls nesting at Three Mile Canyon (Roby et al. 1998:34, Collis et al. 2000:60; 2002:543), where about 11-13 thousand gulls nested during 1997-1998 (Table 1.2). Only 2% of the diet of these California gulls were nonsalmonid fish (Collis et al. 2002:543).
In 1998 and 2000, juvenile salmonids were about 81-86% of the diet of Caspian terns nesting at Three Mile Canyon (Table 5.7), where 520 nested in 2000 and none nested in 2001 (Table 1.3). Many Passive Integrated Transponders that had been inserted into juvenile salmonids were also found at this colony (Collis et al. 2000:25). Thus, salmonids were an important part of the diet of terns at this colony.
--------------------------------------------------------------------------TABLE 5.7. Percentage of juvenile salmonids in the diet of Caspian terns nesting at Three Mile Canyon Island near McNary Dam, Crescent Island near Ice Harbor Dam, and at Solstice Island. Diets were determined by nonlethally observing fish carried by adult terns to their nest. Colony locations are shown in Fig. 1.1. These data are from Collis et al. (2000:56; 2001b:10-11; 2002:542) and Columbia Bird Research (2000a:Diet Composition). N=total number of prey items observed; Juv. Sal.=proportion of total prey items that were juvenile salmonids.
---------------------------------------------------------------------------
Caspian Tern Diet__________________
1998______ 2000______ 2001_____
Colony Juv. Juv. Juv.
No. Colony N Sal. N Sal. N Sal.
---------------------------------------------------------------------
3 Three Mile Canyon Island 60a 81%a 331 86% 0b -b
4 Crescent Island 0 - 846 75% 2,189 68%
8 Solstice Island 0 - 0 - 255 30%
a In 1998, 26 adult terns were also lethally collected for their stomach
contents; 89% of the 26 prey items and 85% of the prey biomass were
juvenile salmonids (Collis et al. 2000:56).
b Three Mile Canyon was abandoned by terns in 2001 (Collis et al.
2001b:6).
---------------------------------------------------------------------------
In 2002, Tiller and Welch (2002) and Columbia Bird Research (2002) observed white pelicans (which are listed by Washington State as Endangered, National Marine Fisheries Service 2002:3-142, 3-146) feeding on unknown prey in the McNary Dam tailrace. The Columbia Bird Research (2002) report was preliminary.
Pelicans were first consistently observed in the tailrace on April 21 with a maximum count of 24 in late May; then numbers declined slightly but later peaked again during the first week of July (Tiller and Welch 2002). The first peak of pelicans occurred during the migration of most juvenile salmonid species at McNary Dam, but the second peak occurred only when mostly subyearling chinook were migrating (Table 1.1).
Although neither Tiller and Welch (2002) nor Columbia Bird Research (2002) documented foraging of pelicans on juvenile salmonids at McNary Dam, Ryan et al. (2002) reported that tags for juvenile salmonids marked with Passive Integrated Transponders were found at an unspecified Columbia Basin pelican nesting colony. So it is possible that pelicans may feed on juvenile salmonids at McNary Dam.
Wildlife Services has been noted as controlling gulls at McNary Dam (which is operated by the U.S. Army Corps of Engineers, Fig. 1.1), but its specific activities have not been reported (section 4-D-1). In early 2002, the Corps installed a water cannon (a 5/8 inch water nozzle) to discourage fish-eating birds from foraging below McNary Dam (Call 2002).
In 2002, Tiller and Welch (2002) observed that bird deterrent practices ("primarily water cannon and firearms for gulls") initially altered pelican behavior but that pelicans rapidly adapted to them and that the amount of time spent by pelicans in the tailrace was more than for other fish-eating birds. In the preliminary Columbia Bird Research (2002) report, changes in bird numbers with hazing in 2002 are also sketched at McNary Dam, but a final report by these researchers is required before making conclusions.
York et al. (2000:216, 219) of Wildlife Services' National Wildlife Research Center (http://www.aphis.usda.gov/ws/nwrc) studied gull predation below Priest Rapids Dam and at Cabin Island (Colony 7 in Fig. 1.1). Cabin Island is where 7,200 ring-billed and California gulls nested in 1995 (Table 1.2) and is 0.9 mile (1.5 km) north of the dam (Pochop et al. 1998:411). They collected gulls to determine stomach contents at both locations (Tables 5.8 and 5.9) and assumed that any fish found in gulls were juvenile salmonids because most fish predation occurred during peak salmon smolt migration in May. They found that ring-billed gulls at the Cabin Island colony had consumed few fish, but that unidentified fish were an important food item for ring-billeds collected below Priest River Dam in May (Table 5.8). In contrast, California gulls collected at the nesting colony took more fish, although fish were only a significant food item on May 25 (Table 5.9). For California gulls collected below the dam, unidentified fish formed 65-85% of their diet during the two days when they were collected in May (Table 5.9).
---------------------------------------------------------------------------TABLE 5.8. Diet (percent by volume) of ring-billed gulls at their Cabin Island colony (Colony 7 in Fig. 1.1) or below Priest Rapids Dam. All data are from York et al. (2000:Tables 1 and 2). Gulls Sampled=number of gulls collected to determine their diet, t=trace, Unidentified Fish=fish not identified to taxon, Misc. Debris=unidentifiable or nonfood items.
---------------------------------------------------------------------------
Ring-billed Gull Diet (% volume) at______________
Cabin Island Colony Below Priest Rapids
in 1995______________ Dam in 1996
Prey 4/19 5/5 5/25 6/13 5/7 5/21
---------------------------------------------------------------------
Unidentified Fish 0 4 t 0 26 41
Grain t t 13 4 26 t
Insect t 1 15 47 15 2
Mammal 1 49 45 1 t 0
Earthworm 0 6 0 1 0 0
Plant Matter 87 35 25 35 31 55
Misc. Debris 12 5 2 13 3 2
SUM 100 100 100 101 101 100
Gulls Sampled 16 18 22 21 21 22
---------------------------------------------------------------------------
---------------------------------------------------------------------------
TABLE 5.9. Diet (percent by volume) of California gulls at their Cabin Island colony (Colony 7 in Fig. 1.1) or below Priest Rapids Dam. All data are from York et al. (2000:Tables 1 and 2). Gulls Sampled=number of gulls collected to determine their diet, t=trace, -=no data, Unidentified Fish=fish not identified to taxon, Misc. Debris=unidentifiable or nonfood
items.
---------------------------------------------------------------------------
California Gull Diet (% volume) at_______________
Cabin Island Colony Below Priest Rapids
in 1995______________ Dam in 1996
Prey 4/19 5/5 5/25 6/13 5/7 5/21
----------------------------------------------------------------------
Unidentified Fish t 13 54 - 65 85
Grain 1 0 0 - 0 0
Insect 0 0 t - 2 t
Mammal 0 0 0 - 1 0
Earthworm 0 0 0 - 1 0
Plant Matter 82 41 44 - 21 15
Misc. Debris 17 46 2 - 10 t
SUM 100 100 100 - 100 100
Gulls Sampled 4 5 13 0 17 6
---------------------------------------------------------------------------
Priest Rapids Dam is operated by the Grant County Public Utility District (Fig. 1.1), and gull control by Wildlife Services below Priest Rapids Dam included exclusion with wire grids, hazing, and lethal control (sections 4-D-1 and 4-D-2).
At Cabin Island, staff of the Animal and Plant Health Inspection Service/Wildlife Services and the National Wildlife Research Center (which is now part of Wildlife Services, http://www.aphis.usda.gov/ws/nwrc) used white mineral oil or corn oil to spray eggs at about 3,190 ring-billed gull nests in 1995 "to control the fecundity of ring-billed gulls on the island" and 95% of oiled eggs did not hatch (Pochop et al. 1998:412). In 1996 and 1997, all eggs at this colony were sprayed with corn oil, and 99.3-99.5% of oiled eggs did not hatch (Pochop et al. 1998:412). Egg oiling may have continued since then because Wildlife Services (2002a) notes that it may use egg addling/destruction for controlling bird damage at mid-Columbia dams in Washington (see quotation in section 4-D-2). However, the number of eggs that are oiled are not included in the number of ring-billed gulls taken in Washington in Wildlife Services' tables (see Table 4.3). Pochop et al. (1998) do not state why it was necessary to reduce gull fecundity; perhaps it was because there was a concern that these nesting gulls were eating juvenile salmonids. They also did not indicate who owns Cabin Island or if Wildlife Services was requested to control nesting gulls there by the landowner.
Ruggerone (1986:736-737) estimated juvenile salmonid mortality by using binoculars to study gulls that were "almost exclusively ring-billed gulls" below Wanapum Dam (which is operated by Grant County Public Utility District, Fig. 1.1) during only 250 minutes (about four hours) of observations from 24 April to 15 May 1982. He did not identify any of the fish taken by gulls but assumed that "most of them probably were salmonids because other juvenile fishes in this area of the Columbia River do not actively migrate downstream in the spring" (Ruggerone 1986:737). He extrapolated his findings to indicate that gulls caught 112,000-119,000 presumed salmonids during his observation period or about 2% of the estimated salmonid migration (Ruggerone 1986:740-741). Ruggerone (1986:741) noted that:
"Although gulls consume large numbers of salmonids each spring, the total effect of gull foraging on the salmonid population is difficult to estimate because a portion of the salmonids consumed by gulls were likely killed by the turbines prior to being consumed."
Ruggerone (1986:741) suggested using monofilament line strung across the tailrace of Wanapum Dam to reduce gull predation. He also suggested that hatchery managers should release fish at night to reduce predation at release sites, so gulls may have been attracted to daytime hatchery releases.
Recently, Wildlife Services has been noted as managing birds at Wanapum Dam by using overhead wires, hazing, and lethal control (sections 4-D-1 and 4-D-2), but some wires may be too far apart to be effective (section 4-D-2).
The Chelan County [Washington] Public Utility District (PUD) operates and owns both of these dams (Fig. 1.1, National Marine Fisheries Service 2002:3-148, 3-149).
BioAnalysts (2000:19) states that fish and bird predation "probably" results in the bulk of the loss of juvenile salmonids through the Rocky Reach Reservoir, but does not provide evidence for this statement. Fish- eating birds present in the Rocky Reach area include gulls, cormorants, Caspian terns, great blue herons, osprey, common mergansers, belted kingfishers, common loons, western grebes, black-crowned night-herons, and bald eagles (BioAnalysts 2000:15).
This PUD has minutes for some of its Board of Commissioners' meetings available on the Internet. These minutes indicate that in 2000, this PUD agreed to pay $160,715 to Wildlife Services to use hazing techniques to scare away gulls and cormorants and to hire fishermen to catch northern pikeminnows at both dams (Gillin 2000). In 2001, the Chelan PUD contracted with Wildlife Services to pay no more than $22,550 to install a wire grid at both dams to deter gulls and terns (Craig 2001a) and $202,900 "to control northern pikeminnow, gulls and cormorants" at both dams (Craig 2001b). In 2002, this PUD agreed to pay $320,295 to Wildlife Services for "animal damage control" at both dams (Gillin 2002). These activities were presumably paid for by this PUD's ratepayers.
In 2002, minutes for the March 11 meeting of this PUD's Board of Commissioners indicated that they would also fund studies of predation (Craig 2002):
"Commissioners approved agreements with the state Department of Fish and Wildlife and the University of Washington for studies to validate the cost-effectiveness of the District's predator control programs to protect juvenile fish at Rocky Reach and Rock Island dams.
"Last week, HCP Implementation Manager Tracy Yount discussed how the PUD has contracted with the U.S. Department of Agriculture for several years to control fish and birds that prey on juvenile salmon at the dams. Contract crews remove northern pikeminnows downstream of the two projects and use various measures to deter or remove cormorants, gulls and terns.
"Similar actions are taken to protect fish in other areas of the Northwest. Although the actions are designed to help protect an endangered species, animal rights groups have raised questions about possible impacts on migratory birds, mostly terns. While Chelan County PUD's actions have not been questioned, the Department of Agriculture recently asked the PUD for help in preparing an environmental assessment on avian predator control.
"The PUD performed a risk analysis that looked at what might happen if the predator control program were abandoned. Yount said the preliminary conclusion was that the program is very cost- effective. But more detailed studies are needed to confirm that.
"The contract with the state is not to exceed $400,000 and the University of Washington agreement is not to exceed $330,000 in the first year. The studies are expected to take three years."
Although details are not given, the minutes for these PUD meetings indicate that the fish-eating birds of most concern at these dams are gulls, cormorants, and terns and that control measures for them include exclusion with wires, hazing, and lethal control. Although the cost of bird control is not separated from that of northern pikeminnow control, predator control and studies of predator control are expensive. For 2002 alone, expenses may total about $1 million.
The degree that this bird control has reduced predation is acknowledged to be unknown (T. West, Chelan PUD in BioAnalysts 2000:15). Some wire grids that Wildlife Services placed at mid-Columbia dams are spaced far apart that they would not be expected to be as effective as more closely spaced wires (section 4-D-2).
The Animal and Plant Health Inspection Service (APHIS; of which Wildlife Services is a part) is cited by National Marine Fisheries Service (2002) as having provided information about bird control at Rocky Reach and Rock Island Dams (section 4-B-2). However, "quantitative data are not available on salmonid mortality by avian predators at the Wells Dam, Rocky Reach Dam, Rock Island Dam, or Wells Hatchery" (National Marine Fisheries Service 2002:3-138). The most numerous predators at dams were reported to be ring-billed gulls, California gulls, and Caspian terns (National Marine Fisheries Service 2002:3-138).
Chelan County PUD is listed as funding bird control, which included wires installed across tailraces, hazing by propane cannons, "other pyrotechnic" methods, and lethal control where deemed necessary (National Marine Fisheries Service 2002:2-29). A diagram of the wire grid at Rocky Reach Dam is included in National Marine Fisheries Service (2002:3-141). At Rock Island Dam "between 1996 and 2001," hazing levels for just ring- billed gulls ranged from 450 to 3,371 per year, and 201-1,075 ring-billed gulls were killed annually (National Marine Fisheries Service 2002:3-140).
Wildlife Services has acknowledged that it has installed bird wires and conducted nonlethal and lethal control at these and other mid- Columbia dams (sections 4-D-1 and 4-D-2). But some of Wildlife Services' overhead wires may be too far apart to be effective in excluding flying birds (section 4-D-2).
Wells Dam is operated by the Douglas County PUD (Fig. 1.1, National Marine Fisheries Service 2002:3-147). Wildlife Services conducted bird control at this dam (sections 4-D-1 and 4-D-2), although estimates of bird predation are not available (National Marine Fisheries Service 2002:3-138). The fish-eating predators of most concern were ring-billed gulls, California gulls, and Caspian terns (National Marine Fisheries Service 2002:3-138). In addition, over 100 white pelicans regularly used the Wells reservoir from May to early October (National Marine Fisheries Service 2002:3-146), and white pelicans have been implicated as predators of juvenile salmonids (section 5-G-4).
Douglas County PUD is listed as funding bird control at Wells Dam (National Marine Fisheries Service 2002:2-29, 3-138), and Steuber et al. (1995) report that Wildlife Services was contacted by this PUD in 1992 for "alleviating a historical gull predation problem." In 1993, Wildlife Services and PUD personnel installed an overhead wire exclusion system over much of the area below the dam (Steuber et al. 1995). There was a dramatic decrease in the number of feeding gulls, especially when wires were placed at 7.5 m (24.6 ft) rather than 15 m (49.2 ft) intervals (Steuber et al. 1995); this is also discussed in section 4-D-2. Other bird control methods at Wells Dam included hazing by propane cannons, "other pyrotechnic" methods, and lethal control when considered necessary (National Marine Fisheries Service 2002:2-29).
During 1997-1998, Collis et al. (2002:543) reported that no fish were found in the diet of ring-billed gulls nesting at Island 18 (see Colony 6 in Fig. 1.1) and that only 3% of the diet of California gulls nesting there were fish, with no identifiable salmonids. In 1997, no salmonids were found in the diet of California or ring-billed gulls nesting at Crescent Island (see Colony 4 in Fig. 1.1) and Richland Island (see Colony 5 in Fig. 1.1) (Roby et al. 1998:34).
1,440 Caspian terns nested at Crescent Island in 2001 (Table 1.3). During 2000-2001, juvenile salmonids were about 68-75% of the diet of terns at this colony (Table 5.7). Many Passive Integrated Transponders that had been inserted into juvenile salmonids were found at this colony (Collis et al. 2000:25, 85-86). Thus, salmonids were a significant food item for these nesting terns, although it is not clear if terns caught salmonids at Ice Harbor Dam or in the Columbia or Snake Rivers after hatchery releases (section 6-B-2).
In 2002, as many as 40 nesting adult double-crested cormorants were to be collected to determine their diet at Foundation Island (which is not shown in Fig. 1.1), just below the confluence of the Snake and Columbia Rivers, where "hundreds of pairs" nested in 2001 (Bonneville Power Administration 2002:3). There may also be other cormorant colonies in the Columbia Basin (section 1-E-4).
Wildlife Services has been noted as controlling gulls at Ice Harbor Dam (which is operated by the U.S. Army Corps of Engineers, Fig. 1.1)(section 4-D-1).
In 1999, Pochop et al. (2001) of Wildlife Services installed silt fencing to discourage ring-billed and California gull nesting at "Upper Nelson Island," where they estimated that there were more than 21,000 gull nests. They found that the zone with fencing had 84% fewer nests than an area without silt fencing. I could not find "Upper Nelson Island" on the topographical map available at http://www.topozone.com, and the TopoZone map indicates that the latitude and longitude Pochop et al. (2001) gave for this island (46 22'50"N, 119 15'05"W) does not match their description that it is 0.3 mile (0.5 km) from the shoreline of Richland, Washington. However, the TopoZone topographical map indicates that Nelson Island is at River Mile 340 (River Kilometer 547), about 0.1 mile (0.2 km) from the Richland shore; an unnamed island just north of Nelson Island is about 0.3 mile (0.5 km) from the Richland shore and could be described as also being at River Mile 340, so maybe it is the island that Pochop et al. refer to as "Upper Nelson Island." Collis et al. (2002:539) searched for nesting gulls to River Mile 343 (River Kilometer 553) during 1997-1998 and did not report any gull nesting on Nelson Island, but they reported that 22,348 gulls nested (which would be 11,174 gull nests) on Richland Island at River Kilometer 547, so Pochop et al.'s "Upper Nelson Island" may be the same as Collis et al.'s Richland Island (see Colony 5 in Fig. 1.1).
Wildlife Services conducted bird control at Ice Harbor Dam (section 4-D-1), but I have not found any specific information about bird predation or control at this dam.
These dams are operated by the U.S. Army Corps of Engineers (Fig. 1.1). According to York in section 4-D-1 and Wildlife Services (2001a:2), Wildlife Services conducts gull control at these dams, though the types of control for each dam is not specified. I have not been able to find more information about bird predation or control at these dams, and Wildlife Services has refused to answer general questions about its control programs (section 4-B-2).
Brownlee Dam is operated by the Idaho Power Company (Fig. 1.1). The only information I have found about bird predation there is for 1959. During large spills in late July-early August and mid-October 1959, "large numbers of seagulls were noted daily" below Brownlee Dam (Haas 1965:47, 50). Haas (1965:50) wrote: "The gulls apparently accumulated below the dam to feed on fish that were killed, injured, or stunned by passing through the turbines or spillway." Idaho Fish and Game staff obtained permits to shoot some of the gulls in 1959 to study their food habits and found that "twenty-seven of 28 gulls shot had at least one identifiable salmon in their stomachs"; no other details are given (Haas 1965:50).
The Columbia River is over 30 miles (48 km) from the Caspian tern colony at Solstice Island in Potholes Reservoir (see Colony 8 in Fig. 1.1)(Collis et al. 2001b:11). Nevertheless, many juvenile salmonids marked with Passive Integrated Transponder, radio, or acoustic tags in 2000 and 2001 that had been released into the Columbia or Snake Rivers were recovered at this colony (Columbia Bird Research 2000a:Diet Composition, Collis et al. 2001b:11). In 2001, 30% of the prey items for Caspian terns nesting at this colony were juvenile salmonids (Table 5.7). It is unclear if terns caught these marked salmonids at dams or in rivers after they were released (section 6-B-2).
Wildlife Services has controlled birds at dams (section 4-D-1) and some bird colonies (sections 5-H-2 and 5-L-3), but it has not been reported if it has done anything at this colony.
A thorough review of bird predation and control at all Columbia Basin dams is not possible because of a lack of accessible information. But an index to known bird predation at dams is in Table 6.1, an index about fish-eating birds in the Columbia Basin away from dams is in Table 6.2, and an index of bird control is in Table 6.3. In the rest of this Chapter, aspects of bird predation and control are examined.
Fish-eating birds that are most often identified at dams are gulls, Caspian terns, and cormorants (Table 6.1). Although some predation has been reported in forebays in front of dams, most predation has been noted in tailraces or outfalls of powerhouses, bypasses, and spillways (Table 6.1). Most predation has been observed during daylight, but at some dams where forebays are illuminated, night-time bird predation has also been reported (Table 6.1).
In some collections of birds at dams, juvenile salmonids or other prey were not identified or were identified only as "fish" and assumed to be salmonids (Table 6.1). In a study in 1959, 3% of prey items of gulls shot at McNary Dam were juvenile salmonids (section 5-G-2), but, in 1997, 66% of the diet of California gulls collected at two dams were juvenile salmonids (section 5-E-5).
During the course of my review I found little information about fish-eating bird predation or control elsewhere in the Columbia Basin. However, the State of Washington has a policy about predator control at its hatcheries (Appendix Table II.1), Schaeffer (1992) reports about bird control at Oregon Department of Fish and Wildlife hatcheries, Associated Press (2002) discusses bird predation and control at the Washington Department of Fish and Wildlife's Ringold Springs hatchery (which is about midway between the Priest Rapids and Lower Monumental Dams), and Wildlife Services (2002a) mentions bird control at mid-Columbia hatcheries. Accordingly, bird predation in at least some Columbia Basin hatcheries occurs and may be significant (e.g., see Associated Press 2002).
Bird predation of juvenile salmonids also may regularly take place in the Columbia or Snake Rivers after large hatchery releases. Schaeffer (1992:9) notes that some Oregon Department of Fish and Wildlife hatchery staff observed birds flocking to sites after large releases, and this has also been observed along the Oregon coast (Bayer 1986). At the Spring Creek hatchery about 22 miles (35 km) upstream of Bonneville Dam, gulls were attracted after two releases in 1995, and the hatchery manager "reported that the gulls usually began feeding about 20 minutes after the start of a hatchery release, then follow the fish downstream" (Jones et al. 1996:12). After the 16 March 1995 release, three cracker shells from a shotgun scared the birds away, but they returned within 20 minutes (Jones et al. 1996:12).
Nesting birds may have caught juvenile salmonids at dams (section 6-B-1) or at hatcheries or in rivers, especially after hatchery releases (e.g., section 6-B-2). Gulls collected at nesting colonies often had not consumed juvenile salmonids, but juvenile salmonids were a significant prey for nesting Caspian terns at several of their colonies (Table 6.2). At one gull colony, the percentage of unidentified fish in the diet of gulls changed during the nesting season (section 5-H-1).
--------------------------------------------------------------------------TABLE 6.1. Index to information about fish-eating bird predation of juvenile salmonids at Columbia Basin dams in Chapters 4-5.
--------------------------------------------------------------------------
Description Section(s)
--------------------------------------------------------------------------
Taxon of Fish-eating Bird at Dams:
cormorants....................5-D-1, 5-E-2, 5-E-3, 5-J-2, 5-J-3, 5-K
gulls.........................4-D-1, 5-C-2, 5-D-1, 5-D-2, 5-D-3,
5-E-2, 5-E-3, 5-E-5, 5-E-7, 5-F-2,
5-F-4, 5-F-5, 5-G-1, 5-G-2, 5-H-1,
5-H-2, 5-I, 5-J-2, 5-J-3, 5-K, 5-N
Caspian terns.................5-E-3, 5-G-1, 5-J-2, 5-J-3, 5-K
American white pelicans.......5-G-1, 5-G-4, 5-K
other birds...................5-D-1, 5-E-2, 5-E-3, 5-G-1, 5-J-3, 5-K
Site of Bird Predation at Dams:
Bypass Tailrace/Outfall.......5-C-2, 5-D-2, 5-E-3, 5-E-4, 5-F-2, 5-G-4
Forebay in Front of Dam.......5-E-2, 5-E-3, 5-F-3
Power House Tailrace..........5-D-1, 5-E-2, 5-E-3, 5-E-4, 5-E-7, 5-K
Spillway Tailrace.............5-D-1, 5-D-2, 5-E-2, 5-E-3, 5-E-7,
5-F-2, 5-N
Ice/Trash Sluice Tailrace.....5-E-2, 5-E-7
Predation Affected by Weather, Stream Flows, and Turbidity: 5-B
Predation After Hatchery Releases: 5-D-1
Night-time Bird Foraging at Dams: 5-E-2, 5-E-3, 5-F-3
Juv. Salmonids as a % of Bird Diet at Dams:
unknown "fish" * ...........5-H-1 (gulls), 5-N (gulls)
1-25%.........................5-G-2 (gulls)
26-50%........................
51-75%........................5-E-5 (gulls)
>75%..........................
Estimate of Bird Predation at Dams:
number of juvenile salmonids..5-C-2, 5-D-2, 5-E-7, 5-I
% of migrating salmonids......5-C-2, 5-D-2, 5-E-4, 5-E-7, 5-I
* "Fish" found in bird stomachs were not identified as juvenile salmonids
or not all prey were identified.
--------------------------------------------------------------------------
--------------------------------------------------------------------------
TABLE 6.2. Index to information about fish-eating birds away from dams in the Columbia Basin from Chapters 1 and 4-6.
--------------------------------------------------------------------------
Description Section(s)
--------------------------------------------------------------------------
Other Sites of Bird Predation:
Hatcheries..........6-B-2
Hatchery Release Site..........6-B-2
Transportation Release Site....5-C-1
Nesting Fish-Eating Bird Taxon:
double-crested cormorant.......1-E-4, 5-L-2
California gull................1-E-2, 5-E-6, 5-G-3, 5-H-1, 5-L-1
ring-billed gull...............1-E-2, 5-E-6, 5-G-3, 5-H-1, 5-H-2,
5-L-1
Caspian tern...................1-E-3, 5-G-3, 5-L-1, 5-O
American white pelican.........1-E-4, 5-G-4
other birds....................1-E-4
Juv. Salmonids as a % of Nesting Bird Diet:
unknown "fish" * .............5-H-1 (gulls)
0% ............................5-E-6 (gulls), 5-G-3 (gulls),
5-L-1 (gulls)
1-25% .........................5-E-6 (gulls)
26-50% ........................5-O (terns), 5-E-5 (gulls)
51-75% ........................5-L-1 (terns)
>75% ..........................5-G-3 (terns)
* "Fish" found in bird stomachs were not identified as juvenile salmonids.
--------------------------------------------------------------------------
--------------------------------------------------------------------------
TABLE 6.3. Index to information about fish-eating bird control near Columbia Basin dams in Chapters 4-5.
--------------------------------------------------------------------------
Description Section(s)
--------------------------------------------------------------------------
Cost of Bird Control: .................4-E, 5-J-2
Timing of Bird Control:
April-September.....................4-D-4
Includes Winter.....................4-D-4, 5-F-4
Nonlethal Control Methods:
Effigies............................4-D-2
Frightening Devices * ..............4-D-1, 4-D-2, 5-F-4, 5-G-5, 5-J-3,
5-K
Effectiveness....................5-F-4, 5-G-5
Night Transportation Releases.......5-C-1
Silt Fencing to Discourage Nesting..5-L-3
Spotlight...........................5-F-3
Water Cannon/Sprayers...............5-C-1, 5-C-2, 5-F-2, 5-G-5
Effectiveness....................5-C-2, 5-F-2
Wire Grids..........................4-D-1, 4-D-2, 5-C-2, 5-D-3, 5-F-2,
5-J-2, 5-J-3, 5-K
Effectiveness....................4-D-2, 5-C-2, 5-D-3, 5-F-2, 5-K
Lethal Control Methods:
Shooting/Unknown....................4-D-1, 4-D-2, 5-F-5, 5-J-2, 5-J-3,
5-K
Effectiveness....................5-F-5
Egg Oiling..........................4-D-2, 5-H-2
* Pyrotechnics, including propane cannon, cracker or screamer shells.
--------------------------------------------------------------------------
The biological impact of bird predation is difficult to determine, so assumptions are made to estimate the amount of predation. The assumption that all fish passing dams during April through early or mid- August are juvenile salmonids appears to be reasonable for mainstem Columbia and Snake River dams, though many, if not most, fish passing by mid-August at lower Columbia dams may be juvenile shad (section 2-B). It is less clear if birds at dams are taking only salmonids, as is commonly assumed by researchers (e.g., Ruggerone 1986, Steuber et al. 1995, Jones et al. 1996-1999) because other fish may be available (section 2-C) and birds collected at dams may (section 5-G-2) or may not (section 5-E-5) include other fish species.
DAMS. At Bonneville, John Day, and Wanapum Dams; estimates of gull predation were about 2% or less of passing juvenile salmonids (sections 5-D-2, 5-E-7, and 5-I).
At The Dalles and/or John Day Dam, birds took 3-11% of radiotagged chinook, but it is not clear if the process of radiotagging increased predation, so that the predation rate for radiotagged chinook may be much higher than for nonradiotagged fish (section 5-E-4). Further, it is unknown if Wildlife Services' control methods have reduced bird predation at these dams since the study of radiotagged chinook (section 5-E-4).
BYPASSES. At the John Day Bypass, 1% of radiotagged chinook were thought to have been taken by birds (section 5-E-4). At the Bonneville Bypass before water cannons and wires were installed to exclude birds, gulls were estimated to take as many as 2% of passing salmonids, though calculations from published figures suggest 0.4% (section 5-C-2). After the bird control methods were installed, gulls were absent from the Bonneville Bypass outfall (section 5-C-2).
MULTIPLICITY OF PREDATION AT DAMS OR BYPASSES. Although predation levels of 1-2% per dam may not appear significant, there are so many dams that these predation levels may become significant, if viable salmonids are taken at each dam. For example, if there was a 1% predation level of viable salmons at each of the nine dams in the Columbia from Wells Dam downstream, then approximately 9% of salmonids that passed Wells Dam may be taken. But if salmonids only pass one dam (e.g., fish released at Spring Creek Hatchery just above Bonneville Dam, Jones et al. 1996:12), then predation is not multiplicative.
-----------[The following additional material was not in the original February 2003 report but was added on 6 April 2003.]
In 2002 at Rocky Reach and Rock Island dams, results of preliminary research reported by C. Thompson (Washington Dept. of Fish and Wildlife) in Wildlife Services (2003:18, 30) indicated that birds may be taking 1-3% of smolts entering the tailraces of these dams.
New reference:
Wildlife Services (WS). 2003. Pre-decisional Environmental Assessment: piscivorous bird damage management for the protection of salmonids on the mid-Columbia River. U.S. Dept. of Agriculture, Animal and Plant Health Inspection Service, Wildlife Services, 7209 O'Leary St. NW, Olympia, Washington; February 2003.
-----------[end of material added on 5 April 2003 since publication of this Review in February 2003]
It is not possible to reasonably estimate the impact of bird predation by just observing birds take fish-like prey at dams and assuming that birds are only taking viable, listed juvenile salmonids and thus have a major negative impact on salmon recovery. With overall direct dam mortality levels of 3-9% at some dams, it is plausible that a portion (perhaps a significant portion) of juvenile salmonids taken by birds below dams may already have been dead or mortally injured (section 2-F-2). Haas (1965:50), Ruggerone (1986:741), and Animal Damage Control staff Steuber et al. (1995:Discussion) acknowledge this, but National Marine Fisheries Service (2002:2-28, 2-29) and Wildlife Services (2002a) do not. Further, the impact of bird predation at one dam is unclear because fish taken at one dam may have died anyway from other sources of downstream compensatory mortality, so reducing bird predation at one dam may not lead to a proportional increase in survival (section 2-G). Finally, most juvenile salmonids do not appear to be from listed stocks (section 1-C). Nevertheless, the possibility that predation may have a significant effect on certain salmon stocks cannot be disregarded.
A significant unanswered question about bird predation at dams is what proportion of juvenile salmonids taken by birds is viable (section 2-F). Studies that only determine if juvenile salmonids have been taken by birds (e.g., see Table 6.1) do not address this question. Research such as Schreck and Stahl (2000:24, 46, 58) that attempted to ascertain if salmonids taken by terns in the Columbia Estuary were smaller, less smolted, or had higher infection levels of disease (e.g., Bacterial Kidney Disease) are essential, but studies should also examine fish taken by birds to determine if they have parasites (e.g., Ceratomyxa shasta, see Bartholomew et al. 1992) or show signs of physical injury (e.g., descaling) or gas bubble trauma from passage at dams. Gas bubble trauma is a concern at Columbia Basin dams, especially for fish with high levels of Bacterial Kidney Disease infection, and can make affected fish more vulnerable to predation (Elston et al. 1997, Mesa and Warren 1997, Mesa et al. 1998, Weiland et al. 1999). Unfortunately, juvenile salmonids taken by birds may be partially digested, so it may be difficult to determine if they were viable.
The level of predation that is unacceptable depends upon public acceptance, which may change with educational efforts and publicity (Decker and Goff 1987, Decker and Purdy 1988, Wildlife Services 2002c:4). In general, any predation may be too much for some people, moderate predation may be acceptable to many, and all predation may be permissible to others; so wildlife managers may suggest a middle ground of allocating some prey for predators (Thompson et al. 1995:187). Wildlife Services (2001c:21) also describes these differences in perspective:
"Local residents who are experiencing damage may want effective methods to be employed, whereas unaffected parties may not see any need for action. Aesthetically speaking, a passerby may view a large flock of migratory birds with great delight, whereas the property owner may view the same birds with disdain."
Acceptance of predation in the Pacific Northwest also depends upon the species of predator and prey (e.g., Independent Scientific Group 1996:333-334). For example:
1) Salmonids are themselves predators of other fish (including juvenile salmonids) and Dungeness crab larvae (Heg and Van Hyning 1951, Anonymous 1959, Angstrom and Reimers 1964, Fresh et al. 1981, Stuart and Buckman 1985, Thomas 1985, Brodeur et al. 1987, 1992; Bayer 1989:36-39, 45-46). But this predation is acceptable, although Thomas (1985) pointed out that salmon predation could be important to the Dungeness crab fishery.
2) Striped bass are a predator of juvenile salmonids, but are also popular for sportfishing (e.g., Morgan and Gerlach 1950, Shapovalov and Taft 1954:257, Johnson et al. 1992:104, Temple et al. 1998). In spite of their predation, the Oregon Fish and Wildlife Commission approved a plan in 1990 to enhance the striped bass population in Coos Bay, Oregon (Johnson et al. 1992, Temple et al. 1998), and the California Department of Fish and Game was also exploring increasing the number of striped bass, although they preyed on an endangered salmonid (Lindley and Mohr 1998).
3) Several species of warmwater gamefish (e.g., walleye and largemouth bass) sometimes prey on juvenile salmonids in the Columbia Basin or along the Oregon Coast but are also popular for sports anglers (Daily 1998, Oregon Department of Fish and Wildlife and National Marine Fisheries Service 1998, Shupp 1998, Temple et al. 1998, Beamesderfer 2000). In 1998, the Oregon Department of Fish and Wildlife proposed reducing this predation of juvenile salmonids by ending fishing restrictions on warmwater gamefish, but the resistance from fishing groups was intense enough for the Oregon Department of Wildlife to quickly withdraw its proposal (Monroe 1998a,b,c).
4) Northern pikeminnows are predators of juvenile salmonids in the Columbia Basin and are controlled in several areas (e.g., Friesen and Ward 1999, Beamesderfer 2000), but they are not popular with anglers. Walleyes also take some juvenile salmonids in the Columbia Basin, but are popular for sports fishing (Temple et al. 1998, Beamesderfer 2000:21). Controlling walleyes appears to be less practical in reducing their predation of salmon (Temple et al. 1998, Beamesderfer 2000:21), and Beamesderfer (2000:21), then of the Oregon Department of Fish and Wildlife, also notes:
"A walleye removal program would be less acceptable than the pikeminnow program, especially to an active and vocal group of walleye anglers."
The level of acceptance may also depend upon whether predation is on private or public lands. Some groups have suggested that the unacceptable level of predation be higher on public than private property (e.g., Humane Society of the United States in Wildlife Services 1998a:section 3.3.5). In the case of bird predation at Columbia Basin dams, the damage is not of a private resource on private property. Public opinion about the predator control is relevant because Wildlife Services is a governmental agency that is trying to control a public resource (bird predators) to protect a publicly owned resource (juvenile salmonids) on public waters, and operators of dams are public entities, with predator control publicly funded by taxes or electric rates (section 4-B-3). Wildlife Services' policy is to do predator control only if requested (Appendix Table II.2), so public input to dam operators as well as Wildlife Services may influence decisions about predator control. For the Columbia Basin, Wildlife Services conducted bird control by 1992 (Steuber et al. 1995), but an opportunity for public input to Wildlife Services about this control does not appear to have come until at least 10 years later (Wildlife Services 2002a). However, Wildlife Services (2002a) allowed public input only by letters for 18 days after it was first announced. In contrast, Wildlife Services' (2000a:6-1, 2000b:17) requests for public input in Kentucky and Wisconsin included a 30-40 day comment period, and comments were accepted by fax and letters as well as by e-mail in Wisconsin. Further, the Wildlife Services' (2002a) request for public comments about Wildlife Services' predator control projects at mid-Columbia dams had little information about its control projects, which makes it impossible for the public to provide knowledgeable comments about Wildlife Services' programs.
One criterion of an unacceptable level of predation is if more than 5% of juvenile salmonids are taken by predators. The Northwest Power Planning Council (1999) stated that its goal was to reduce Caspian tern predation to less than 5% of juvenile salmonids in the Columbia River Estuary. As pointed out in section 6-C-2, the few estimates of predation at Columbia Basin dams is 2% or less per dam and perhaps a significant portion of this predation is of dead or dying salmonids killed while going through turbines or bypasses or over spillways.
At Wells, Rocky Reach, and Rock Island Dams, the current National Marine Fisheries Service goal for juvenile salmonid mortality is 5% or less during dam passage (i.e., passage through the forebay, dam, and tailraces) and 7% or less for project passage (i.e., forebay, dam, tailraces, and reservoir)(National Marine Fisheries Service 2002:2-51, 2- 58). Bird predation is one component of this mortality (which would also include predation by fish), though it is unclear if bird predation is of viable juveniles or of juveniles killed or mortally injured during dam passage (section 2-F-2). In any case, if bird predation is greater than 5% of all juvenile salmonids for dam passage or greater than 7% for project passage, it would be clearly unacceptable for the current National Marine Fisheries Service objective for these dams.
[Also see section added in this online file: "UNACCEPTABLE PREDATION: 5% OR MORE OF DIET OF BIRDS"]
At Cabin Island, fish that were assumed to be juvenile salmonids were found to only be 0-4% of the diet of nesting ring-billed gulls during April 19-June 13 (Table 5.8). Yet all ring-billed gull eggs at this colony were oiled to control gull fecundity during 1995-1997, perhaps because of assumed gull predation of salmonids (section 5-H-2).
At Upper Nelson Island (which is probably Richland Island), Wildlife Services installed silt fences in 1999 to discourage gull nesting and reduce their predation of salmonids (section 5-L-3). However, none of the diet of ring-billed and California gulls nesting at Richland Island in 1997 and nearby Island 18 during 1997-1998 were juvenile salmonids (section 5-L-1).
Another criterion of unacceptable predation is if the costs of control are less than the costs of predation. Many wildlife management professionals and government staff, including some Wildlife Services' staff, recommend that the need for predator control be determined before conducting control and that control be cost effective, or the predation may not be significant enough to warrant the costs of predator control (Appendix II-D-1). However, in practice, Wildlife Services states that its control does not have to be cost effective (Appendix II-D-2), and Washington Wildlife Services indicates that it has the legal mandate to respond to all requests for wildlife damage, irregardless of the amount of loss (section 4-E).
When determining the benefits of predation control and in estimating its economic, biological, aesthetic, and social costs, it is important to use realistic estimates rather than unrealistically inflate the benefits and diminish the costs, but this has not always been done (e.g., Cain et al. 1972:12, 24-29; Beasom 1974, Peek 1986:254-261, Dolbeer et al. 1996:476).
At Columbia Basin dams, the costs of predation can be inflated by:
1) assuming that damage reports of predation losses by those requesting predator control are underestimated (section 4-E)
2) assuming all fish taken by birds are listed salmonids, though this is doubtful (sections 2-C, 2-D, and 2-E)
3) assuming birds are feeding only on viable salmonids, but many salmonids taken by birds may have been killed or mortally injured while passing through dam turbines or bypasses or crossing spillways (section 2-F-2)
4) assuming that there is no compensatory mortality of juvenile salmonids saved from birds; however, fish saved from birds can die from other mortality factors such as disease that can directly or indirectly lead to the death of salmonids (section 2-G).
The costs of predator control at Columbia Basin dams has been minimized by:
1) not publicizing the costs, so that the total cost is unknown and consequently less likely to be a controversial issue. The costs of predator control have only been reported for two dams in the Columbia Basin, where contracts with Wildlife Services for bird and fish control have been $160,715-$320,295 each year from 2000-2002, and studies of bird predation were not to exceed $330,000 in 2002 (section 5-J-2).
2) not examining all aspects of predation. For example, all ring-billed gull eggs were oiled at Cabin Island, although a maximum of 4% of the diet of gulls collected at the colony was fish (which were assumed to only be juvenile salmonids), but a much larger part of their diet was of insects and mammals (sections 5-H-1 and 5-H-2). Was the cost of salmonid predation offset by predation of insects and mammals injurious to agriculture? I do not know, and Wildlife Services does not address this issue (e.g., Wildlife Services 2001a, 2002a).
3) assuming that predator control methods have reduced predation rather than determining predation losses before and after control measures were implemented (section 6-G-1).
Wildlife Services states that predation does not need to be above any particular level nor be cost effective before it conducts control (Appendix II-D-2), so its position about predator control is more extreme than that of many wildlife management professionals (Appendix II-D-1). If Wildlife Services is requested to do predator control, it feels that it is required by law to do so (Appendix II-C).
For mid-Columbia dams, Woodruff (2002:2), which is given in Appendix III, writes that the goal of the National Marine Fisheries Service is to have "100% no net impact" to passing salmon and:
"The predation of smolt in the tailrace reduces the percent of juvenile fish surviving dam passage, which directly affects the smolt survival standard that each hydroproject must meet. The hydroproject managers have direction from the Federal Energy Regulatory Commission, NMFS, Northwest Power Planning Council, and others to use available mitigative measures to increase anadromous fish survival based upon the best scientific information available, and have identified predator control as a mitigation measure which is likely to increase smolt survival through each hydroproject on the mid-Columbia River."
One interpretation of Woodruff's statement is that a "100% no net impact" goal means a goal of 0% predation. However, this is not what the National Marine Fisheries Service (2002:2-50, 2-51, 2-58) state in its Final Environmental Impact Statement; it (2002:2-50) writes about the "no net impact standard":
"This term takes into account the fact that 100 percent equivalent survival cannot be achieved at the projects alone, requiring additional mitigation through off-site measures to increase salmonid productivity (e.g., hatchery supplementation programs and tributary habitat improvements."
For three mid-Columbia dams, National Marine Fisheries Service (2002:2- 50) notes that its goal is 91% combined adult and juvenile project survival (i.e., forebay, dam, tailraces, and reservoir) with "compensation for the 9 percent unavoidable project mortality" provided by hatchery programs (7%) and tributary habitat programs (2%). Further, its current goal for juvenile salmonid mortality at these three dams is 5% or less during dam passage (i.e., forebay, dam, and tailraces) and 7% or less for project passage (National Marine Fisheries Service 2002:2-51, 2-58). Reducing bird predation may help reach these goals if bird predation is only of viable juveniles, but this has not yet been demonstrated (section 2-F-2).
To conclude, National Marine Fisheries Service (2002:2-58, 2-59, 2- 60, 3-138, 3-140) discusses bird control at three mid-Columbia dams but does not state that its goal is 0% predation.
-----------[The following additional section was not in the original February 2003 report but was added on 6 April 2003.]
UNACCEPTABLE PREDATION: 5% OR MORE OF DIET OF BIRDS
Draulans (1988:252) wrote that in some early studies that if less than 5% of the diet of birds was "valuable" fish, then it was often assumed that "no damage was done." He also notes that determining the impact of bird predation on fish populations should also include calculations of fish abundance to determine what proportion of the fish population is taken by birds. Derby and Lovvorn (1997:1490) also cited Draulans (1988), but noted that it was unclear in their study if bird predation was an additional source of mortality or if fish taken by birds would have died anyway from other sources of mortality.
In the Columbia Basin, juvenile salmonids can sometimes be 5% or more of the diet of birds (Tables 6.1 and 6.2), but it is unclear if juvenile salmonids saved from this bird predation would die from other causes (section 2-G).
At Columbia Basin Dams, bird predation is generally 3% or less of passing juvenile salmonids (section 6-C-2).
-----------[end of material added on 5 April 2003 since publication of this Review in February 2003]
Wildlife Services (2002a) indicates that its control at mid-Columbia River dams is during April-September, but its control at lower Columbia dams continued outside of this time period, since Jones et al. (1999:14) recommended that Wildlife Services' control activities be reduced or discontinued during winter. It seems questionable that control conducted from mid-August through March would be very helpful in protecting juvenile salmonids at lower Columbia dams because mostly juvenile shad, not salmonids, are passing these dams then (sections 1-D and 2-B). At dams further upstream, there appear to be relatively few migrating juvenile shad (section 2-B), but there are also few juvenile salmonids migrating after mid-August (section 1-D). Thus, bird control in September at these dams may also not be useful in protecting juvenile salmonids. Further, it is questionable that controlling wintering birds that leave before juvenile salmonids begin migration would have any effect on reducing bird predation of salmonids.
Wire grids and frightening devices have been the most extensively employed forms of nonlethal control, although water cannons or sprayers have also been used in certain circumstances (Table 6.3).
Wires may injure a few birds (e.g., possibly two in 1997, Jones et al. 1998:6), so Jones et al. (1997:16, 1998:9) suggested adding mylar streamers to the wires to help make the wires more visible.
Overhead wires can often keep away flying birds, but not swimming birds. At three lower Columbia dams, overhead wires installed by Wildlife Services appeared to be 100% effective in keeping out flying birds when the wires were maintained and were close enough together; however, it was not possible to install them in all areas where flying birds were feeding (sections 5-D-3 and 5-F-2). The effectiveness of wire arrays was also reduced when they were vandalized (Jones et al. 1997:15, 17; 1998:9, 1999:14) or contractors at John Day Dam removed and did not replace an array (see footnote ** in Table 5.4).
At mid-Columbia dams, Wildlife Services reported that wires were not 100% effective because gulls learned to feed under them (Appendix III). However, wires at these dams were spaced up to 50 ft (15.2 m) apart, which is not as effective as more closely spaced wires (section 4-D-2).
Elsewhere, wires have also kept birds away if spaced near each other, but overhead wires can not always be installed (section 3-B-1).
Water cannons and frightening devices were two popular ways to keep birds away from dams (Table 6.3). Water cannons or sprinklers could be very effective in keeping fish-eating birds away, if they could be operated and if it was not too windy (sections 5-C-2 and 5-F-2). Frightening devices such as cracker shells or propane cannons reduced the number of birds present when used but kept birds away for a limited period of time (sections 5-F-4, 5-G-5, and 6-B-2). This is similar to results elsewhere (section 3-B-2).
The use of lethal control is controversial, especially if many individuals are taken (section 3-C-1, Appendix II-E-2). Washington Wildlife Services uses lethal control of many species of fish-eating birds more than all other states combined (Table 4.1), but it declined to say how much of its lethal control occurred at dams (section 4-B). Nevertheless, there is some information about its lethal control at Columbia Basin dams (sections 4-D-1, 4-D-2, 5-F-5, 5-J-2, and 5-J-3). A research group from the U.S. Army Corps of Engineers observed the effects of Wildlife Services' lethal control at The Dalles and John Day Dams and found that other birds often returned to the area after a bird was shot (section 5-F-5), so lethal control needs to be continued to keep birds away. An indication that lethal control is of uncertain effectiveness in reducing predation of salmonids is that 75% of a sample of gulls shot by Wildlife Services during its control program at The Dalles and John Day Dams in 1997 did not contain juvenile salmonids (section 5-E-5), so it is not clear that shooting these gulls reduced predation, though it may have deterred other gulls that may have fed on salmonids.
Wildlife Services oiled ring-billed gull eggs in at least one Washington colony (section 5-H-2) and refused to answer a question about whether it currently oils eggs at bird colonies (section 4-B-2). The number of eggs oiled was not included in tables of Wildlife Services' "take" of animals (Table 4.3, section 5-H-2).
Washington Wildlife Services does not appear to have measured predation losses before and after using lethal control to see if its lethal control actually reduced predation of juvenile salmonids. It has counted the number of birds taken and assumed that predation was thereby reduced.
In 2002, studies of predation at Rocky Reach and Rock Island dams with the Washington Department of Fish and Wildlife and the University of Washington were funded and initiated (section 5-J-2). In 2002, David Wesley of the U.S. Fish and Wildlife Service's Migratory Birds and State Programs emailed Gerald Winegrad that Wildlife Services is not killing terns for damage control at dams in 2002 but is only taking "terns in coordination with Washington Department of Fish and Wildlife and wildlife research."
Wildlife Services' Directive 2.201 (Wildlife Services 1993a:section 4b) indicates that Animal Damage Control/Wildlife Services will assess a wildlife damage request; this Directive states:
"First, a determination should be made as to whether the problem is within the authority of ADC. If it is, damage information should be gathered and analyzed to determine factors such as what species was responsible for the damage; the type, extent, and magnitude of damage; the current economic loss and potential losses; the local history of damage; and what control methods, if any, were used to reduce past damage and the results of those actions."
Washington Wildlife Services does not appear to have complied very well with this Directive since during 1996-2000 a total of only $400 of damage to natural resources in Washington was reported and verified by Wildlife Services (section 4-E). Washington Wildlife Services appears to have assumed that damage was occurring because fish-eating birds were observed.
Directive 2.201 also indicates that Wildlife Services will monitor and evaluate the results of control actions to determine "whether further assistance is required or whether the problem has been resolved" (Wildlife Services 1993a:section 4f). Washington Wildlife Services may not have followed this Directive because it has not released any evidence that it has (section 4-B). Staff of the U.S. Army Corps of Engineers, not Wildlife Services, has monitored and reported on some Washington Wildlife Services' control actions (sections 5-D and 5-F), including a possible reduction in predation of juvenile salmonids as a result of the installation of overhead wires. Wildlife Services' staff have only reported that they have controlled birds at dams or nesting areas (e.g., sections 4-D-1, 4-D-2, 5-H-2, 5-K, and 5-L-3), not that there has been a reduction in the predation of juvenile salmonids. Thus, it appears that the original salmon recovery goal of increasing juvenile salmonid survival by reducing predation may have been changed to the Washington Wildlife Services' goal of decreasing numbers of fish-eating birds, with measures of success being the number of birds hazed or taken or the number of eggs that are oiled.
Bird (especially Caspian tern) predation of juvenile salmonids in the Columbia River Estuary has received considerable research and publicity (e.g., Roby et al. 1998, 2002; U. S. Army Corps of Engineers 1999-2002; Collis et al. 2000, 2001a,b; 2002). Bird control in the Estuary has proceeded differently than at Columbia Basin dams. In the Estuary, many of the guidelines for justifying control in Appendix II-D-1 have been followed. For example, research was conducted in 1997 to determine if control was needed (Roby et al. 1998) and as a result of this research, possible management options were proposed that incorporated public comments as well as the input of many federal, state, and tribal agencies (U.S. Army Corps of Engineers 1999-2002). After a need was established, predator control was initiated in 1999, using nonlethal methods of habitat manipulation of nesting areas and putting bird excluders on pilings to discourage bird perching (e.g., U.S. Army Corps of Engineers 2001:3, Roby et al. 2002).
At lower Columbia Basin dams, bird predation has been studied at The Dalles and John Day Dams during 1995-1998 (section 5-E), but bird hazing and lethal control occurred at these two dams in 1995 (Jones et al. 1996:3) and perhaps earlier (Steuber et al. 1995). Thus, control began before predation studies were complete and before a need for control was established. Control also appears to have commenced before public comment was solicited for an Environmental Assessment, since I have found no record of an Environmental Assessment being proposed or done for these dams. At mid-Columbia dams, control began as early as 1993 (Steuber et al. 1995), but I have found no evidence of a predation study other than by Ruggerone (1986) and public comments were not requested until 2002 (Wildlife Services 2002a). At Columbia Basin dams, it also does not appear that Washington Wildlife Services has involved or consulted other government agencies as much as the U. S. Army Corps of Engineers and National Marine Fisheries Service has about proposed bird control measures in the Columbia Estuary (e.g., U. S. Army Corps of Engineers 1999-2002).
Because of the deliberate process in the Columbia Estuary, there are estimates of how much predation has been reduced after management actions (Collis et al. 2001b:13, U.S. Army Corps of Engineers 2001:7, 2002:23; Roby et al. 2002:669). In contrast, Wildlife Services has only assumed that its bird control projects at dams have reduced predation.
Salmon recovery in the Columbia Basin is important, and predator control may sometimes be appropriate to improve adult returns of salmonids. However, predator control, especially lethal control, needs to be based on a realistic consideration of costs and benefits as has been suggested in Appendix II-D-1. Because predators and prey are public resources and predator control occurs at Columbia Basin dams that are operated by public entities on public waters and predator control is publicly financed (section 4-B-3), the public should be part of the decision-making process. To do so, accurate and complete information about the biological value and the cost effectiveness of nonlethal and of lethal control needs to be publicly available. Although a court case precludes Wildlife Services from giving information that would identify individual dam owners, Washington Wildlife Services could release more information than it has (section 4-B-2).
In my opinion, Washington Wildlife Services' refusal to answer general questions (section 4-B) and decision that its predator control programs do not have to be cost effective (sections 4-E and 6-D-4) can make conducting appropriate wildlife control measures more difficult because of public distrust (see Appendix II-F). In addition, its lack of candor seems unacceptable for a government agency (section 4-B-3).
Common Name Scientific Name bass, largemouth Micropterus salmoides bass, striped Morone saxatilis cormorant, double-crested Phalacrocorax auritus crab, Dungeness Cancer magister crow, American Corvus brachyrhynchos eagle, bald Haliaeetus leucocephalus goose, Canada Branta canadensis grebe, eared Podiceps nigricollis grebe, pied-billed Podilymbus podiceps grebe, western Aechmophorus occidentalis gull, Bonaparte's Larus philadelphia gull, California Larus californicus gull, Franklin's Larus pipixcan gull, glaucous-winged Larus glaucescens gull, great black-backed Larus marinus gull, Heermann's Larus heermanni gull, herring Larus argentatus gull, laughing Larus atricilla gull, ring-billed Larus delawarensis gull, western Larus occidentalis heron, black-crowned night- Nycticorax nycticorax heron, green Butorides virescens heron, great blue Ardea herodias kingfisher, belted Ceryle alcyon lamprey, Pacific Lampetra tridentata lamprey, western brook Lampetra richardsoni loon, common Gavia immer merganser, common Mergus merganser merganser, hooded Lophodytes cucullatus merganser, red-breasted Mergus serrator osprey Pandion haliaetus pelican, American white Pelecanus erythrorhynchos pikeminnow, northern Ptychocheilus oregonensis salmon, chinook Oncorhynchus tshawytscha salmon, coho Oncorhynchus kisutch salmon, sockeye Oncorhynchus nerka shad, American Alosa sapidissima steelhead Oncorhynchus mykiss tern, Caspian Sterna caspia tern, Forster's Sterna forsteri walleye Stizostedion vitreum
There has been public controversy about some Animal Damage Control/Wildlife Services' programs (e.g., search the Internet for "animal damage control" or "predator control"). This has resulted in investigative reports by the U.S. General Accounting Office (1990, 1995, 2002b) and attempts to cut funding of Wildlife Services during 1998-2000 in the U.S. House (Swanson 1998, Eilperin 1999, Taugher 1999, DeFazio 2000). It is beyond the scope of this review to examine all these controversies, but it is relevant to examine some of the background of Wildlife Services that is relevant to bird predation at dams in the Columbia Basin. Thus, complaints that some Wildlife Services' projects are a subsidy to private businesses (e.g., U.S. General Accounting Office 1990:15, Deeble and Stadler 1993, O'Toole 1994, Eilperin 1999, Taugher 1999, DeFazio 2000) are not examined because private businesses do not operate most Columbia Basin dams (see Fig. 1.1 legend). Also complaints about Wildlife Services' programs in suburban areas (Portland Audubon Society 2000) are not discussed because Columbia Basin dams are not in suburban areas.
The history of Wildlife Services is examined in Deeble and Stadler (1993:6), O'Toole (1994:4-5), and U.S. General Accounting Office (2002b:47-52). Here, I only discuss some of the name changes that are relevant to this paper.
In 1939, federal predator control was moved from the U.S. Department of Agriculture to the U.S. Department of Interior and renamed Predator and Rodent Control (PARC)(U.S. General Accounting Office 2002b:48). The name of PARC was later changed to Animal Damage Control, and, in 1986, Animal Damage Control was transferred back to the Department of Agriculture, Animal and Plant Health Inspection Service (Deeble and Stadler 1993:6). In August 1997, the name of Animal Damage Control was changed to Wildlife Services (Wildlife Services 1999:1-1).
The authority and mission of Wildlife Services come from the U.S. Animal Damage Control Act of 1931 (7 U.S.C. 426-426c; 46 Stat. 1468), which states (Wildlife Services 1998a:section 1.5.1):
"The Secretary of Agriculture is authorized and directed to conduct such investigations, experiments, and tests as he may deem necessary in order to determine, demonstrate, and promulgate the best methods of eradication, suppression, or bringing under control on national forests and other areas of the public domain as well as on State, Territory or privately owned lands of mountain lions, wolves, coyotes, bobcats, prairie dogs, gophers, ground squirrels, jackrabbits, brown tree snakes and other animals injurious to agriculture, horticulture, forestry, animal husbandry, wild game animals, furbearing animals, and birds, and for the protection of stock and other domestic animals through the suppression of rabies and tularemia in predatory or other wild animals; and to conduct campaigns for the destruction or control of such animals. Provided that in carrying out the provisions of this Section, the Secretary of Agriculture may cooperate with States, individuals, and public and private agencies, organizations, and institutions."
Wildlife Services (2000a:section 1.7.1.1) writes about its changing philosophy:
"Since 1931, with the changes in societal values, Wildlife Services policies and its programs place greater emphasis on the part of the Act discussing 'bringing (damage) under control,' rather than 'eradication' and 'suppression' of wildlife populations."
In 2001, the Animal Damage Control Act of 1931 was amended in the Agriculture Appropriations Bill and states (Wildlife Services 2002c:19):
"The Secretary of Agriculture may conduct a program of wildlife services with respect to injurious animal species and take any action the Secretary considers necessary in conducting the program."
Wildlife Services' major constituencies are those citizens and businesses requesting wildlife control. For example, Wildlife Services (2002c:19) states:
"The USDA is directed by law to protect American agriculture and other resources from damage associated with wildlife."
Many wildlife management professionals have recommended that animal damage control be conducted only if the economic loss or impact on the resource by predation is greater than the economic, biological, aesthetic, and/or social costs of conducting control (e.g., Berryman 1972:397, 399; McCabe and Kozicky 1972:393, Dolbeer et al. 1996:474). Some Animal Damage Control/Wildlife Services' staff have also recommended that before predator control is initiated an environmental as well as an economic cost effective analysis be conducted (Slate et al. 1992:57-59) and that wildlife management actions be used that are cost effective "not only with respect to economics, but to biological, physical, social, and legal parameters, as well" (Owens and Slate 1991:26). Further, the International Association of Fish and Wildlife Agencies' promotes:
"thorough planning of control programs with justification, implementation, and evaluation on the basis of total social benefits" (Belanger 1988:200).
At fish hatcheries, it has also been recommended that predator control be economically cost effective (i.e., the costs of control are not greater than the costs of predation). For example, predation losses at Oregon Department of Fish and Wildlife hatcheries have been estimated, so that it can be determined if the magnitude of loss justifies the cost of covering ponds or other control measures (Schaeffer 1992:10-12). Washington Department of Wildlife policy is also that the cost of a predator control system not exceed the value of the fish taken by predators (Appendix Table II.1#5). Animal Damage Control/Wildlife Services' staff have also recommended that predator control at aquaculture facilities only be done only if it is economically justifiable (e.g., Glahn 1997:16- 17, Littauer et al. 1997:1). Finally, in a review of predation at Pennsylvania trout hatcheries, Parkhurst et al. (1992:415) write:
"Clearly, an inaccurate assessment of the extent of depredation or a misidentification of what predator is responsible could lead to improper management of predation losses and a potential waste of capital on needless or ineffective control methods."
Nevertheless, predator control programs have sometimes been initiated on the basis of assumed benefits and unrealistic predato