Microclimate of the nest and egg water loss of the Eider Somateria mollissima and other waterfowl in Spitsbergen H E R M A N N R A H N . JOHN K R O G AND FRIDTJOF M E H L U M Rahn. H . . Krog, J . & Mehlum, F. 1983: Microclimate of the nest and egg water loss of the Eider Somareria mollissima and other waterfowl in Spitsbergen. Polar Research I n s . 171-183. Temperature gradients in the nest and within the egg. nest humidity as well as eggshell conductance and rate of egg water loss of the Eider Somareria mollissrma and other uaterfowl were studied at Ny-Alesund. Spitsbergen (78%" latitude). These studies suggest a specific interrelationship between eggshell con- ductance and maintenance of an appropriate temperature and humidity environment of the nest. resulting in an egg water loss rate which is optimal for hatching success. I n spite of low ambient temperatures of less than 3°C and very low absolute humidities of less than 4 torr (similar to those found in hot deserts). the nest's microclimate and rate of mater loss were similar t o those reported for nests and eggs in temperate climates. Hermann Rahn. Department of Physiolog)). Srare Uniuersiy of N e w York. Buffalo, A'. Y . 13214. L ' . S . A ; John Krog. Deparrmenr of Zoophysiology, Unruersiy of Oslo. Blindern. Oslo 3 , Norway; Fridrlof Mehlum. Norwegian Polar Research Institure. Rolfstangueien 12, 1330 Oslo Lufrhaun. Sorway: Nouember I982 (accepred May 19831. R Introduction While much has been written about the distri- bution and nesting behavior of birds in the Arctic, little if anything is known about the microclimate of nests of birds that colonize particularly the farthest reaches of the polar region. What are the egg temperatures and nest humidities that such birds can maintain during incubation under severe climatic conditions encountered in the high arctic area (79" North), and how d o these compare with nesting conditions in more southern latitudes. Irving & Krog (1956) examined nest air tem- peratures of arctic birds in Alaska (68" North) to see whether the eggs were exposed to lower or more variable temperatures than in temperate regions. Using thermocouples affixed to the nest they determined the nest air temperature in seven incubating species and found that the median incubation temperature of 49 records was between 33°C and 3YC, and in 74% of the records, between 33°C and 37°C. These values were similar to those obtained by Huggins (1941) in the Cleveland, Ohio, area (41" latitude); Hug- gins found a mean temperature among eggs of 37 species to be 34°C. S.D. 2.4. Irving & Krog con- cluded that 'during the periods of incubation and brooding the growth of birds proceeds at well- regulated temperatures in the changeable weather of temperate and arctic climates. The regulation of temperature during incubation and brooding is accomplished by the behavior of the parent birds'. In this study we examined the microclimate of the nest in three species of waterfowl nesting in Kongsfjord o n Spitsbergen (79" latitude). during the latter half of June (see map. Fig. 1). The daily mean air temperatures were considerably lower than those in the Alsaka study and did not exceed 4"C, while the mean ambient absolute humidity was below 4 torr (mm Hg), equivalent to an air moisture content of 4 m g l - I . Not only did we examine the temperature gradients within the eggs as well as the nest and the absolute humidity of the nest's microclimate. but also the egg water loss and eggshell conductance. Most of the infor- mation was obtained for the eggs and nests of the Eider (Somateria ntollissima), with additional data from the nests of the Barnacle Goose (Branfa leucopsis) and the Long-tailed Duck (Clangula hyernalis) and the eggs of the Pink-footed Goose (Anser brachyrhynchus) . 172 Hermann Rahn, John Krog and Fridtjof Mehlum -7 j I LOCATION M A P P Fig. I. Location map. the nest. This was of sufficient length so that the incubating parent would not leave the nest when readings were made at various times of the day. Temperatures were read from a Kane-May Tem- perature Recorder. As each nest was visited, the various leads were individually plugged into this hand-held digital read-out device. On one occa- sion egg temperatures of Eider eggs were meas- ured by direct insertion of a thermistor into the egg at various levels. Readings at the top, center, and bottom of the egg were made in succession within 30 seconds of the female leaving the nest. Brood patch temperature. - A small piece of Styrofoam plastic was mounted on top of the egg to serve as insulation from the remainder of the egg. A small thermocouple was then mounted on top of the insulation to come in contact with the skin of the brood patch. The leads from this thermocouple were affixed with wax to the outside of the shell as shown in Fig. 2. Nest air temperature. - These were recorded con- tinuously in the nest of the Long-tailed Duck using a thermistor which was mounted as shown in Fig. 2. Body temperature. -Female Eiders were captured using a 4 m-long handle net on the nest. While one person held the body, the other opened the bill and inserted a wooden bit to lodge across the angle of the bill so that it stayed open. A small Methods Temperature measurements Egg tempernture. - Copper-constantan thermo- couples were secured at various levels in either fresh Eider or chicken eggs. To prevent such eggs from turning in the nest, they were secured on a small board with epoxy resin as shown in Fig. 2 . The thermocouple leads exiting beneath the board were buried and extended 10-15 m beyond Fig. 2. Schematic section through Eider egg showing locations of thermocouples and their leads brought out through a hole in the wooden platform which serves to keep the egg in the upright position once placed in the nest. Thermocouple mounted on a Styrofoam insulation was used to measure the brood patch temperature. Its leads were attached with wax to the outside of the egg. Next to the egg is mounted a thermistor for air temperature recordings. On the right is shown the wooden bit in the angle of the Eider's bill. A polyethylene tube (outside diameter 3 mm) is advanced through a hole in the bit down the esophagus a distance of 36cm to the proventriculus. The thermocouple initially retracted is then advanced 3 m m beyond the end of the tube for the temperature reading. Nest microclimate of the Eider 173 E g g humidity. - The water vapor pressure in the eggs provides the pressure head for the flux of water vapor which leaves the egg during incu- bation. This vapor pressure is essentially satu- rated and thus can be estimated from the egg temperature using standard water vapor pressure-temperature tables. hole in this bit allowed a third person to advance a ( 3 m m o . d . ) polyethylene tube into the prov- entriculus (see Fig. 2). After placement a 0.9 mm thermocouple junction inside the tube was advanced 3 mm beyond the terminal end of the tube and the temperature read on the Kane-May Temperature Recorder. The whole operation from the time of capture to the temperature read-out was accomplished in less than one minute. Nest and ground temperatures. - Temperatures at the bottom of the down layer and a few cm below the down layer were measured with thermocouple junctions in Eider nests from which the female had been flushed. The depth of the compressed down layer and the ground beneath the nests were also measured. Ambient air temperatures. - These were daily averages taken from the official log of the local weather station located with 0.5 km radius of the Eider nests. Humidity measurements Nest humidity. - Egg diffusion hygrometers were prepared as previously described (Rahn et al. 1977). Briefly, chicken eggs were emptied, cleaned, dried, and filled with a desiccant, silica gel. They were then calibrated by exposure to a constant water vapor pressure in a sealed con- tainer. The daily gain in water vapor was meas- ured and provided the necessary calibration of the diffusive eggshell conductance expressed in mg HzO day-' torr-' (1 torr = 1 mm Hg vapor pressure). Once calibrated, the eggs were weighed just prior to insertion into the nest and again 2-4 days after removal from the nest. From the average daily weight gain in the nest and the eggshell conductance, the mean water vapor pres- sure can then be calculated, since PN = &f;IH20/G~z0, where PN = the mean water vapor pressure in the nest, torr; =the mean daily weight gain of water, mg day-'; G = the shell conductance, mg day-' torr-'. The chicken egg hygrometers were readily accepted in all the nests. Ambient humidity. - The absolute ambient humidity was calculated from the mean air tem- perature and the relative humidity values of the weather station records, and is also expressed in terms of the water vapor tension, torr. Weight changes of eggs during incubation These were determined at various intervals using Pesola scales. Egg and shell dimensions Initial egg mass. - Since eggs lose appreciable amounts of water during incubation, the air cells were displaced by water in order to establish the initial egg mass; this technique was recently vali- dated (Grant et a]. 1982). Weight measurements were made on a torsion balance to within + 3 mg. Initial egg density. - This value was calculated from the egg dimensions provided by Schonwetter (1960), as recently reported by Rahn et al. (1982). Surface area. - This value was derived from the allometric relationship between surface area and egg mass (Paganelli et al. 1974). Length and width. - These were measured using calipers as previously described (Rahn et al. 1976). Shell weight. - After eggs were emptied the shell was dried, first in air and then for at least one day in a desiccator before weighing on a torsion balance within + 3 mg accuracy. Shell thickness. - This was measured to three significant figures with a ball point micrometer. Twenty-one measurements from representative areas were taken and averaged for each egg. Pore counts of shell Shell fragments were boiled in a 2.5 KOH solution to remove all proteinaceous material. They were then etched in concentrated nitric acid for several seconds, the time depending upon the thickness of the shell. The etching process enlarges the pores so that painting the inside of the shell with a water soluble dye will draw the solution to the 174 Hermititi Rahti, John Krog utid Fricirrof Mvhlitrn outside rim of the pores T h e pores in each of twenty 1/4 cm' areas h e r e counted and averaged. and the total number per egg calculated on the basis ot its surface area The average coefficient of variation of the 20 pore counts for each egg I S about 2 0 7 Egg rliell r o t i c i i ~ t m i c e This was measured according to the method of Ar e t al. (1974). Eggs were placed at constant room temperature in desiccators with silica gel. The daily weight loss was established and divided by the saturation water vapor pressure at that temperature to yield the eggshell conductance to water vapor expressed i n mg H:O day-' torr ~ ' _ Results Teniperatitres Body temperature of ten female incubating Eiders was determined at about 2400 on 2 July. The values ranged from 38.2 t o 395°C. with a mean value of 39.1"C. S.D. 0.4. T h e brood patch tem- peratures had a mean value of 38.5"C in Eiders and 37.7"C in the Barnacle Goose (Table 1). The width of an Eider egg is 5 cm (Table 5 ) and the implanted thermocouples were situated 0.5 cm from the top, i n the center. and 0.5 cm above the T d d r 2 Local temperatures in 12 Eider nests. S . D . in parenthe5es. Bottom of Gravel-sand down layer below down Distancc helow egg or below down. cm 3 . 3 (0.8) 4.1 (1.8) >lean temperature. "C 20.4 (3.8) 10.3 ( 3 . 3 ) bottom of the shell. From Table 1 i t can be seen that a 7.6"C temperature difference was observed over a distance of about 4 c m . T h e thermocouple in the center of the egg registered a mean value of 33.6"C about halfway between the top and the bottom thermocouples, and indicates a linear temperature gradient over the extreme distance. It is of interest to note that by direct insertion of a thermistor into six Eider eggs the mean top to bottom temperature difference was only 4.1"C. The implanted thermocouples in the Barnacle Goose egg showed a mean difference of 5.3"C. A t the bottom of the down layer, 3.3 cm below the bottom of the egg, the temperature had fallen to 20.4"C. In the gravel-sand substrate 4.4cm below the down, the temperature averaged 10.3"C (Table 2 ) . T h e ambient mean daily air temper- atures during the period average about 4°C. Humidities The absolute humidities of the nest's microclimate Tubk i. . \ . l e a s m m c n t of brood patch and egg temperatures h! implanted thermocouples ~n thc eggs of Eider (between 25 June and J Jul! and the Barnacle Goose ( h e r w e n 30 Junc and 3 Jul! I . Egg ccmperatures Here also ohtained in Eider eggs h y direct insertion ot '4 thermistor into top. center. and bottom of egg ( 2 8 J u n c ) Brood Egg Method of pdtch Top Centei Bottom meawrement EiJcr Previously 29.7 implanted thcrmocouplcs 7 0.8 6 h Insertion of 31 9 thermistor 11 ~ Previousl) implanted So nests 1 1 I 6 - Barnacle Sn. obserLation5 6 6 GO<)% hlcan temp . "C 37 7 35.9 - 30.6 S . D . 0 7 11.7 1.2 ~ thcrmocouplss Nest microclimate of the Eider 175 T a h k 3. Absolute humidity values in nest of Eider, Barnacle Goose. and Long-tailed Duck as well as ambient absolute humidit! and temperature values. Date Absolute humidity Nest Ambient Ambient renip. Hyg. N o . (torr) (torr) ("C) Eider 27-30 June ( 2 nests) 3C1 June-3 July 7 6 18.0 4.2 3.6 20.3 4 . 3 4.3 Mean 1Y.2 4.3 4.0 14-17 June 17-20 June 2&23 June 23-27 June 27-30 June Barnacle Goose (1 nest) 30 June-2 July 3 18.7 4.0 1 .4 6 20.6 3.7 1.9 1 14.1 4 . 2 1 7 5 16.0 4.0 1 . 1 9 16.9 4.2 3.6 2 17.6 3.3 4.3 Mean 17.3 4 . 1 1 5 S D 2.2 0.2 I ? Long-tailed +13* J u n e Duck 13-16 June (1 nest) 1 6 2 0 June 1 10.9' 2 15.7 5 14.0 Mean 14.9 3 I " 1.8' 3.8 2.0 3.8 1.7 3.8 1.9 * Incubation had not started during this period. Values omitted from mean. for the three species are shown in Table 3, giving the date during which a particular egg hygrometer was in the nest. In the Barnacle Goose nest we have a continuous record of nest humidity over an 18-day period, with remarkably little change and maintaining an average value 13 torr higher than the absolute ambient humidity. During six days of integrated records of the Eider nest the nest humidity was maintained 15 torr above the ambient humidity. T h e nest humidity of the Long-tailed Duck during seven days of recording was somewhat lower than for the first two species, maintaining an 11 torr difference above the ambient humidity. Table 4. Average weight losses of eggs during incubation Weight loss of eggs The mean weight loss for Eider eggs was 0.59 g/day. Assuming this t o b e a n average throughout incubation, this would result in 25 days in a net loss of 14.8g or 14% of the initial egg mass of 103 g (Table 5 ) . F o r the Barnacle Goose the daily weight loss of 0.57 g for 24.5 days of incubation would result in a net loss of 1 4 g , or 13% of the initial mass of 109 g. I n the Long-tailed Duck the average value after the initiation of incubation was 151 mg day-'. However, this mean rate dur- ing the first 15 days does not reflect the progressive increase in the daily rate from 100 to 1 8 0 m g day-' during four successive weighings during this Interval Weight loss No. nests No. eggs (days) (mg.day-') SE Eider Barnacle Goose Long-tailed Duck 8 1 1 27 4 4 19 6 15 590 32 572 69 151 3 176 Hermann Rahn, John Krog and Fridtjof Mehlum period, and thus does not provide a good basis for an overall estimate of incubation water loss. Egg dimensions These are shown in Table 5 and agree within 3% or less with the values for egg mass, length, width, shell mass, and shell thickness provided by Schonwetter (1960). The indicated typical clutch size and incubation times are taken from Johns- gard (1978). Shell conductance and pore dimensions The mean values for water vapor conductance of the shell, G , the pore length or shell thickness, L, and the number of pores In each egg are shown in Table 6. Since water vapor transport across the shell obeys Fick's first law of diffusion, the shell conductance is proportional to the total pore area, A,. and inversely proportional t o the length of the pore, L. (Paganelli 1980). Substituting the measured values for G and L allows one to cal- culate A, = 0.447 G.L (Rahn et al. 1976; Paga- nelli 1980) as shown in Table 6. Compared with the surface area of the egg (Table 5 ) , A, is very small; for example, in the Eider it is equal to (3.7/10,700) or 0.03% of the egg's surface area. All gas exchange during incubation is limited to this small fraction of the shell's surface. When the total pore area, A,, is divided by the total number of pores, N , one obtains the average cross sectional area of the individual pores as well as the pore radius, assuming that all pores have uniform diameter throughout their length. The actual shapes of the pores for the Eider, Pink- footed Goose, and Barnacle Goose are drawn to scale in Fig. 3. They were copied from the pore casts made by Tyler (1964). According to Tyler (1964) and Board et al. (1977), examination by light and electron microscopy reveals that these pores are partially obstructed by the remains of the cuticle and organic spherules which fill the upper cup of the pores. This is shown diagra- matically in the center of the lower row in Fig. 3. To the left are three pore casts of the Eider and to the right are the calculated dimensions. Since this calculation is based on a functional performance, namely, the conductance, the derived dimensions are smaller than the actual pores becawe of the obstruction and therefore one must speak of an effective pore area and an Fig. 3. Pore casts reproduced from Tyler (1964). 1. Pink-footed Goose, 2. Barnacle Goose, and 3. Eider. In the lower row are shown from left to right three pore casts of Eider, a diagram- matic cross-section through pore showing the organic spherules which obstruct the pore cup according to Tyler (1964) and Board et al. (1977). On the right are the calculated effecfiue pore dimensions for Eider eggs. See text. effectiue pore radius (Paganelli 1980; Board 1982). Our shell conductance measurements of Eider eggs were done on freshly laid eggs before incu- bation had started. To our surprise these values were very much lower than one would have expected (Ar & Rahn 1978) and lower than those that H.R. had previously obtained in incubated Eider eggs in Alaska and Canada. We believe that these low conductance values were due to the presence of cuticular covering plugging the pores which had not yet been removed by pro- Nest microclimate of the Eider 177 longed contact with the brood patch. Similar dif- ferences in conductance value were observed between unincubated and incubated eggs of the Long-tailed Duck. For this reason the shell con- ductance values for the Eider listed in Table 6 are based on the average value of nine eggs pre- viously measured. As far as we know this is the first observation in non-passeriform eggs that the water vapor conductance will change after incu- bation is initiated. That such changes occur in the small eggs of passeriform birds has been pre- viously reported by Carey (1979), Hanka et al. (1979), and Sotherland et al. (1980). Discussion Body temperatures Our mean value of 39.1"C for the incubating Eider is similar to the 'abdominal temperature of the sitting Eider' of 39.6"C reported by Rol'nik (1970). Kossack (1947) reported a 'cloaca1 tem- perature for incubating Canada Geese' of 40.9"C (n = 2) and Caldwell & Cornwell (1975) men- tioned a 'deep body temperature' of 41°C for the Mallard. Egg temperatures Outer surface measurements. - By attaching tem- perature recording devices to the outer surface of natural eggs during natural incubation large temperature differences have been recorded Table 6. Shell conductance, pore length (= shell thickness), pore number per egg, and calculated total pore area and individual pore dimensions. Standard error in parentheses. Conductance mg.day-'.torr-' (GI Eider 21.6. Pink-footed Goose 29.5 Barnacle Goose 22.7 Long-tailed Duck 11.6 (0.8) (1.8) - (0.4) Measured Pore length mm (L) 0.381 (0.007) 0.451 0.460 0.271 (0.004) (O.oa8) - Number of pores (N) Calculated Total pore area Individual pores (AP) Area Radius mm2 mP2 mlc 14,266 16,277 11,927 9,373 (684) (544) (939) - ~ ~ 3.7 260 9.1 6.0 368 10.8 4.1 394 11.2 1.4 149 6.9 * See text 178 Herniann Rahn, John Krog and Fridtjof Mehlum between the top and bottom surfaces. Both Rol'nik (1970) and Drent (1970) have reviewed the literature. For example. Burke (1925) meas- ured these differences daily in five nests of the chicken throughout the whole incubation period, while Drent (1970) made similar observations in the Herring Gull (Larus argentatus). In both species the top of the egg under the brood patch maintained remarkably stable temperatures between 39.2 and 39.5"C, while the bottom sur- face of the eggshell gradually increased in tem- perature during the incubation period from about 30 to 34°C in the chicken and from about 27 to 35°C in the Herring Gull. In the Kittiwake (Rissa tridactyla). the t o p of the egg maintains a fairly constant temperature of 38°C throughout most of the incubation period, while the nest's surface below the egg (not the bottom surface of the egg) was approximately 11°C lower (Barrett 1980). Temperature differences within the egg. - A few of these have been reported in natural eggs during natural incubation and are given in Table 7 includ- ing o u r own observations. T h e egg width of each species is also given. allowing o n e to calculate the approximate temperature gradient within the egg, A"C/cm. As may be noted. these vary greatly and a r e not necessarily related to egg size but more likely reflect the particular stage of embryonic development. In our study the thermocouples were implanted in freshly laid eggs with no appreciable heat production of the embryo nor chorioallantoic circulation. Such eggs might be expected to have the largest gradients of l°C/cm or more as shown in Table 7. I n the other reported studies the age of t h e embryo, the degree of chorioallantoic development. and the possible disturbance of the normal metabolism and cir- culation by implantation of t h e thermocouple are unknown quantities. Such factors would influence not only the absolute temperature but also the temperature gradient within the egg. Central and air cell egg temperatures. - Egg tem- peratures frequently reported are assumed to be central egg temperatures. The average value reported by Drent (1975) for 16 non-passeriform species is 35.8"C. S . D . 1.1. O n e approach is to record the air cell temperature, which does not interfere with the normal development of the embryo. Caldwell & Cornwell (1975) measured air cell temperatures in Mallard eggs (Anas playrhynchos) during natural incubation. During the early stages these temperatures were about 33 t o 34°C and during the last week of incubation averaged 38"C, S . D . 1.2, with a n overall average of 36.3"C. S . D . 2.5, for t h e whole incubation period. Similar studies in t h e Canada Goose (Branta canadensis) gave values of 38.5"C. S.D. 0.7 (Kossack 1947) and 37.9"C, S.D. 1.1 (Cooper 1978). T h e latter study also showed a gradual increase in air cell temperature as incubation pro- ceeded as well as a marked circadian rhythm, with lowest temperature occurring between 0400 and 0500 and the highest value between 1400 and 1600 hours, with a mean difference of 3.0"C. Table 7 . Temperature differences within natural eggs during natural incubation. their egg width and calculated temperature gradients. A"C;cm. Temperatures. "C Egg width Temp. grad. Ref. T o p Center Bottom cm AT/crn 37.3 33.6 29.7 5 . 1 1.5 30.6 5.1 1 .0 34.6 6.9 0.3 Eider - /Somareria mollissimaJ Barnacle Goose - 35.9 - /Branta kitcopsis) Laysan Albatross 1 36.8 - (Diomrdea mmurabilisi Royal Tern '. 39.2 37.8 - 4.4 0.6 (Srerna maxima) Kittiwake 3 38.0 36.4 - lRissa rridacFla) Great Egret 36.9 34.3 - (Cavmerodrur alhrcc) Cattle Egret 2 37.5 36.8 - /Buhu/rus ibis) 1 4.2 0.8 4 1 1.3 3.3 0.5 7 References: 1 . Grant. G . tpers. comm.). 2. Vleck et al. 1983. 3. Barrett 1Y80. Nest microclimate of the Eider 179 "C 39.1 38.5 37.3 33.6 29.7 20.4 10.3 Fig. 4. Composite of various temperatures measured to show the heat gradients in the Eider's nest. A distance scale is on the left. Nest air temperatures. - This is the temperature found between the eggs. These were measured continuously over a 5-hour period in the nest of the Long-tailed Duck during the first week after clutch completion and varied between 33 and 34°C. As pointed out by Drent (1970, 1975), nest air temperature of the Herring Gull increases gradually over the first ten days of incubation and during this time may actually exceed the central or air cell egg temperature. Measurements of nest air temperatures for other waterfowl have been reported for Anus acuta, 35.9"C (Irving & Krog 1956), for Mergus serrutor, 36.9"C (Barth 1949), and for the Domestic Goose. 33.4"C (Koch & Steinke 1944). Lomholt (1976) measured the air temperature and relative humidity 'below the eggs' of Eiders and reported a value between 27 and 29°C. which is close to our bottom values inside the egg, namely, 29.7"C (see Fig. 4 and Table 7). Nest air values for other non-passeri- form birds have been reported by Drent (1975), and for 16 species these average 34.8"C, S.D. 1.3. Overall temperature gradient. - In Fig. 4 we have combined our various temperature measurements for the nest and the egg of the Eider. The scale on the left indicates the spatial arrangement and on the right the continuous temperature gradient. These nests were located on a moraine close to the bay and the daily average air temperature varied between 3 and 4°C. It is interesting to note that the temperature gradient of the egg, as pre- viously noted, was approximately lS"C/cm, while the gradient across the 3.8 cm of down as well as across 4cm of soil was similar, namely, ca. 2 . PC/cm. When these temperatures are compared with those of other waterfowl nesting in more tem- perate climates there are no essential differences, thus confirming the observations of Irving & Krog (1956) that the arctic environment does not com- promise normal egg and nest temperatures in these species. As far as our egg temperatures are concerned, it is important to bear in mind that these represent the situation during the early stages of incubation. As the chorioallantoic cir- culation becomes established, it will have a great effect on redistributing the heat more evenly throughout the egg, thus diminishing the large gradients seen in Fig. 4. The chorioallantois begins to make contact with the inner shell mem- brane about day 6 in the chicken and finally envelops the whole shell by day 12, that is, after only 60% of the incubation period has been com- pleted. It is at that time that the rapid growth of the embryo begins to contribute to the heat bal- ance of the egg. It is therefore not until the last quarter of the incubation that one would expect to see temperature gradients within the egg reach their minimal values. As we shall see. these are important considerations for predicting the mean temperatures of the egg, since they determine the water vapor pressure head responsible for the incubation water loss. Nest humidity The rate of water loss from the egg is equal to the product of the shell conductance and the water vapor pressure difference between the egg and that of the nest air. The vapor pressure of the nest air is therefore an important factor in the regulation of water loss. Since arctic climates have a very low absolute humidity or vapor pressure, it was of particular interest to see how nest humidity in the Spitsbergen area compared with that of nests in more humid and temperate climates.In Table 8 we have compared our aver- age values with those obtained in the nests of other waterfowl nesting in temperate regions. 180 Hermann Rahn, John Krog and Fridtjof Mehlum Table 8. Absolute nest humidities (torr) in nests of waterfowl measured with egg hygrometers. The values reported for Lomholt (1976) and Koch 8. Steinke (1944) were calculated from their measurements of relative humidity and temperature of nest air. Nest humidity (torr) No. obs. Reference - Eider 19.2 2 (Somateria mollissima) 21.0 - Lomholt (1976) Barnacle Goose 17.3 6 (Branra leucopsisl - - Long-tailed Duck 14.9 2 (Clangula hyemalrrJ Domestic Goose (Anser sp. 1 Snow Goose (Anser caerulescens) Greylag Goose f A n s e r anser) Egyptian Goose fAIopochen aegyptiacus) Mallard Duck ( A n a s platyrhynchos) 14.0 24.2 22.3 19.2 17.4 Koch & Steinke (1944) Rahn et al. (1977) Rahn et al. (1977) Rahn et al. (1977) Rahn et al. (1977) There appears to be no essential difference and this suggests that these arctic nests are so well insulated against the dissipation of water vapor through the nesting material and the protecting by 24.5 days of incubation (Johnsgard 1978) pre- dicts a total weight loss of 14g, or 13% of their initial mass of 109 g. feathers that values characteristic of temperate climates are maintained. Relation between egg water loss, egg temperature and nest humidity To ensure a given incubation water loss of the Egg water loss during incubation egg, the shell conductance must be exposed to a Although the rate of water loss is now known for finite nest temperature and humidity. During more than 80 species (Ar & Rahn 1980). it development, the oxidation of solids and the for- includes only one species of waterfowl, namely, mation of metabolic water will each contribute the Eider. Hagelund & Norderhaug (1975) stud- to an appreciable increase in the relative water ied the weight loss of Eiders in Spitsbergen over content of the eggs unless it is removed by dif- a period of 20 days. In seven broods (33 eggs), fusion through the pores. Knowing the metabolic the average weight loss was 625 mg day-'. rate throughout incubation, one can calculate that Extrapolated over their reported incubation to achieve a relative water content at the end of period of 24.3 days this represents 14.5% of their incubation equal to that of the fresh egg, about initial egg mass of 104.5 g. Belopolski (1961) 15% of the initial egg mass must be lost as water. reported the weight loss of eleven Eider eggs in Recent measurements of the relative water con- the Barents Sea area with an initial mass of 102 g. tent of precocial eggs have shown that indeed this Over a 25-day incubation period these lost 16.6 g, value is the same (72%) for hatchlings and the an average daily weight loss of 660 mg day-'. fresh egg in spite of a water loss equal to about These values may be compared with our value of 15% of the initial egg mass. O n the basis of these 590mg day-' over a four-day period (Table 4). observations Ar & Rahn (1980) have proposed Assuming this to represent the average value over that a particular incubation water loss is man- a 25-day period of incubation, their total loss datory for optimal hatching success. would be 14.8 g, or 14.3% of their initial mass of We may now consider the nest conditions pre- 103g, very close to the average value reported requisite for incubating eggs losing the proper for birds in general (Ar & Rahn 1980). The Bar- amount of water. The rate of egg water loss is a nacle Goose eggs over a 19-day interval lost on physical phenomenon dependent on the escape the average 572 mg day-'. Multiplying this value of water vapor by diffusion across the pores of Nest microclimate of the Eider 181 the shell. It can be described as follows (Rahn et al. 19761: MH20 = Gegg (PA - PN)H2O (1) where MH,O = rate of water loss, mg day-', Gegg = water vapor conductance of the shell, mg day-' torr-', PA = water vapor pressure in the egg, torr, and PN = water vapor pressure in the nest air surrounding the egg, torr. Substituting into equation (1) the values for the Eider, where MH10 =590mg day-' (Table 4) and G,,= 21.6mg day-' torr-' (Table 6), one obtains the water vapor pressure difference (PA - PN) which must have existed between the egg and nest air and is equal to 27.3 torr. Since we also know the vapor pressure of the Eider nest, PN = 19.2 torr (Table 3) we can add these two values (19.2 + 27.3) to obtain the water vapor pressure of 46.5 torr that must have existed in the egg, PA. The water vapor pressure in the air spaces of the membranes beneath the shell is essentially satu- rated, and thus one obtains from this value the average temperature of the egg, which according to the vapor pressure-temperature tables is 36.8"C. Similar calculations for the eggs of the Barnacle Goose predict a temperature of 35.2"C. Both of these values are similar to the central or Vapor Pressure 5o PH20rtorr 7 4 0 10 20 30 Y 40 Fig. 5. Simultaneous values of temperature and absolute humidity values found in the egg of the incubating Eider, A , the nest air, N , and the ambient environment, 1. Curve on the left is saturation water vapor pressure at various temperatures equivalent to 100% relative humidity. Curves to the right are isopleths for relative humidity values of 80, 60, 40, and 20%. (For discussion see text.) air cell temperature of other waterfowl eggs dis- cussed above as well as the average egg temper- ature value for non-passeriform eggs in general, namely, 35.8"C (Drent 1975). In Fig. 5 are shown the water vapor tempera- ture coordinates where the left hand curve rep- resents saturation water vapor pressure (= dew point or 100% relative humidity). The other curves represent isopleths for relative humidity of 80, 60, 40, and 20%. Point A represents the air space benetah the shell which is saturated at 46.5 torr at 363°C. Point N represents the vapor pressure of the nest = 19.2 torr. This point was arbitrarily placed at a temperature of 34"C, an average value for nest air temperature (Drent 1975). However, this point can be moved to the right or the left along the 19.2 vapor pressure isopleth when better values for nest air temper- ature are eventually obtained. The exact location is of importance only if one wants to obtain a more reliable value for calculating the relative nest humidity. At present, this is shown at ca. 50%. For our purposes, the important difference is the vapor pressure difference between the egg and the nest (PA - PN), where PA is determined by the egg temperature and PN by the nest con- struction, feather cover, and brooding behavior of the parent. For a given value of (PA - PN), the rate of water loss depends on the magnitude of the shell conductance, Gegg. By rearranging equation ( l ) , (PA - PN) = M H ~ o / G ~ ~ ~ as shown on the right ordinate of Fig. 5. In the case of the Eider, the required conductance is established in utero long before incubation begins by the for- mation of 14,000 pores which have an average length of 370 pm and an effective radius of 9 pm (Table 6). Nest conductance Water vapor released from the egg cannot accu- mulate in the nest without raising its vapor pres- sure. It must therefore escape by diffusion and/ or convection across the nest and feather barrier. In the steady state the rate of water vapor escape from the nest must equal the escape rate from the egg, assuming that additional release of water vapor from the brood patch is negligible. Thus one can write an equation for water loss from the nest, where 182 Hermann Rahn. John Krog and Fridtjof Mehlum where MHiO =rate of water loss for each egg from the nest and is assumed t o be equal t o the rate of water loss from the egg. mg day-', G,,,, = the water vapor conductance of the nest, assumed to be by diffusion only. mg day-' torr-', PN = water vapor pressure of the nest, torr, and PI = water vapor pressure of the ambient environment, torr. The mean ambient vapor pressure. PI, was equal to 4.0 torr. Substituting this value in equa- tion (2) allows one to obtain the water vapor conductance of the nest per egg. G,,,, = 590/ (19.2 - 4.0) = 39 mg day-' torr-'. nearly twice the value of the egg conductance. O n the right ordinate of Fig. 5 the bracket indicates the (PN - PI) difference which is determined by the ratio of MH20/G,,,,. Comparison of dry and moist climates In view of a prescribed incubation water loss it is of interest to compare the effects of a dry climate such as Spitsbergen's with a moist climate of the tropics o n the development of nest and eggshell conductance. Point I in Fig. 6 described the ambient atmosphere, PI. at Spitsbergen at YC, relative humidity 7096, and an absolute Vapor Pressure, torr 7 4 5 0 40 30 20 10 0 0 ' 0 20 30 'C 4 C Fig. 6. Comparison of the total water vapor pressure difference ( P a - P , ) in dr!-cold a n d in moist-warm climates. (For discus- sion 5ee text.) humidity of 4 torr. During t h e nesting season the ambient atmosphere of Enewetak Island in t h e Pacific has a mean temperature of 28"C, relative humidity 80%, and a n absolute humidity of 24 torr (Fig. 6) ( R a h n e t al. 1976). Each value, 4 and 24 torr, represents the final sink for water vapor diffusion. Assuming in each region a n egg temperature of 36°C equivalent to an egg vapor pressure of 44 torr, then t h e total pressure dif- ference (PA - PI) available for t h e dissipation of water vapor is (44 - 4) o r 40 torr in the dry climate of Spitsbergen and (44 - 24) or 20 torr in the moist climate of Enewetak as illustrated in Fig. 6. In each case this pressure difference must b e divided between that acting across the eggshell and that acting across the nest-feather barrier, since PA - PI = (PA - PN) + (PN + PI). Since the flux rate of water vapor is assumed t o be the same for the egg and the nest, equations (1) and (2) can be substituted and rearranged to yield P A - P I - 1 ~ 1 MHXJ Gqg Gnes, and since 1/G = the resistance, R (3) -- PA - PI - Re,, + R,,,, M H : O Thus for a given water loss rate in both climates, (Re,, + R,,,,) is half as large (and G twice as large) in the moist climate as in the dry. This does not mean that the eggshell resistance in the dry climate is necessarily half as large (or Geg, twice as large), since the particular value of the Regg will also depend on t h e changes in R,,,, (equation ( 3 ) ) , and t h e ratio of Reg, to R,,,, will vary with each species. However, it emphasizes that each climate imposes a finite total water vapor pressure difference. PA - PI (PA deter- mined by the egg temperature and PI by the climate), and that the adaptive shell resistance depends upon the nest resistance, as their sums must equal (PA - P[)/&0 according t o equation (3). Egg resistance in each species is determined by the number of pores, their length and cross- sectional a r e a , while a given nest resistance is achieved by the nest construction and the brood- ing behavior and attentiveness of the parent. Acknowkdgemenrr. - These studies were carried out in June and J u l y 1982 at the Norwegian Polar Research Institute at Ny-Alesund. Svalbard. and were supported in part by the Bell Nest microclimate of the Eider 183 Fund and the Distinguished Professor Fund of the State Uni- versity of New York at Buffalo to Hermann Rahn. The authors take pleasure in acknowledging the help of many others who assisted in these studies, particularly 0ivind Tsien. References A r , A , , Paganelli, C. V . , Reeves, R . B . , Greene, D. G. & Rahn, H. 1974: The avian egg: water vapor conductance, shell thickness and functional pore area. Condor 76, 153- 158. Ar. A. & Rahn, H . 1978: Interdependence of gas conductance, incubation, length, and weight of the avian egg. Pp. 227-236 in J. Piiper (ed.) Respiratory Function in Birds, Adult and Embryonic. Springer Verlag. Ar, A . & Rahn. H . 1980: Water in the avian egg: Overall budget of incubation. A m . Zool. 20, 3 7 s 3 8 4 . Barrett. R . T . 1980: Temperature of Kittiwake Rissa tridacpla eggs and nest during incubation. Ornis scandinauica 11, 50-59. Barth, E . K. 1949: Redetemperaturer og rugevaner. Naturen (Bergen) 73, 81-95. Belopolski, L. 0. 1961: Ecology of sea colony birds of the Barents Sea (Transl. from Russian edition, 19.57). Israel Prog. Sci. Tranrl., Jerusalem. Board, R . G . 1982: Properties of avian egg shells and their adaptive value. Biol. Reu. 57, 1-28. Board, R. G . . Tullett, S . G . & Perrott. H. R . 1977: An arbitrary classification of the pore systems in avian egg shells. J. Zool. Lond. 182, 251-26s. Burke, E . 1925: A study of incubation. Monr. Agric. Exp. Sra. Caldwell. P. J. & Cornwell, G . W. 1975: Incubation behavior and temperatures of the Mallard Duck. Auk 92, 706731. Carey, C. 1979: Increase in conductance to water vapor during incubation in eggs of two avian species. J. Exp. Zool. 209, 181-186. Cooper. J . A . 1978: The history and breeding biology of the Canada Geese of Marshy Point, Manitoba. Wildlife Mono- graphs No. 61, 1-87. Drent, R. H . 1970: Functional aspects of incubation in the Herring Gull. Behauiour, Suppl. 17. 1-132. Drent, R . H . 1975: Incubation. pp. 334-420 in Farner. D . S . & King. J. R . (eds.): Avian Biology Vol. 5 . Academic Press, N.Y. Grant, G . S . , Paganelli, C. V.. Pettit, T . N. & Whittow, G. Bull. 178, 1-44. C. 1982: Determination of fresh egg mass during natural incubation. Condor 84, 121-122. Hagelund, K. & Norderhaug, M. 1975: Studies of population changes and breeding processes in a colony of Eiders (Somateria mollissima (L)) in Svalbard. Norsk Polarinstitutt Arbok 1973, 141-162. Hanka, L. R . , Packard, G . C . , Sotherland. P. R . , Taigen, T. L . , Boardman, T. J . & Packard, M. J . 1979: Ontogenetic changes in water-vapor conductances of eggs of yellow- headed blackbirds (Xanthocephalus xanrhocephalus) . 1. Exp. Zool. 210, 1 8 s 1 8 8 . Huggins, R . A . 1941: Egg temperatures of wild birds under natural conditions. Ecology 22, 148-157. Irving, L. & Krog, J . 1956: Temperature during development of birds in arctic nests. Physiol. Zool. 29, 195-205. Johnsgard, P. A . 1978: Ducks, Geese. and Swans of the World. Univ. Nebraska Press, Lincoln, Nebraska. 404 pp, Koch, A . & Steinke, L. 1944: Uber Temperatur and Feuchtig- keit bei Natur und Kunstbrut. Arch. Kleintierzucht 3 , 153- 203. Kossack, C . W. 1947: Incubation temperatures of Canada Geese. J . Wildlife Manage. 11, 119-126. Lomholt, J . P. 1976: Relationship of weight loss to ambient humidity of bird eggs during incubation. J. Comp. Physiol. 105. 189-196. Paganelli, C. V. 1980: The physics of gas exchange across the avian eggshell. A m . Zool. 20, 329-338. Paganelli, C. V . , Olszowka, A . & Ar, A . 1974: The avian egg: surface area, volume and density. Condor 76, 319-325. Rahn, H . , Paganelli, C. V., Nisbet, I . C. T. & Whittow, G. C. 1976: Regulation of incubation water loss in eggs of seven species o f Terns. Physiol. Zool. 49, 245-259. Rahn, H . , Ackerman, R. A . & Paganelli, C . V . 1977: Humidity in the avian nest and egg water loss during incubation. Phys- iol. Zool. 50, 269-283. Rahn, H . , Parisi. P. & Paganelli, C. V. 1982: Estimating the initial density of birds’ eggs. Condor 84, 339-341. Rol’nik, V. V. 1970: Bird embryology (Transl. from the Russian edition, 1968). Israel Prog. Sci. Transl., Jerusalem. Schonwetter, M. 1960; Handbuch der Oologie. Meise. W. (ed.). Akademie Verlag. Berlin. Sotherland, P. R . , Packard, G . C. & Taigen, T . L. 1980: An altitudinal cline in conductance of Cliff Swallow eggs to water vapor. Auk 97, 177-185. Tyler, C. 1964: A study of the egg shells of the Anatidae. Proc. Zool. S O C . Lond. 142, 547-583. Vleck, C. M . , Vleck, D . , Rahn, H . & Paganelli, C. V 1983: Nest microclimate, water vapor conductance and water loss in heron and tern eggs. Auk 100, 76-83.