young. We measured the quantity and quality of milk produced by female house mice (Mzs musculus) dluring a period of 28 days after birth of a litter. We aimed ...
J. Zool.,Lond.(1988)216,195-210
Maternal care in house mice (Mus musculus);II. The energy cost of lactation as a function oflitter size Ba n s A.n A .Kö NI c*, J. Rtn s r E n aN n H . Ma nrr Fakultätfür Biologie, (Jniuersityof Konstanz, Postfach 556A,D-7750 Konstanz, Federal Republic of Germany (Accepted 25 January 1988) (With 4 figures in the text) In mammals, lactation performance may influence future reproduction of a mother and her young. We measured the quantity and quality of milk produced by female house mice (Mzs musculus) dluring a period of 28 days after birth of a litter. We aimed to analyse how females cope with the energy demands of different sized litters as a tradeoff between current and future reproduction. Litter sizesexaminedwere small (six young), intermediate (7'3 * 0'7), and large (12 young). Females met the energy demand of a growing litter both by increasing the amount of milk given to the young and by improving the quality (through an increase in total solid and fat concentrations)to a peak during days 9-16. At the onset ofweaning (day 17), milk production decreasedand young shifted to solid food. The main energy source in the milk of house mtce was fat, which provided more than 80% of the energy to the suckling young. Average lipid concentrations werc 20ok, and peak values of 33Vn were reached during days 13 16. Protein concentrations were 6 8% and carbohydrate (lactose) concentrations 3 4%u. With larger litters, females increased both the amount of milk and the absolute amount of major nutrients (lipids, proteins, lactose). However, regulation was imperfect. When litter size was doubled from six to 12, amount ofmilk produced roseby 44ok, and energycontent only rose by 30%. The reduction in milk supply for individual young in larger litters was reflectedin slower growth and lower weaning weight. The efficiency of conversion of milk energy into biomass of young was highest for intermediate litters (51%; for small litters,43%; for large ones, 36%). During days 5 16, a single pup in an intermediate litter needed lessenergy for metabolism and growth than pups in the other litter sizes examined. Although individual young of small litters have a relatively high weaning weight which might improve their future reproduction females gain higher reproductive success by dividing the energy available per litter between the largest number ofyoung they can raise to a weaning weight of on average9 g. For the life-time reproductive successof a female house mouse-at least for the strain used in this study-an intermediate litter size of sevenseemsto be best becauseof a favourable ratio of energy cost of lactation to number and size of young produced.
Contents Introduction Methods Milk i n g t e chnique ..
..
Page 196 . . 197 .. 191
*Present address:ZoologischesInstitut, University of Basel, Rheinsprung 9, CH-4051 Basel, Switzerland
19 5 09s2-8369/88/0l0l9s + 16$03.00
@ 1988 The Zoological Society of London
196
B AR BAR A
KÖ NI G,
J. RIESTER AN D H . M ARKL
M i lkan al yse s. . .. Body weight changesofyoung during suckling Results Body weight and food consumption . . . . .. .. Q u an ti tyo fm ilk Quality of milk Body weight changesofyoung during suckling
D isc us si on .. .. Amount and quality of milk produced Energy budget of growing young Energy cost of lactation and reproductive success R efe re nce s.. ..
197 198 199 199 t99 200 202 203 203 206 201 208
Introduction A striking characteristicof mammals is that femalesfeed their young with milk. Via the milk, offspring are provided with calories, nutrients, vitamins and minerals for growth and for metabolism (Peaker, 1977).As a consequence,a long lactation period always should be profitable for the offspring. For the mother, on the other hand, the cost of reproduction may influence her survival and her fecundity. A neg?tive correlation between current and future reproduction is a basic element of some theories of life-history evolution (Williams, 1966; Stearns, 1976; see Reznick, 1985for a review ofthe evidence).It is not yet known to what extent the energycosts of nursing (or breeding) reflect the reproductive costs (Clutton-Brock, 1984), but lactation performancedoesinfluencelater reproduction in somemammalian species(Fuchs, 1981;CluttonBrock, Guinness& Albon, 1982;Trillmich, 1986). In rodents, the energy cost of lactation has usually been evaluated by measuring food consumption and body weight of the female and her litter, or by measuring metabolism and the gross chemical components of the body of young at birth and at weaning (Kaczmarski, 7966; Migula, 1969; Millar, 1978, 1979; Studier, 7979; Lochmiller, whelan & Kirkpatrick, l9B2; Mattingly & McClure, 1982).In housemice, energeticcostsof lactation are more than four times higher than energeticcosts of gestation (Myrcha, Ryszkowski & walkowa, 1969). Especiallyfor small animals, the metabolic demandsof lactation are enormous.In an animal the sizeof a mouse, daily energy output in milk per unit body weight is approximately l6-fold higher than in an animal the size of a cow (Hanwell & Peaker, 1977). A high and sustained milk production in small animals is possible becauseof a high metabolic rate which also relates to a decreasein life span (Pearson,1948;McNab, 1983). A high reproductive output in housemice can be interpreted as an adaptation to a colonizing life strategy which has to cope with variable environmental conditions and high mortality (Bronson, 1984;König & Markl, 1987). In this study, we analysedthe quantity and quality of milk produced during lactation in house mice with small, intermediate, and large litters, to investigatehow femalescope with the energy demandsof different sizedlitters as a tradeoffbetweencurrent and future reproduction. The study therefore extends and refines Fuchs' (1982) work, in which body weight oiyonng was used as a correlate to the amount of milk produced. However, female mammals could further change the quality of milk and thus invest different amounts of energyper g milk as a function of age and size of their young.
E N ERG ET IC S O F LAC T AT I ON IN HOUSE MICE
197
Methods Experimental animals came from an inbred strain of housemice (Mus musculus;BALB/c/Han). Pregnant lemales were separatedfrom their mates and kept in Macrolon cages(22 x l6 x 14 cm) at 2l "C in a 12:12 L:D cycle (light on at 08:00). Food pellets (Altromin 1310)and water were offered ad libitum. The day of birth of a litter corresponded to day I of lactation. All females used for experimental analysis were multiparous; we used their second or third litter. Experimental litter sizeswere eithersmall (6 young; n:69 different litters), intermediate(7 9 young, mean: 7.3 +0.7; n:22litterd), or large (l2young;n:26litters). In most cases,number of young correspondedto the actual litter size at birth. To achieve a sufficient number of litters with 6 or l2 offspring, some litters were reducedor enlargedby adding same-agedyoung at day 1 of lactation. Femalescaring for manipulated litters were never milked during the first 4 days to allow for adaptation of milk production according to the number of sucklingyoung.Ofthefemales,2T%(44)weremilkedtwice(afteranaverageof ll'4+4'l days;seeTable I). This had no effecton weaning weight of young. All other femaleswere milked only once during lactation. Body weight of females and of offspring, and food consumption (amount of food disappearing lrom the food rack in24h, minus 10% estimated food loss in bedding material) were evaluated daily to the nearest o '1 g .
Milking
technique
Lactating femaleswere isolated from their litters at 09 : 00 and milked at 13: 00 to evaluatethe amount of milk produced drrring a 4-h period. Longer separation can result in decreasingmilk production becauseof lack of stimulation by the offspring (Bateman, 1957;Hanwell &Lin2e11,1972).Milk production was assumed not to change significantly dtring24 h (Bateman, 1957).Females were anaesthetizedby an intraperitoneal injection of 100mg ketaminhydrochloride and 4.6 mg xyladine per kg body weight, followed by an injection of I unit oxytocin. The mice were milked with an apparatus developedby Hoffmann, Sawatzki, Schmitt & Kubanek (1982). One milking sessionlasted 10 15 min, until all available milk was removed from every mammary gland. The milk collection head of the milking apparatus was weighed before and after milking to measure the amount of milk produced (mg). Five (20%) of the femaleswith 12 young died during, or shortly after, milking. In thesecases,milk samples only were used for qualitative but not lor quantitative analyses.There were no lossesin the other litter size groups. The number of femalesmilked is siven in Table I.
Milk
analyses
Milk sampleswere stored at - 16'C until the following ingredients were analysed: -Total solids: gravimetrically (milk sample dried to constant weight at 102'C. Sample size: 50 mg), -Lipids: gravimetrically (Baverstock, Spencer& Pollard, 1976. Sample size: 50 mg). Carbohydrates(lactose)'.titrimetrically (Folin & Denis, 1917-1918.Sample size: 50 mg). Proteins'.photometrically (Bradford, 1976. Sample size: l0 mg). -Calcium'. by atomic absorption spectrophotometry (sample size: 10 mg). -Iron'. by atomic absorption spectrophotometry (sample size: 10 mg). Calcium and iron concentrations were evaluated only for small and large litters. A minimum of 180 mg milk was needed to carry out all analyses.In most of the samples,milk yield was sufficient to al1owfor double testing (preferencewas given to lipid and protein analyses).In caseof double testing, the mean values are used. The numbers of samplesanalysed are given in Table I.
198
BA R BARA
K ÖN IG,
J. RIESTER AND
H. MARKL
Ta s L r I Number of milk samples taken from lactating females, and number of different milk samplesanalysedfor total solids (TS ) , lipid- (L), carbohydrate- (C), and protein- (p) contents, as a .function of litter size and age of young (days). Numbers of females milked twice are giuen in brackets; mean number of days between two milking se.rslons..small litters, l2.I +4.0 d; intermediatelitters, 10.2+4. 1d; large litters, ll.9-t3.5 d Litter size (range)
Small 6 Days
Intermediate 7.3+ 0.7 (7e)
Large l2
Number of milk samples
1, 4 58 9 12 13 16 11-22 23 28 Total
l8 (0) 20 (0) l4 (3) 14(4) l4 (8) 7 (3) 87 ( 18 )
5 (0) 6 (2)
5 (2)
e (0) l1 (0) 6 (2) 7 (:6) 6 (s) 3 (3)
32(10)
42 (16)
s (1) 6 (3) s (2)
Number of samplesanalysed for:
Days TS L t 4 17 58 t 7 9 12 t 2 1 3- r 6 1 0 17-22 t2 2 3 28 5
C
16 t5 t7 1 5 t2 I t 8 10 t2 l1 5 36
P
TSL CP
1'/ 19 1l 13 12
2 5 3 2 3 5
53 66 55 o-J
54 43
TSL
599 6 lt 555 666 566 533
C
P
89 11 11 1l 65 57 66 IJ
Differences in the composition of sampleswere tested with non-parametric tests. Correlations between groups of data were evaluated by the Spearman-correlation coemcient (Sachs, 1978). Differences were regarded as significant for 2-tailed error probabilities P
