PHYSIOLOGY AND REPRODUCTION. Development of the Reproductive System in Turkeys with a High or Low Susceptibility to Prolapse of the Oviduct

Transcription

1 PHYSIOLOGY AND REPRODUCTION Development of the Reproductive System in Turkeys with a High or Low Susceptibility to Prolapse of the Oviduct S. Buchanan, G. W. Robertson, and P. M. Hocking 1 Roslin Institute, Roslin, Midlothian EH25 9PS, United Kingdom ABSTRACT Lines of turkeys selected for rapid growth ventral ligament between the two strains that could predispose and high meat yield have an increased incidence of prolapse of the oviduct compared with unselected or traditional strains of turkeys. The development of the reproductive system and changes in plasma estrogen concentrations were compared in sire line and traditional turkeys with the aim of identifying any morphological or hormonal differences that could be associated with the high incidence of prolapse in the male line. Four turkeys from each strain were killed weekly from 0 to 7 wk postphotostimulation, and samples from prolapsed birds were obtained from field cases. There were no differences in the rate of development of the ovary, oviduct, uterus, vagina, sphincter ani muscle, or muscular cord of the the sire line to prolapse. Histological investiga- tion of the uterus, vagina, muscular cord of the ventral ligament, and sphincter ani muscle 5 wk postphotostimulation in traditional, sire line, and prolapsed sire line turkeys did not reveal any differences that could be associated with prolapse. No prelay peak in plasma estradiol concentration was observed in either strain, and there was no evidence to suggest that plasma estradiol was higher in the sire line compared with the traditional turkeys. It was concluded that prolapse of the oviduct in sire line turkeys was not associated with any anatomical abnormalities or high plasma estradiol during reproductive development. (Key words: turkeys, prolapse, oviduct, ovary, estradiol) 2000 Poultry Science 79: INTRODUCTION Prolapse of the oviduct is common in females of sire line breeding turkeys that have been heavily selected for rapid growth and high meat yield. Prolapse can result in the loss of up to 10% of a sire line flock, whereas it is rare in dam lines selected for egg production and traditional unselected turkeys. Prolapse is most common around the onset of lay, about 4 to 5 wk after the onset of photostimulation (N. A. French, 1997, British United Turkeys Ltd., Chester CH40EW, UK, personal communication). Prolapse of the oviduct is characterized by eversion of the vagina and protrusion through the ventral opening. The rest of the oviduct gradually moves out of the body cavity through the vent, and the condition worsens over 1 to 2 d. Affected birds are always culled. The etiology of prolapse in turkeys is not understood. Hocking (1993) compared egg weight, pelvic width, vent diameter, and abdominal fat pad weight of sire line and dam line turkeys at the onset of lay. No differences in relative size were found, and it was concluded that these factors were not involved in the higher incidence of prolapse in the sire line. Lilburn and Nester (1993) compared the development of the oviduct in two strains of turkeys, one selected for body weight (sire line), and the other selected for egg production (dam line). They showed that, whereas the oviduct of their dam line had almost reached full weight at first egg, the oviduct of their sire line continued to increase in weight throughout the first 49 d of lay. Their results led them to suggest that the oviduct of the sire line might not be as reproductively mature as the oviduct of the dam line. If the oviduct of the sire line is relatively immature at the onset of lay, then it is possible that it is not sufficiently developed to cope with egg production and could predispose the strain to prolapse. Melnychuk et al. (1997) also compared the development of the reproductive system in sire line and dam line turkeys. They looked at the changes in oviduct weight and ovary weight for 6 wk following photostimulation and at first egg, and hypothesized that the weights of these organs at first egg were the fully developed weights. They suggested that although in the dam line the oviduct and ovary reached mature weight at the same time, in the sire Received for publication September 7, Accepted for publication May 24, To whom correspondence should be addressed: paul.hocking@ bbsrc.ac.uk. Abbreviation Key: D = hours of dark; L = hours of light; RMS = residual mean square. 1491

2 1492 BUCHANAN ET AL. line the oviduct achieved mature weight 3 d later than the ovary. From these results, they proposed that ovulation in the sire line starts before the oviduct is fully mature and hypothesized that this was responsible for the greater number of unreconciled ovulations observed in the sire line. Rao et al. (1985) investigated prolapse in a New Hampshire strain of chickens prone to prolapse of the oviduct. They showed that the ventral ligament of the oviduct was more prominent in hens with prolapse. Histological techniques were used to show that this muscular cord contained muscle and collagen fibres. They found no obvious ventral ligament in the control hens. From these results, they hypothesized that over-development of the ventral ligament caused increased torsion at the uterovaginal junction of the oviduct, and that this could impede the passage of an egg down the oviduct and result in prolapse. It was thought that estrogens might be involved with prolapse in domestic hens, because administration of estradiol caused vent feminization and intestinal prolapse in cockerels (Wheeler and Hoffman, 1948). More recently, it was suggested that plasma estradiol was low in prolapsed hens (Shemesh et al., 1982) and that estradiol injections aided recovery from prolapse (Shemesh et al., 1984), whereas higher concentrations of estradiol hindered recovery from prolapse (Shore et al., 1984). Estradiol administration stimulated degradation of uterine collagen in pregnant rats (Pastore et al., 1989; Bienkiewicz et al., 1996) and production of collagen-degrading enzymes in uterine cell cultures from pregnant guinea pigs (Rajabi et al., 1991). The residual ovary is the major source of estradiol in birds (Etches, 1996), and was five times heavier in the sire line compared with the traditional turkeys (Buchanan et al., 1999). It was hypothesized that estradiol from the ovary stimulated vaginal collagen degradation and predisposed the sire line to prolapse. Plasma estradiol rises in hens as they become sexually mature, and peaks 2 to 3 wk before the onset of lay (Senior, 1974). Bacon et al. (1980) measured plasma total estrogen and estradiol concentrations in turkeys from 1 wk prior to photostimulation through the laying period. They found that total estrogens increased from photostimulation, peaked at first egg, and then levelled off during lay, but that there was no change in plasma estradiol concentration. We have previously shown that plasma estradiol was lower in sire line turkeys compared with traditional turkeys 5 wk after photostimulation, when prolapse is most common (Buchanan et al., 1999). These results suggested that a high plasma estradiol concentration was not involved in predisposing the sire line to prolapse. 2 British United Turkeys Ltd., Chester CH40EW, UK. 3 Bred from a flock maintained at the Roslin Institute, Roslin EH259PS, UK. 4 Inverclyde Biologicals, Strathclyde ML43JU, UK. 5 Euthatal Veterinary Drugs Company Ltd., Falkirk FK29HJ, UK. However, it is possible that a high prelay peak in estradiol in the sire line before 5 wk of photostimulation could stimulate vaginal collagen degradation and predispose the strain to prolapse. The aim of this experiment was to investigate the development of the reproductive system from photostimulation until after the onset of lay in a sire line and a traditional strain of turkeys. We wished to identify any differences in the anatomy and physiology of the lines that could underlie the predisposition of the sire line to prolapse of the oviduct. Specifically, we tested the hypotheses that the oviduct of the sire line was not fully mature at the onset of lay and that there was a difference in the structure of the ventral ligaments of the two strains. Plasma estradiol concentration was measured to determine whether there was a prelay peak in plasma estradiol in turkeys, and to compare the plasma estradiol profile of the two strains from photostimulation and throughout the onset and early period of egg laying. Samples from contemporary sire line turkeys that had prolapsed were obtained for comparison with contemporary turkeys that had not prolapsed from the same commercial flock. In addition to these analyses, the data from Melnychuk et al. (1997) on the growth of the ovary and oviduct in sire line and dam line turkeys were re-analyzed to compare the results of the two studies. MATERIALS AND METHODS One-day-old poults were obtained from the Big 5 sire line 2 and the traditional, unselected Nebraska Spot (traditional) 3 line of turkeys. At photostimulation, 32 birds from each line were allocated at random to be killed for autopsy at one of eight ages. The traditional turkeys were reared in eight pens measuring 3.6 m 2, littered with wood shavings, and containing a suspended feeder and drinker. The sire line birds were hatched at the same time as the traditional birds and reared among a large commercial flock of turkeys, from which prolapsed sire line turkey females were also obtained. Both lines of turkeys were subject to similar management conditions. The photoperiod was 14 h light:10 h dark (14L:10D) from 1 to 18 wk, 8L:16D at 18 wk, and 14L:10D at 29 wk and 4 d (the commercial age for photostimulation of the sire line). The birds were fed standard turkey breeding rations throughout the experiment. Starting from photostimulation, each wk for 8 wk, four birds were randomly selected from each strain. A sample of blood was obtained from the wing vein of each bird, transferred to a 5-mL heparinized blood collection tube, 4 and placed on blood tube rollers. The birds were then killed with an overdose of sodium pentobarbitone. 5 The blood samples were centrifuged for 10 min at 2,000 g, and the plasma was removed and stored at 20 C. The body weight of each bird was recorded. The abdominal cavity was opened, and the muscular cord of the ventral ligament was dissected and weighed. The oviduct was removed, and its length and weight were recorded. The uterus and vagina were removed from the oviduct and weighed. The ovary was dissected from the

3 REPRODUCTIVE DEVELOPMENT AND PROLAPSE IN TURKEYS 1493 abdominal cavity, any follicles greater than 0.5 g were detached, the numbers of normal and atretic mature yellow follicles were counted, and the normal follicles were weighed. The residual ovary weight was also recorded. The sphincter ani muscle (Harvey et al., 1968) was removed from the outside of the vent, and the surrounding skin and fat were dissected from it before it was weighed. Samples were taken for histological investigation of the uterus, vagina, muscular cord of the ventral ligament, and sphincter ani muscle of turkeys 5 wk after photostimulation. Samples were also taken from four sire line turkeys that had prolapsed 5 wk after photostimulation. The samples from the uterus and the ventral ligament were pinned to balsa wood to keep their shape. All samples were placed in individual Tissue Tek Mega-cassettes 6 and fixed in buffered neutral formalin. They were subsequently processed using a Shandon Hypercentre XP tissue processor 7 and embedded in paraffin wax. Sections were cut, mounted on albumin-prepared slides, and stained with Hemotoxylin and Eosin or Van Gieson s stains. The estradiol concentration of the plasma samples was measured by a double antibody radioimmunoassay for estradiol, as described by Webb et al. (1985). Antibodysepharose extraction was used to extract the estradiol, because the high fat content of laying turkey plasma can interfere with the assay. The samples were measured in a single assay to eliminate inter-assay variability. The lowest sensitivity of the assay was 8 pg/ml and the intraassay coefficient of variability was 12%. Changes with age in the mean lengths and weights of the reproductive structures of the two strains were compared using regression analyses. A range of nonlinear models was evaluated, and the one with the smallest residual mean square (RMS) was selected to describe the data for a specific trait. The best models for the results presented here were the logistic and exponential curves. The logistic curve was described by the equation y = C + A/(1 + exp[ R(x M)]), where A represented the upper asymptote; C, the lower asymptote; R, the rate of exponential decay of growth; and M, the point of inflection. The exponential curve was described by y = A + BR x, where A represented the upper asymptote; B, the scaling factor; and R, the rate of exponential decay of growth. The logistic curve is symmetric, and the point of inflection gives an indication of the time when half of the growth is complete. The rate parameter in the exponential is a measure of the rate of development of the trait over time. Statistical comparisons of these parameters are a measure of the similarity of the rate of development in the two lines. Values for plasma estradiol concentrations were transformed to natural logarithms to ensure normal distribution of the residuals. Analysis of variance was used to 6 Diagnostics Division, Miles Inc., Elkhart, IN Shandon Scientific Limited, Runcorn, WA7 1PR, UK. 8 alw.nih.gov/pub/image. test for significant effects of strain or week (time). The graphs of Melnychuk et al. (1997) were scanned into TIFF files, and the original data were reconstructed using NIH Image Analysis software. 8 Multiple regression analyses were used to compare growth curves of the ovary and oviduct in the sire and dam lines represented in their experiment. Gross Anatomy RESULTS The sire line turkeys were significantly heavier (16.0 ± 0.31 kg) than the traditional turkeys (5.4 ± 0.07 kg). Figures 1a to 1d show the changes in weights of the oviduct, uterus, vagina, and residual ovary in the traditional and sire lines from 0 to 7 wk of photostimulation. A logistic curve gave the best model (smallest RMS) for oviduct weight, residual ovary weight, and weight of the uterine and vaginal sections of the oviduct in both lines. The fitted curves are shown in the figures; the parameters of the curves and their standard errors are given in Table 1. Analyses of each of these five traits showed a significant change (P 0.001) in the RMS when the linear parameters (upper and lower asymptotes) were fitted separately for each strain, but there were no significant changes when separate nonlinear parameters (point of inflection and rate parameter) were examined. There was no significant strain difference in the point of inflection for oviduct weight, residual ovary weight, or weight of the uterus and vagina. The point of inflection was greater in both strains for ovary weight compared with oviduct weight, but there was no difference between the traditional and sire lines. The point of inflection was lowest for the vaginal weight in both strains, which suggests that the vagina developed faster than the uterus, residual ovary, and total oviduct. The mean number of mature yellow follicles (>8 mm) in the traditional and sire line turkeys and the changes in weight of the muscular cord of the ventral ligament for 7 wk after photostimulation are presented in Figure 1e and 1f. Exponential models were fitted to the weekly means for the number of mature yellow follicles and ventral ligament weight. The estimates for the parameters of the exponential curve are presented in Table 2. The RMS for number of mature yellow follicles and weight of ventral ligament showed a significant change (P 0.001) when the linear parameters of the model were fitted separately, but no significant change was found when the nonlinear parameters were fitted separately for each strain. The results demonstrated that although there were significantly more mature yellow follicles and a heavier ventral ligament in the sire line, there were no differences in the rate of maturation of these structures between the two strains. The total ovary weights in the two strains for 7 wk following photostimulation are presented in Figure 1g. Regression analysis did not result in a biologically sensible model for the changes in total ovary weight from

4 1494 BUCHANAN ET AL. FIGURE 1. Mean (± SEM) change in oviduct weight (a), uterus weight (b), vagina weight (c), residual ovary weight (d), number of follicles (e), ventral ligament weight (f), total ovary weight (g), and sphincter ani weight (h) from 0 to 7 wk of photostimulation in traditional and sire line turkeys. Key: = sire line; = traditional line. Weeks 0 to 7. However, an exponential model was fitted to the data for Weeks 1 to 7, as shown on Figure 1g. The estimated parameters from this model are shown in Table 2. There was a significant change (P 0.001) in the RMS when the linear parameters (upper asymptote and scaling factor) were fitted separately for each strain, but no significant change was found when different non-linear parameters were evaluated. The mean weights of the sphincter ani muscle in the two strains from 0 to 7 wk of photostimulation are shown on Figure 1h. There was little change in the weight of this muscle during the course of the experiment, and we were unable to obtain a satisfactory fit to any nonlinear regression equation. The data were therefore analyzed by ANOVA. There was no interaction between strain and week. The sphincter ani muscle was significantly heavier

5 REPRODUCTIVE DEVELOPMENT AND PROLAPSE IN TURKEYS 1495 TABLE 1. Estimates of fitted parameters (± SE) of logistic curves for oviduct length, weight, residual ovary weight, and weights of the uterine and vaginal portions of the oviduct in traditional and sire line turkeys Rate Point of Upper Lower Item parameter inflection 1 asymptote asymptote R 2 Oviduct length, mm Traditional 1.1 ± ± ± ± Sire line 1.1 ± ± ± ± 142 Oviduct weight, g Traditional 1.8 ± ± ± ± Sire line 1.3 ± ± ± ± 14.0 Residual ovary weight, g Traditional 0.9 ± ± ± ± Sire line 1.4 ± ± ± ± 2.67 Uterus weight, g Traditional 2.6 ± ± ± ± Sire line 1.6 ± ± ± ± 3.17 Vagina weight, g Traditional 1.8 ± ± ± ± Sire line 2.0 ± ± ± ± Weeks postphotostimulation. (P 0.001) in the sire line compared with the traditional turkey line. There was also a significant effect of week (P 0.001). The sphincter ani weight increased at the start of photostimulation, and there was a slight decrease during the final weeks of the experiment. Analysis of Data from Melnychuk et al. (1997) The data of Melnychuk et al. (1997) was reconstructed, and the oviduct and ovary weights were plotted separately, with each curve showing data for both the sire and dam lines as shown in Figure 2. A logistic curve was the best-fitting model for both oviduct weight and ovary weight. There was no significant effect on the fitted model for either oviduct weight or ovary weight when linear or nonlinear parameters were fitted separately for the sire line and dam line. The parameters of the curves are given in Table 3. Plasma Estradiol Concentration Mean plasma estradiol concentrations from 0 to 7 wk postphotostimulation are shown in Figure 3. There was no statistical evidence to support the fitting of a curvilin- ear model over a linear response to the mean plasma estradiol concentration for either the male line or traditional turkey line. The ANOVA showed a significant effect of week on plasma estradiol (P 0.001). This effect was due entirely to the low levels of plasma estradiol at Week 0. There was no significant change in plasma estradiol from Week 1 to 7 following photostimulation. The sire line turkeys had a lower overall mean plasma estradiol concentration (189 ± 21 pg/ml) than the traditional turkeys (202 ± 13 pg/ml), although this difference was not significant. Histology Histological investigation of the uterus, vagina, sphincter ani muscle, and muscular cord of the ventral ligament revealed no differences between the traditional, sire line, and prolapsed sire line turkeys. The vagina of both lines was shown to consist of an inner epithelial layer surrounded by two muscle layers, an inner circular layer, and an outer longitudinal layer. The surrounding circular and longitudinal muscle layers were also found in the uterine sections, although they were much thinner than in the vagina. The sections from the sphincter ani muscle showed a tight band of circular muscle fibers surrounded TABLE 2. Estimates of fitted parameters of exponential models for weight of the muscular cord of ventral ligament and the number of maturing follicles in traditional and sire line turkeys Rate Scaling factor Asymptote, g R 2 Number of follicles Traditional 0.63 ± ± ± Sire line 0.69 ± ± ± 3.72 Ventral ligament weight, g Traditional 0.57 ± ± ± Sire line 0.53 ± ± ± Total ovary weight, g Traditional 0.52 ± ± ± Sire line 0.56 ± ± ± 32.8

6 1496 BUCHANAN ET AL. FIGURE 2. Ovary (a) and oviduct (b) weights as percentages of mature organ weights from 206 to 245 d of age in dam and sire line turkeys (data derived from Melnychuk et al., 1997). Key: = sire line, = dam line. by connective tissue. Sections of the muscular cord of the ventral ligament from traditional, sire line, and prolapsed sire line turkeys demonstrated a distinct muscular cord of the ventral ligament in all three groups that consisted mainly of muscle fibers and connective tissue (color sections are reproduced in Buchanan, 1999). The pattern of collagen distribution was similar in all three groups of turkeys. There was no evidence of inflammation, infection, edema, hematoma, or apparently damaged muscle fibers in the prolapsed samples. DISCUSSION The development of the oviduct followed the same pattern of growth in both lines of turkeys. Because there were no significant differences in the fitted models when the nonlinear parameters were fitted for each strain, it is clear that growth occurred over the same time course in each strain. The significant effect of strain on the upper asymptotes for oviduct weights, total and residual ovary weights, and weights of the uterus and vagina reflects the greater size of these structures in the sire line compared with the traditional line. These results are not surprising, because the sire line has a much greater body weight than the traditional turkey line. Lilburn and Nester (1993) suggested that growth of the oviduct continued for 7 wk after the onset of lay in their sire line, but not in their dam line. The results of our experiment showed that the growth of the oviduct reached a plateau within 7 wk of photostimulation. In the sire line, the first egg is normally laid 3 wk after the onset of photostimulation, and the results are, therefore, not consistent with those of Lilburn and Nester (1993). The proposal that oviduct growth in turkeys selected for meat increases through the first 49 d of egg production was based on the significantly greater weight of the oviduct of the sire line at 49 d than at the onset of lay, and does not necessarily reflect growth throughout that period. The time course of growth of the residual ovary was not significantly different for the traditional and sire line turkeys, as shown by the lack of significant difference in the fitted model when the non-linear parameters were fitted separately for the two strains. The total ovary consists of two components, the residual ovary and the number of mature yellow follicles. The results of this experiment showed that residual ovary growth followed a logistic model, whereas the number of mature yellow follicles increased exponentially. The combination of these two components with the total ovary may explain why neither model gave a satisfactory fit to the data for total ovary weight. However, there was no significant difference in the maturation of either the residual ovary or the number of mature yellow follicles between the two strains. It was also shown that there was no difference in the maturity of the total ovary between the two strains from 1 to 7 wk of photostimulation. There was, therefore, no evidence to suggest a difference in the developmental rate of the ovary between the traditional and sire lines. Statistical analysis of the reconstructed data from Melnychuk et al. (1997) did not support the conclusion that there was a difference in the relative growth rate of the TABLE 3. Estimates of fitted parameters (± SE) of logistic curve models for oviduct and ovary as percentages of mature organ weights in sire line and dam line turkeys (data extracted from Melnychuk et al., 1997) Point of Upper Lower Rate 1 inflection asymptote asymptote R 2 Oviduct % mature organ weight Sire line 0.65 ± ± ± ± Dam line 0.51 ± ± ± ± 3.1 Ovary % mature organ weight Sire line 0.28 ± ± ± ± Dam line 0.29 ± ± ± ± 6.7

7 REPRODUCTIVE DEVELOPMENT AND PROLAPSE IN TURKEYS 1497 FIGURE 3. Mean (± SEM) plasma estradiol concentration from 0 to 7 wk postphotostimulation in traditional and sire line turkeys. Key: = traditional, shaded box = sire line. ovary and oviduct in their sire and dam lines. In fact, the fitted curve for the dam line lies to the left of that of the sire line, with its inflection point numerically 1 d earlier than that of the sire line. Clearly, the suggestion that the ovary of the dam line reached mature weight 3 d after that of the sire line is not substantiated. The oviduct in both lines matured 4 to 5 d after the ovary, as measured by the point of inflection. The same group investigated oviduct and ovarian development in bronze turkeys, and again suggested that the ovary reached maturity 3 d after the oviduct (Renema et al., 1998). However, their analysis used linear regression to fit straight lines to the growth curves, and the authors did not provide statistical evidence to support their statement. We prefer to conclude that there is no evidence from the results of the experiment described here or those of Melnychuk et al. (1997) and Renema et al. (1998) to support the suggestion that the ovary in turkeys reaches maturity 3 d after the oviduct. Furthermore, the results of this experiment and the reanalysis of published data failed to provide evidence for differences in the relative growth patterns or rates of maturity that could influence the stability of the reproductive system and predispose females of the sire line to prolapse of the oviduct. The structures of the vagina were consistent with the observations of Verma and Cherms (1964) in bronze broad-breasted turkeys. Detailed histological investigation of the uterus, vagina, muscular cord of the ventral ligament, and sphincter ani muscle did not show any differences that could be involved in prolapse of the oviduct. These results disagree with the finding in prolapsed and non-prolapsed hens (Rao et al., 1985) that the presence of a muscular ventral ligament was associated only with prolapse. There was a distinct muscular cord of the ventral ligament in the traditional, sire line, and prolapsed sire line turkeys that consisted of longitudinal muscle fibres and connective tissue. The hypothesis suggested by Rao et al. (1985) that over-development of the ventral ligament caused increased torsion in the oviduct and predisposed the birds to prolapse is not supported by the present data. The frequency of prolapse is variable and unpredictable, and in order to obtain prolapsed sire line turkeys that were contemporary to unaffected birds, we had to rely on a commercial flock. Therefore, we hatched traditional birds at the same time and reared them in a manner similar to the commercial birds. Selection of the birds was conducted at random, and the data on the relative growth of the oviducts were assessed by regression analysis. It is unlikely that obtaining the samples from different locations would affect the parameters of the growth curves of the two lines significantly, although absolute weights of individual tissues might vary. Furthermore, the results are consistent with comparable data from traditional and sire line turkeys reared together and examined at the onset of lay (Buchanan et al., 1999). The results of our experiment do not provide any evidence for a peak in plasma estradiol concentration, either just prior to the onset of lay as observed in domestic hens (Senior, 1974), or around the onset of lay, when prolapse is most common. This result is consistent with the previous findings of Bacon et al. (1980). The lower plasma estradiol concentration of the sire line turkeys is surprising considering the larger residual ovary weight and the greater number of yellow follicles in this strain. These results provide no evidence of any underdevelopment in the growth of the oviduct or its supporting structures in the sire line compared with the traditional strain, and this was confirmed by the absence of differences in the histological examination of the tissues. We have shown elsewhere that there is a lower content of collagen in the vagina, but not the uterus, in the sire line compared with the traditional strain, and that the vaginal collagen content of prolapsed turkeys was lower still, compared with nonprolapsed sire line turkeys (Buchanan et al., 1999). The physiological explanation for these differences is presently unknown, but it is clear that relatively high plasma estradiol concentrations are unlikely to be the cause of the higher predisposition to prolapse of the oviduct in sire line turkeys. It is possible that genetic differences in receptor populations or efficiencies of steroid hormones associated with selection for high growth rate and breast muscle yield are associated with prolapse of the oviduct. Alternatively, selection might have changed the status of growth factors that affect collagen metabolism and predispose the sire line to prolapse of the oviduct. ACKNOWLEDGMENTS This work was funded by a CASE studentship with the Ministry of Agriculture, Fisheries and Food (London SWIP 4PQ, UK), and British United Turkeys Ltd. (Warren Hall, Chester, CH40EW UK). Sire line turkeys were kindly donated by British United Turkeys Ltd. Thanks are due to Gerry Baxter for technical advice on the estradiol- 17β assay. REFERENCES Bacon, W. L., K. I. Brown, and M. A. Musser, Changes in plasma calcium, phosphorus, lipids, and estrogens in turkey hens with reproductive state. Poultry Sci. 59:

8 1498 BUCHANAN ET AL. Bienkiewicz, A., J. Welfel, and E. Kus, The influence of estrogens, progesterone and their antagonists on the collagen content in the pregnant rat uterus. Horm. Metab. Res. 28: Buchanan, S., Physiological basis of prolapse of the oviduct in turkeys. Ph.D. Thesis, University of Edinburgh, Edinburgh, Scotland. Buchanan, S., G. W. Robertson, and P. M. Hocking, The relationships between vaginal collagen, plasma oestradiol and uterine prolapse in turkeys. Res. Vet. Sci. 67: Etches, R. J., The ovary. Pages in: Reproduction in Poultry. CABI, Oxon, UK. Harvey, E. B., H. E. Kaiser, and L. E. Rosenberg, Trunk and torso. Pages in: An Atlas of the Domestic Turkey (Meleagris gallopavo) Myology and Osteology. Bull no. 56. U. S. Government Printing Office, Washington, DC. Hocking, P. M., Relationships between egg size, weight and pelvic dimensions in turkeys. Anim. Prod. 56: Lilburn, M. S., and K. E. Nestor, The relationship between various indices of carcass growth and development and reproduction in turkey hens. Poultry Sci. 72: Melnychuk, V. L., F. E. Robinson, R. A. Renema, R. T. Hardin, D. A. Emmerson, and L. G. Bagley, Carcass traits and reproductive development at the onset of lay in two lines of female turkeys. Poultry Sci. 76: Pastore, G. N., L. P. DiCola, N. R. Dollahon, and R. M. Gardiner, The effect of estradiol on collagen structure and organization in the immature rat uterus. Proc. Soc. Exp. Biol. Med. 191: Rajabi, M., S. E. Solomon, and A. R. Poole, Hormonal regulation of interstitial collagenase in the uterine cervix of the pregnant guinea pig. Endocrinology 128: Rao, A. G., L. N. Das, S. K. Chanda, V.S.C. Bose, and S. C. Mishra, Prolapse of the oviduct in a New Hampshire strain. Indian J. Poult. Sci. 20: Renema, R. A., V. L. Melnychuk, F. E. Robinson, H. L. Classen, and R. D. Crawford, Reproductive organ morphology and carcass traits in unselected naturally mating female bronze turkeys at onset of lay. Can. J. Anim. Sci. 78: Senior, B. E., Oestradiol concentration in the peripheral plasma of the domestic hen from 7 weeks of age until the time of sexual maturity. J. Reprod. Fertil. 41: Shemesh, M., L. Shore, S. Lavi, M. Ailenberg, U. Bendheim, A. Totach, and Y. Weisman, The role of 17β-estradiol in the recovery from oviductal prolapse in layers. Poultry Sci. 63: Shemesh, M., L. Shore, U. Bendheim, S. Lavi, and Y. Weisman, Correlation between serum estrogen concentrations, light and prolapsed uterus in laying hens. Adv. Pathol. 1: Shore, L. S., M. Shemesh, U. Bendheim, and A. Totach, The effect of pharmacological doses of oestrogen on oviductal prolapse in the hen. Refu. Vet. 41: Verma, O. P., and F. L. Cherms, Observations on the oviduct of turkeys. Avian Dis. 8: Webb, R., G. Baxter, D. McBride, G. D. Nordblom, and M.P.K. Shaw, The measurement of testosterone and oestradiol- 17β using iodinated tracers and incorporating an affinity chromatography extraction procedure. J. Steroid Biochem. 23: Wheeler, R. S., and E. Hoffman, Prolapse and pickout in diethylstilbestrol-treated cockerels. Cornell Vet. 38:89 92.