Response of Pacific loons (Gavia pacifica) to impoundments at Prudhoe Bay, Alaska.
ABSTRACT. Reproductive success and foraging effort of Pacific loons (Gavia pacifica) were compared between impoundments and natural ponds in the Prudhoe Bay oil field, Alaska, in 1992 and 1993. Pacific loons successfully reproduced on both impoundments and ponds. Though success tended to be lower on impoundments, no significant differences were detected between the two water body types. The principal cause of reproductive failure on both impoundments and ponds in 1993 appears to have been predation by arctic foxes (Alopex lagopus) during incubation. Predation losses on impoundments often were associated with large water-level drawdowns, which apparently increased opportunities for predation of stranded clutches. I found no differences in chick mortality or adult foraging effort between impoundments and ponds. Over 99% of the prey items delivered to chicks were invertebrates captured in freshwater wetlands. Reproductive output of Pacific loons may be improved by increasing the stability of water levels at impoundments used for nesting. I found no evidence that draining impoundments would improve habitat for loons (or invertebrate-eating waterfowl species) at Prudhoe Bay.
Key words: Pacific loon, Gavia pacifica, Prudhoe Bay, impoundments, impact assessment, Alaska
RESUME. En 1992 et 1993, on a etabli une comparaison du taux de reussite de la reproduction et de la quete de nourriture du huart du Pacifique (Gavia pacifica) entre les bassins de retenue et les etangs naturels situes dans le champ petrolifere de Prudhoe Bay en Alaska. Le huart du Pacifique se reproduisait bien sur les bassins comme sur les etangs. Malgre un taux ayant tendance a etre plus faible sur les bassins, on n'a pas remarque de differences notables entre les deux nappes d'eau. En 1993, la raison principale de l'echec de la reproduction sur les retenues comme sur les etangs semble avoir ete la predation effectuee par le renard arctique (Alopex lagopus) au cours de l'incubation. Les pertes dues a la predation sur les bassins etaient souvent liees a de grands rabattements de nappe qui, selon toute apparence, facilitaient l'acces des predateurs aux pontes echouees. On n'a trouve aucune difference dans la mortalite juvenile ou dans la quete de nourriture entre les bassins et les etangs. Plus de 99 p. cent des proies rapportees aux poussins etaient des invertebres captures dans des marais d'eau douce. On pourrait rendre la reproduction du huart plus efficace en preservant la stabilite du niveau des eaux dans les bassins qui servent a la nidification. On n'a pas trouve de preuves a l'effet que le drainage des bassins ameliorerait l'habitat du huart (ou des especes d'oiseaux aquatiques se nourrissant d'invertebres) a Prudhoe Bay.
Mots cles: huart du Pacifique, Gavia pacifica, Prudhoe Bay, bassins de retenue, evaluation environnementale, Alaska
Traduit pour la revue Arctic par Nesida Loyer.
Within oil fields in arctic Alaska, water impounded beside gravel roads and pads constitutes one of the major human-induced landscape disturbances in terms of area affected (Walker et al., 1987). Some scientists and regulatory agency personnel have expressed concern about the potential negative effects of impoundments on wildlife populations due to habitat loss or alteration. Options for mitigating such potential effects, including draining impoundments through installation of culverts, are being considered by the oil industry. However, for mitigation plans to be successful, they must be based on knowledge of how wildlife populations are actually affected by impoundments. When such knowledge is unavailable, selected species are sometimes used as "indicators" to infer impacts for a wider array of species. An indicator species is defined by Landres et al. (1988:317) as "an organism whose characteristics (e.g., presence or absence, population density, dispersion, reproductive success) are used as an index of habitat attributes too difficult, inconvenient, or expensive to measure for other species or environmental conditions of interest." Thus, an indicator can serve as a surrogate for other species.
Pacific loons have attributes that make them one of the few arctic waterbirds for which impacts of oil development can be measured easily. They rely almost entirely on freshwater habitats for food during the breeding season (Bergman and Derksen, 1977; McIntyre, 1994), and often spend the entire reproductive season within one oil field. During this time their life-support needs may be met within a single small wetland (Davis, 1972; Bergman and Derksen, 1977). They are easy to observe and respond to habitat differences in ways detectable to observers. They are relatively easy to relocate following dispersal from nest sites. Finally, they are abundant in arctic wetlands, providing resource managers with the opportunity to collect data in a variety of locations and habitat types.
Several researchers have studied loons to assess habitat conditions. Strong (1990) concluded that habitat conditions for breeding common loons (Gavia immer) are reflected most accurately by reproductive output, and that the best measurement of reproduction may be a count of prefledged chicks 7-10 weeks old. Dickson (1993) similarly indicated for redthroated loons (Gavia stellata) that reproductive success, as measured by young that survive to fledging, was a good indication of local habitat conditions. Andres (1993) suggested that local foraging behavior of Pacific loons should be determined before they are used as an indicator of freshwater habitat conditions. The relationship between prey availability and the foraging tactics of vertebrate consumers in aquatic systems has been used to assess habitat quality not only for loons (Alvo, 1986) but for other species as well, including sea otters (Enhydra lutris) (Estes et al., 1986) and dippers (Cinclus cinclus) (Omerod and Tyler, 1987).
This paper compares reproductive success and foraging effort of Pacific loons on impoundments and natural ponds, and thereby evaluates impoundments as waterbird habitat at Prudhoe Bay.
I studied Pacific loon biology in the Prudhoe Bay oil field, between the Kuparuk River and the Putuligayuk River, on the Arctic Coastal Plain of Alaska (Fig. 1). The oil field borders the Beaufort Sea and is located in a region of low relief covered by numerous shallow lakes and ponds (Walker et al., 1980).
Temporary and permanent impoundments together account for a major proportion of the acreage disturbed in the Prudhoe Bay region. Most instances of impounded water occur where gravel roads and pads block drained thaw-lake basins or other low-lying areas (Alexander and Miller, 1978; Brown et al., 1984; Walker et al., 1986; Walker et al., 1987). The amount of acreage flooded by impoundments usually peaks as snowmelt ends around mid-June. In the Prudhoe Bay oil field, many of the flooded areas are temporary; they drain completely by mid-summer after the ice plugging road culverts melts and surface run-off rates subside (Alexander and Miller, 1978; Klinger et al., 1983; Walker et al., 1987). However, other flooded areas are permanent, draining only partially.
Water Body Classification
Some terms require definition. I often refer to impoundments and natural ponds as water bodies. Wetland refers to a class or type of impoundment or pond (e.g., Shallow-Arctophila) and therefore differs from the regulatory definition of wetland (Environmental Laboratory, 1987). Impoundments are water bodies created by anthropogenic alterations to the landscape surface.
Water bodies with territorial or nesting Pacific loons were plotted on a 1991 hydrological map (1"=500') of the Prudhoe Bay oil field produced from aerial photography taken in July and August. These water bodies were then classified as natural ponds or impoundments on the basis of a 1983 map (Lederer et al., 1984) depicting impoundment locations in the oil field. A recent analysis of pre-development aerial photographs (Noel et al., 1995) indicates that some water bodies depicted as "impoundments" by Lederer originally were natural ponds that were enlarged by the presence of impounded spring runoff following the construction of nearby roads and drilling pads. I found that the accumulation of surface water not only enlarged temporarily the size of these water bodies, but, more importantly, resulted in earlier thaw dates, apparently by enhancing heat absorption (Truett and Kertell, 1992). They were therefore used earlier by loons than nearby natural ponds. Furthermore, because the majority of loons initiated nests in early spring, several nests placed within temporarily impounded portions of these water bodies were left stranded above water as drying of spring meltwater progressed. Therefore, the presence of impounded water has the potential to impact loon use and reproduction, whether or not these water bodies are modified natural ponds.
In 1993, water body size was measured from a digital version of the 1991 hydrological map using MapInfo[R] (MapInfo Corporation, Troy, New York). For impoundments with large drawdowns, boundaries prior to drawdown were estimated visually, transposed onto the digital hydrology map, and measured using MapInfo[R]. Percent open water was estimated daily in 1992 at 15 randomly selected impoundment/pond pairs with and without nesting loons until the water bodies were 100% ice-free. I classified water bodies with nesting loons in the field according to a classification system developed by Bergman et al. (1977), except that I did not use the Basin-complex wetland type because it may contain any or all of the other wetland classes (Derksen et al., 1979; Taylor, 1986).
I located territorial Pacific loon pairs within approximately 0.8 km of gravel roads and pads by scanning water bodies from a parked vehicle with 10x binoculars and a spotting scope with a 20-45x zoom lens. In 1992, I did not search for nests until the beginning of July, 7-10 days after peak nest initiation that year. In 1993, I started nest surveys on 10 June, eight days before the first nest was initiated.
Nest construction and clutch initiation were observed for the majority of pairs in 1993. For nests located after laying had taken place, clutch initiation dates were calculated by back-dating from the date of egg-hatching. For this purpose, I assumed an incubation period of 26 days (Johnson and Herter, 1989). A clutch was considered to be a replacement if it was found within an identified territory following the known failure of an earlier clutch. In such cases, I assumed that both clutches were initiated by the same pair of adults.
Loon pairs were checked daily with binoculars or spotting scope from a parked vehicle to determine stage of nesting. Following nest initiation, nests were checked daily to monitor their fate. To avoid disturbing nesting birds, which might have increased the chances of predation on eggs, I did not visit active nests on foot to observe breeding behavior or determine clutch sizes. Nest sites were classified as (1) natural island, (2) platform (constructed by loons from mud and/or emergent vegetation), or (3) mainland shoreline (including peninsulas).
Nest success is defined as the number of breeding pairs that hatched at least one young. When nesting was successful, I determined hatching date, number of young, survival of young, and frequency and timing of movements from the nest site to adjacent water bodies. Hatching was considered to have occurred on the day when at least one young was first observed, and production estimates were based on the number of young observed immediately after hatch. Newly hatched young were sometimes difficult to observe because they either hid or were brooded by adults; thus hatching dates were sometimes approximated on the basis of adult behavior (e.g., adults roosted near the nest site) or by counting forward from the date of nest initiation. Through daily visits to water bodies, I was able to track movements of individual families by noting the number and size of young. Surviving young were monitored daily until 5 August in 1992 and until 27 August in 1993. Survival to 27 August was considered a reasonably good indication of breeding success at Prudhoe Bay since by then most chicks were over five weeks old, hence within about 10 days of fledging (Parmelee et al., 1967).
A nest that appeared inactive (i.e., the adults were absent from the water body) or continued to be attended by an incubating adult well beyond the normal incubation period was visited on foot to determine nest fate. Each failed nest was assigned to one of four categories:
(1) Disappeared. No trace of shell or membrane was found.
(2) Predated. Traces of membrane were found in or near the nest bowl or where predators were observed at the nest.
(3) Deserted. The clutch was found intact and there was no evidence of activity for at least two days.
(4) Unknown. The nest was inaccessible to researchers because of water body depth.
TABLE 1. Wetlands used by Pacific loons for nesting in 1993, Prudhoe Bay, Alaska (wetland type after Bergman et al., 1977). Wetland type Ponds (%) Impoundments (%) Flooded tundra 0 1 (2.4) Shallow-CAAQ (1) 20 (24.7) 19 (45.2) Shallow-ARFU (2) 38 (46.9) 17 (40.5) Deep-ARFU 23 (28.4) 4 (9.5) Beaded stream 0 1 (2.4) Coastal wetlands 0 0 Deep-open lakes 0 0 Total 81 (100) 42 (100) (1) CAAQ = Carex aquatilis (2) ARFU = Arctophila fulva
In 1992 and 1993, adult loons with young were selected for detailed observations of foraging effort. Observations were conducted at six pairs of ponds and impoundments in 1992, and at eight pairs of ponds and impoundments in 1993. At these sites, time spent foraging, time spent feeding young, and number and duration of foraging "bouts" were recorded during four-hour observation periods. A bout refers to a period during which relatively brief feeding behaviors were repeated several times in succession. Two loon pairs, one at a natural pond and one at an impoundment, and each with the same number of young of similar age, were observed concurrently by two observers during morning and afternoon observation periods. Three observation periods (12 hours total) were conducted for each pair. To estimate total prey delivered to young at each site, prey delivery rates (number of prey per 10-minute period) were recorded at intervals of 30 minutes when loons were foraging. Although I could not accurately identify individual prey items during prey exchange, I could distinguish between invertebrates and fish by size and shape.
I used contingency table analysis to compare impoundments with ponds relative to loon choice of wetland type and nest substrate, nest success, and productivity. Yates's correction was applied to all two-by-two contingency tables. Paired-sample t-tests were used to compare foraging times and prey delivery rates. I considered tests to be statistically significant at p < 0.05. Standard deviations (SD) are provided as the measure of variability throughout the text.
I located 46 nests in 1992 (24 on natural ponds and 22 on impoundments) and 123 nests in 1993 (81 on natural ponds and 42 on impoundments). In 1993, the three most commonly used wetland classes on both impoundments and ponds were Shallow-Carex, Shallow-Arctophila, and Deep-Arctophila (Table 1). No nests were placed on Deep-open Lakes. The impoundments and ponds that had nests in 1993 showed wide variation in size (Table 2).
TABLE 2. Sizes of water bodies used for nesting by Pacific loons in 1 993, Prudhoe Bay, Alaska. Water body type (n) Mean area (ha) SD Range Pond 81 10.71 15.92 0.35 - 65.30 Impoundment (1) 42 8.20 9.64 0.50 - 43.10 (1) size before drawdown
Presumably many birds returned to the same wetland in 1993 that they had used the previous year. Ninety percent of the impoundments (n = 22) and 79% of the ponds (n = 24) that were used for nesting in 1992 were used again in 1993. In 1993, 50% of the pairs on impoundments (n = 20) and 32% of the pairs on ponds (n = 19) nested on water bodies where nests had failed in 1992.
Nest Site Selection
Pacific loons tended to prefer natural islands and platforms as nesting sites (Table 3). In 1993, the majority of nests on impoundments (79%) and natural ponds (73%) were located on these two substrate types (Table 3). Choice of nest substrate in 1993 was independent of water body type (2 = 2.37, d.f. = 2, p = 0.31).
In 1993, nest success by substrate type tended to be different between impoundments and ponds (Table 3). For nests located on islands and platforms, a greater percentage were successful on ponds (28 of 59, 48%) than on impoundments (9 of 33, 27%). A large percentage (17 of 21, 81%) of mainland nests on ponds failed. On impoundments, 50.0% of mainland nests were successful, but the sample size was small (n = 8).
Timing of Breeding
Mean clutch initiation dates and mean hatch dates were similar on ponds and impoundments (Table 4). In 1992, impoundments (n = 15) were 100% ice-free on 9 June, six days earlier than ponds (n = 15), but first clutches were initiated on 18 June and 19 June, respectively. Thaw dates were not recorded in 1993.
In 1993, Pacific loons laid replacement clutches on both impoundments (n = 5) and ponds (n = 5). Mean date of replacement clutches for all water bodies pooled was 10 July (range 3-19 July). Average time between the loss of the original clutch and the initiation of the second clutch was 12.4 days ([+ or -] 3.7).
= 0.65, d.f. = 1, p = 0.42). The majority of failed nests on both water body types contained no eggs or egg remains in 1993 (Table 6). Because 1992 studies began after clutch initiation, results likely underestimate predation losses early in the incubation period.
[Part 1 of 2] TABLE 3. Nest success by substrate type in 1993, Prudhoe Bay, Alaska. Impoundment nests Nest substrate Successful Unsuccessful Total (%) Natural island 7 19 26 (61.9) Vegetation platform 2 5 7 (16.7) Mainland shore 4 4 8 (19.0) (including peninsula) Unknown 1 0 1 (2.4) Total 14 28 42 (100.0) [Part 2 of 2] TABLE 3. Nest success by substrate type in 1993, Prudhoe Bay, Alaska. Pond nests Nest substrate Successful Unsuccessful Total (%) Natural island 19 20 39 (48.2) Vegetation platform 9 11 20 (24.7) Mainland shore 4 17 21 (25.9) (including peninsula) Unknown 1 0 1 (1.2)
Although nest success tended to be higher on ponds than on impoundments in 1993 (Table 5), it was not significantly different (41% versus 33%, but 2
Total 33 48 81 (100.0)
[Part 1 of 2] TABLE 4. Nesting chronology of Pacific loons in 1992 and 1993, Prudho e Bay,
Water body Mean clutch initiation date [+ or -] SD (range) Year type First attempts All nests 1992 Pond 22 June [+ or -] 2.4 (n = 14) 23 June [+ or -] 5. 4 (n = 15) (18 - 27 June) (18 June - 11 July) Impoundment 23 June [+ or -] 3.5 (n = 12) 26 June [+ or -] 7. 8 (n = 14) (19 June - 2 July) (19 June - 13 July) 1993 Pond 23 June [+ or -] 2.3 (n = 63) 26 June [+ or -] 8. 1 (n = 80) (19 June - 1 July) (19 June - 6 August) Impoundment 21 June [+ or -] 2.1 (n = 33) 25 June [+ or -] 7. 7 (n = 42) (18 - 27 June) (18 June - 15 July) [Part 2 of 2] TABLE 4. Nesting chronology of Pacific loons in 1992 and 1993, Prudho e Bay,
Year Hatch date [+ or -] SD (ranges) 1992 18 July [+ or -] 2.8 (n = 13) (15 - 25 July) 19 July [+ or -] 3.8 (n = 10) (15 - 28 July) 1993 22 July [+ or -] 6.9 (n = 32) (14 July - 9 August) 23 July [+ or -] 9.1 (n = 14)
(13 July - 7 August)
Natural ponds tended to produce more young than impoundments (Table 5). However, the proportion of pairs with 0, 1, and 2 young at hatch was not significantly different between the two water body types in 1992 (2 = 0.37, d.f. = 2, p = 0.83) or 1993 (2 = 2.40, d.f. = 2, p = 0.30).
In 1993, water-level drawdowns affected nest success on impoundments but not on ponds. Impoundment drawdowns occurred between 15 June and 15 July. Drawdowns were the result of culvert thaw or drying of recently flooded tundra bordering deeper, more stable pools. At some impoundments, drawdowns were extensive and rapid. At other impoundments, where culverts were absent or ineffective, water levels remained relatively constant, fluctuating slightly with yearly rainfall cycles in a manner similar to that observed at natural ponds in 1993. At the eight impoundments with the largest drawdowns, reductions in surface area ranged from 25% to 69%, and averaged about 48%. Of the 13 nests located on these eight impoundments, only 15% were successful. In 1992, two impoundment nests were predated following large drops in water level.
Nest predation was rarely observed in 1992 or 1993. In 1993, nests where predation was directly witnessed, or where there was circumstantial evidence of predation (i.e., egg remains or fox tracks at the nest), accounted for 31% of failed pond nests and 14% of failed impoundment nests (Table 6). Direct predation was witnessed on two occasions when parasitic jaegers (Stercorarius parasiticus) were observed feeding at nests which were later found to contain shell fragments. From the number of "disappeared" and "predated" clutches in 1993, my best estimate of the overall impact of predators is that they were responsible for the loss of over 95% of failed clutches on ponds (n = 48) and impoundments (n = 28). Predation estimates for impoundments may have been too high because of the difficulty in determining whether nests had been deserted because of water-level drawdowns or other events prior to predation.
Survival of Young
Although chick survival tended to be lower on impoundments than on ponds (Table 5), the proportion of loon pairs with 0, 1, and 2 young was not significantly different at the end of field observations (5 August in 1992 and 27 August in 1993) in either 1992 (2 = 1.81, d.f. = 2, p = 0.40) or 1993 (2 = 2.40, d.f. = 2, p = 0.30). In 1993, the majority of chick mortality on impoundments apparently occurred immediately after hatch. On the first day of observations, 58% (n = 33) of loons pairs at successful pond nests had two chicks, but only 36% of loon pairs at successful impoundment nests had two chicks. Otherwise, the percentage of young that survived to the end of field observations in August was similar on the two water body types in 1993: 71% on ponds and 74% on impoundments (Table 5).
In 1993, Pacific loons commonly dispersed from both natal ponds and natal impoundments (Table 7). Half of the families that moved from nest sites on impoundments (n = 8) and 63% of the families that moved from nest sites on ponds (n = 16) returned to the natal water body at least once prior to the end of observations. Although loons tended to disperse more frequently from small impoundments and ponds than from large ones, there was wide variation in water body size (Table 7).
TABLE 5. Pacific loon productivity on ponds and impoundments in 1992 and 1993, Prudhoe Bay, Alaska. Pond Impoundment 1992(1) 1993 1992(1) 1993 Number of nests found 24 81 22 42 Nest success (%) 58 41 50 33 Number of young at hatch 20 52 16 19 Number of young per nest at hatch 0.83 0.64 0.73 0.45 Number of young in August(2) 15 37 7 14 Hatched young that fledged (%) _ 71 _ 74 Number of young per nest in August(2) 0.63 0.46 0.41 0.33 (1) Because 1992 studies began after clutch initiation, results likel y underestimate predation losses early in the incubation period. Also, 1992 studies ended in early August well before young fledge. (2) Studies ended on 5 August in 1992 and on 27 August in 1993.
Oil field infrastructure (i.e., gravel roads and pads) did not prevent overland movement from either impoundments or ponds. I observed one chick from a pond nest crossing a gravel road, and I have indirect evidence (families observed on opposite sides of roads on alternate visits) that five additional families with flightless chicks (four from impoundments and one from a pond) crossed roads during disperal.
In both 1992 and 1993, the amount of time spent foraging by adults was similar on impoundments and ponds (Table 8). During an average four-hour observation period, adults for-aged longer in 1993 (60% on impoundments and 54% on ponds) than in 1992 (54% on impoundments and 48% on ponds). Foraging times were not significantly different on the two water body types in either 1992 (t=0.73, d.f.=5,p=0.50) or 1993 (t=1.00, d.f.=7, p=0.35).
Although loons on impoundments tended to deliver more prey to young than did loons on ponds (Table 8), prey deliveries per four-hour period were not significantly different between the two water body types in 1992 (t=0.89, d.f.=5, p=0.41) or 1993 (t=0.73, d.f.=7, p=0.49). In both years, prey delivery rates were much more variable on impoundments than on ponds (Table 7). There was no pattern between years (higher at ponds in 1992 and higher at impoundments in 1993) in the percent of time both adults simultaneously delivered prey to young (Table 8).
Loon chicks were fed a diet consisting almost entirely of freshwater invertebrates. Over 99% of prey items delivered in 1992 (n = 5146) and 1993 (n = 23 363) were invertebrates captured in freshwater wetlands. Small fish (approximately 2-4 cm in length) were the only other prey items I observed being delivered to young. A total of 4 fish and 11 fish were delivered by adults in 1992 and 1993, respectively. Fish were brought to the nest site by adults returning from unknown locations, presumably the nearshore Beaufort Sea (where adult loons were often observed foraging) or large inland lakes.
TABLE 6. Fate of failed nests in 1993, Prudhoe Bay, Alaska. Nest Fate Pond (%) Impoundment (%) Disappeared 31 (64.6) 23 (82.1) Predated 15 (31.2) 4 (14.3) Avian 3 2 Fox 2 2 Unknown 10 0 Deserted 1 (2.1) 1 (3.6) Unknown 1 (2.1) 0 (0.0)
Total 48 (100.0) 28 (100.0)
TABLE 7. Dispersal of Pacific loons on ponds and impoundments in 1993 , Prudhoe Bay, Alaska. Ponds Impoundments Successful nests 33 14 Families that dispersed (%) 49 57 Average age of chick when family dispersed (days) 10.9 17.8 Average size of water body from 5.6 [+ or -] 6.23 7.6 [+ or -] 14 .4 which families dispersed (ha) (n = 16) (n = 8) Average size of water body at 24.0 [+ or -] 20.6 10.2 [+ or -] 9 .1
which families remained (ha) (n = 17) (n = 6)
Invertebrate availability apparently plays an important role in nest site selection by Pacific loons. At Prudhoe Bay, loons nested only in wetland types in which the depth of the water is less than 1 m (Hobbie, 1984). Maximum depth in Deep-open Lakes, on which no loons were found nesting in 1992 or 1993, is about 2 m. Wetlands more than 1.7 m deep (the thickness of the ice cover) contain fish (Hobbie, 1984), and therefore have a lower abundance and diversity of invertebrates (Butler et al., 1980).
Selection of shallow water bodies in this study area also may have been related to factors other than food for the young. In addition to food availability, nest pond requirements for common loons include (1) isolation from conspecifics, (2) adequate nest sites, (3) abundant aquatic vegetation, and (4) protection from wave action (McIntyre, 1983). It is possible that many of these requirements are better provided in shallow wetlands than in Deep-open Lakes.
Nest Site Selection
Throughout their range in North America, Pacific loons prefer to nest on islands (Davis, 1972; Bergman and Derksen, 1977), apparently because foxes visit island nests less frequently than mainland nests (Petersen, 1979). Selection of island nest sites may be less adaptable on impoundments than on ponds because of hydrological differences between the two water body types. Water levels in natural arctic wetlands tend to be relatively stable and predictable (Hohman et al., 1992), but impoundment water regimes often are not (Truett and Kertell, 1992). Therefore, there is increased risk that island nest sites on impoundments may become attached to the mainland and shoreline nests may become stranded on dry land as water levels fall during the course of the breeding season.
[Part 1 of 2] TABLE 8. Foraging effort of Pacific Loon pairs in 1992 and 1993, Prud hoe Bay, Alaska. Average Average length of Aver age number of complete time a dults foraging bouts foraging bouts fed y oung Water body Number of per 4-hr per 4-hr per 4 -hr Year type pairs period period (min) period (min) 1992 Pond 6 3.0 ([+ or -]1.1) 50.8 ([+ or -]3 2.3) 113.9 ([+ or -]32.5) Impoundment 6 2.4 ([+ or -]0.6) 78.1 ([+ or -]2 3.6) 129.1 ([+ or -]32.5) 1993 Pond 8 2.9 ([+ or -]0.9) 48.9 ([+ or -]1 4.0) 128.7 ([+ or -]25.2) Impoundment 8 2.7 ([+ or -]0.7) 45.3 ([+ or -]2 2.2) 144.0 ([+ or -]47.3) [Part 2 of 2] TABLE 8. Foraging effort of Pacific Loon pairs in 1992 and 1993, Prud hoe Bay, Alaska. Average percent of time Average Estimated total young fed by number of prey deliveries 1 and 2 adults prey deliveries per 4-hr per 4-hr per 10-min period Year period period (excluding fish) 1992 48 ([+ or -]20), 52 ([+ or -]20) 62.2 ([+ or -]27.5) 665. 1 ([+ or -]282.7) 63 ([+ or -]39), 37 ([+ or -]39) 67.8 ([+ or -]31.3) 853. 8 ([+ or -]414.0) 1993 73 ([+ or -]16), 27 ([+ or -]16) 95.4 ([+ or -]19.3) 1216. 6 ([+ or -]311.4)
40 ([+ or -]27), 60 ([+ or -]27) 100.1 ([+ or -]54.6) 1462.5 ([+ or -]871.5)
Timing of Breeding
Timing of breeding at Prudhoe Bay was similar to that reported at Point Storkersen, Alaska. From 1971 to 1973 at Point Storkersen, dates of clutch inititation on natural ponds were between 20 and 23 June (Bergman and Derksen, 1977).
Water level might be more important than timing of spring thaw in determining breeding dates at Prudhoe Bay. In a study of common loons, Barr (1986) found that adults waited for specific nest sites to become available as water levels receded. Although impoundments were available earlier than ponds at Prudhoe Bay, loons that selected territories on impoundments nested at the same time as those using ponds. Like the common loon, they might have been waiting for spring water levels to stabilize before nesting.
Schamel and Tracy (1985) have suggested that the ability of loons to produce replacement clutches is a valuable adaptation to occasional high rates of arctic fox (Alopex lagopus) predation in western Alaska. However, in northern Alaska young loons from late clutches often do not fledge before lakes and ponds freeze (Johnson and Herter, 1989). Thus, replacement of clutches at Prudhoe Bay may be less adaptive than at subarctic locations.
In both 1992 and 1993, predation by arctic foxes apparently was the major cause of clutch loss on both impoundments and ponds in this study. Foxes typically remove entire eggs from the vicinity of the nest before eating or caching them (Tinbergen, 1972; Kertell, pers obs.), and it is likely that they were responsible for the majority of empty nests. However, glaucous gulls and common ravens (Corvus corax) will sometimes fly with whole eggs before eating them (Davis, 1972; Petersen, 1979; Kertell, pers. obs.) and cannot be totally discounted as a potential cause of empty nests.
It was impossible to determine if loss of most clutches was the result of direct predation or of scavenging following nest desertion. Stranded clutches (clutches left on dry land following water-level drawdowns) are less accessible to incubating loons and more difficult to defend against foxes (Gotmark et al., 1989). For common loons, Barr (1986) found that 33% more clutches hatched successfully in territories with water-level fluctuations of less than 1.5 m. Belant and Anderson (1991) reported nest failure when receding water levels increased the distance of the nest to water by 3 m. Although flooding was a major cause of arctic loon nest failure in northern Sweden (Gotmark et al., 1989) and Scotland (Mudge and Talbot, 1993), I observed no instances of loon nest loss due to rising water levels at Prudhoe Bay. Pacific loons at Prudhoe Bay do not begin to initiate nests until around the third week of June, after peak water levels have occurred.
The degree to which reproductive failure was the result of human disturbance at Prudhoe Bay could not be determined. Impoundments are located immediately adjacent to oil field roads and drilling pads, where the potential for human disturbance is greater than for the majority of natural ponds. Temporary abandonment of nests following disturbance likely results in increased opportunities for predation by glaucous gulls and parasitic jaegers (Gotmark et al., 1990), which are common at Prudhoe Bay (Kertell, pers. obs.). I observed workers displace loons from their nests at Prudhoe Bay on two occasions, but neither instance resulted in loss of eggs or nest abandonment.
Survival of Young
Since Pacific loons almost always lay two eggs (Johnson and Herter, 1989), chicks on impoundments apparently suffered higher mortality immediately after hatch, either from higher rates of partial predation or from starvation, than did chicks on ponds. At other arctic locations, most chick mortality apparently resulted from starvation of the youngest chick in the brood. Davis (1972) and Bergman and Derksen (1977), for example, found that the youngest chick often died within two weeks of hatch because of competition for food with its older sibling. A chick weakened by lack of food is especially vulnerable to hypothermia and fatigue during this period (Alvo, 1986). Therefore, although prey delivery rates at Prudhoe Bay were not significantly different between impoundments and ponds, there may have been differences between the two water body types in the type or nutritional quality of prey.
In this study, avian predators likely were responsible for most chick predation at ponds and impoundments in both 1992 and 1993. Glaucous gulls, for example, are known to prey heavily on young birds (Johnson and Herter, 1989; Barry and Barry, 1990). In northwest Canada, the glaucous gull was the only predator species seen taking the young of red-throated loons (Dickson, 1993). Predation on red-throated loon chicks was apparently limited to those less than 10 days old, and the greatest loss of young occurred in a year when there was a significantly large number of glaucous gulls in the study area (Dickson, 1993).
In 1993, production rates at Prudhoe Bay were within the range of production rates (0.17-0.41 young/territorial pair) reported for arctic loons (see Mudge and Talbot, 1993). However, arctic loon production was based on number of nearly fledged young per territorial pair rather than on young per nesting attempt. I was unable to determine the percentage of territorial birds that failed to initiate nests.
At Prudhoe Bay, dispersal from impoundments and ponds may have been related to food availability. Davis (1972) found that the average size of 35 territories at McConnell River, Northwest Territories, was 3.66 ha, considerably larger than the average size (2.22 ha) of nest ponds there. Consequently, 57% of territories included more than one pond, and birds that nested on small ponds often moved to nearby ponds to feed.
For visual predators like loons, water body selection may represent a balance between food abundance and foraging efficiency (Heglund et al., 1994). Eriksson (1985) considered water transparency to be a factor important in the selection of lakes by arctic loons, and suggested that their use of oligotrophic lakes in southwest Sweden was consistent with the need for high water transparency to maintain an adequate prey capture rate. Based on similar average prey delivery rates at impoundments and ponds, prey detectibility appears to have been similar in the two water body types in my study. I can provide no reasons for greater variability in prey delivery rates on impoundments than on ponds.
Invertebrate prey choice could not be determined at Prudhoe Bay. Invertebrate prey items reportedly taken by Pacific loons breeding in the Arctic include fairy shrimp (Order Anostraca), tadpole shrimp (Order Notostraca), and caddisfly larvae (Order Trichoptera) (Davis, 1972; Bergman and Derksen, 1977). Among these and other potential prey, probably only fairy shrimp and large cladocerans occur at densities high enough to account for the high rates of prey delivery (over 300 prey per 10-minute period) observed by loons subsurface dabbling in open water at Prudhoe Bay. Fairy shrimp are frequently clumped in dense swarms (Stross et al., 1980) and are high in protein content (71.9% of dry weight: Driver et al., 1974; Krapu and Swanson, 1975). In 1992, I collected a sample of large cladocerans (Eurycerus lamellatus, 4 mm long) at two water bodies, and, because they also occur at high densities (Stross et al., 1980), the presence of these organisms may have resulted in high prey delivery rates.
TABLE 9. Monthly mean temperatures ([degrees]C) at Deadhorse, Alaska in 1992 and 1993. Weather data were provided by the Alaska State Climate Center, University of Alaska, Anchorage. Year April May June July August 1992 -17.4 -5.9 3.6 8.3 6.7 1993 -14.2 -4 3.7 10.3 6.2
Difference (1993 - 1992) +3.2 +1.9 +0.1 +2.0 -0.5
Greater prey delivery rates at both impoundments and ponds in 1993 than in 1992 may have resulted from greater invertebrate production in 1993. At Barrow, Alaska, chironomid larval growth was more rapid at water temperatures of 10[degrees]C and 15[degrees]C (Butler et al., 1980). Although I did not measure water temperatures, monthly air temperature readings averaged higher in 1993 than in 1992 in all months but August in the Prudhoe Bay area (Table 9).
Foraging observations during this study do not support the conclusion of Andres (1993) that Pacific loons rely on marine resources (i.e., marine fish or invertebrates) to the same degree as red-throated loons, which are known to feed their young exclusively fish (Bergman and Derksen, 1977). Andres' conclusion was based on the observation that red-throated and Pacific loons nested at equal densities and were observed flying north to marine waters at the same frequency in his study area on the Colville River Delta, Alaska.
Adult loons in my study did not usually leave the nest pond to capture food for the young, but may have been foraging to supplement their own diet. I frequently observed both Pacific and red-throated loons flying northward from the breeding area toward marine waters and judged it likely that many of the loons performing these flights were breeders, considering that often only one adult was present at the nest pond during chick feeding. However, if adults were foraging in the Beaufort Sea, only on 15 occasions in 1992 and 1993 did they return to nest sites with fish. In Canada, sticklebacks (Pungitius pungitius) and grayling (Thymallus thymallus) were among fish prey captured by Pacific loons during the breeding season (Davis, 1972).
Particulate composition and organic content of the sediments are correlated with distribution and abundance of certain invertebrates (Wetzel, 1983). Even if benthic habitats in impoundments contain higher levels of peat and drowned tundra vegetation than benthic habitats in ponds, these differences do not appear to have significantly influenced foraging efficiency. Loons in my study foraged successfully in both impoundments and natural ponds despite any potential differences in substrate type.
It is reasonable speculation, based on observations made during this study, that impoundments increase the abundance and diversity of resource patches potentially available to breeding Pacific loons. However, although the importance of wetland diversity to breeding waterbirds is widely recognized (Swanson et al., 1979; Weller, 1988, 1990), it is not known whether or not the loon population at Prudhoe Bay is limited by available foraging habitat. Furthermore, it is not known whether loons that nest on impoundments would normally have nested on ponds, or whether the presence of impoundments provides nesting opportunities for birds that would not have nested otherwise.
The principal cause of nest failure on both impoundments and natural ponds appears to have been predation by arctic foxes during incubation. Predation losses on impoundments, however, were often associated with large water-level drawdowns, which apparently increased opportunities for predation of stranded clutches. Considering the high degree of site fidelity exhibited by some loon species (McIntyre, 1974; Alvo, 1986; Strong, 1990), it is possible that many Pacific loon pairs at Prudhoe Bay will return in the future to impoundments with large water-level fluctuations and that nest success will continue to be lower than at natural ponds, even though impoundments are otherwise productive and suitable for nesting.
Depending on the management objective, the presence of some impoundments may be consistent with wildlife management goals on the Coastal Plain of Alaska. For Pacific loons and waterfowl that nest on islands or pond shorelines, overall reproductive output may be improved by maintaining stable water levels on those impoundments that have the largest water-level drawdowns and a history of previous nesting. At other impoundments, the needs of waterbird species, particularly shorebirds and dabbling ducks, may be best served by allowing drawdowns to occur and increasing the availability of invertebrate-rich sediments. I found no evidence to suggest that draining impoundments would improve habitat value for loons (or invertebrate-eating waterfowl species) at Prudhoe Bay.
This study was funded by BP Exploration (Alaska) Inc. (BPX). I thank Chris Herlugson (Supervisor, Environmental Assessment, BPX) and Michelle Gilders (Environmental Scientist, BPX) for their interest and support. I am grateful to John Luginbuhl, Mark Miller, Jeff Wickliffe, Carlos Paez, and Steve Johns for assistance in the field. Special appreciation is extended to John Luginbuhl for managing the loon study in 1993 during periods when I was in Anchorage. I am grateful to Bob Pollard and Bob Rodrigues for help with logistics. Joe C. Truett contributed valuable insight during early discussions on impact assessment research. Joe C. Truett, Lynne Dickson, Stephen Johnson, and Lynn Noel reviewed and improved the final manuscript. The study was conducted while I was employed by LGL Alaska Research Associates, Inc., Anchorage, Alaska.
|Printer friendly Cite/link Email Feedback|
|Date:||Dec 1, 1996|
|Previous Article:||Can regulation of freshwater runoff in Hudson Bay affect the climate of the North Atlantic?|
|Next Article:||Method to determine lake depth and water availability on the north slope of Alaska with spaceborne imaging radar and numerical ice growth modelling.|