USVI Public Records

Document text

60 Sheep & Goat Research Journal, Volume 25, 2010 ©2010, Sheep & Goat Research Journal Summary There is little information available on methods to estimate fetal age of hair sheep. The objective of this study was to determine whether crown-rump length (CRL) and/or fetal heart rate (FHR) could be used to predict fetal age in hair sheep using transrectal B-mode and Doppler ultrasonography. St Croix White, Barbados Blackbelly and Dorper X St. Croix White ewes (n = 54) were scanned weekly beginning 28 d after a successful mating. Linear array B-mode ultrasonography (5 MHz) was used to measure CRL and visual FHR, and Doppler ultrasonography was used to measure audible FHR. Due to the size of the fetus CRL was not measurable after 42 d of gestation, and visual FHR was Evaluation of Ultrasonography to Measure Fetal Size and Heart Rate as Predictors of Fetal Age in Hair Sheep R.W. Godfrey1,2, L. Larson2, A.J. Weis2 and S.T. Willard3 1 Corresponding author: Email: rgodfre@uvi.edu 2 Agricultural Experiment Station, University of the Virgin Islands, St Croix, VI; 3 Department of Biochemistry and Molecular Biology, Mississippi State University, Mississippi State, MS; Volume 25, 2010 not measurable after 70 d. Audible FHR was not consistently measurable before 35 d but could be measured through 140 d of gestation. Single fetuses had greater CRL (P < 0.01) than multiple fetuses at d 35 and d 42 of gestation. There was no difference (P > 0.10) in visual or audible FHR between single and multiple fetuses. Visual FHR was higher (P < 0.0001) than audible FHR at 49 d of ges- tation. Both CRL and audible FHR had a linear relationship with days of gesta- tion for single and multiple fetuses. The relationship between days of gestation and visual FHR was best described by a cubic equation for single fetuses and a quadratic equation for multiple fetuses. Accuracy of the regression equations and the software in the ultrasound machine was evaluated by scanning a set of ewes (n = 51) also with known breeding dates. Both the equations and the software underestimated actual age of single fetuses by 2.5 d (P < 0.01). For multiple fetuses the equation overestimated the age by 1 day, and the software underesti- mated age by over 2 d (P < 0.01). Over- all, the regression equations underesti- mated fetal age by 1 d and the software underestimated fetal age by more than 2 d (P < 0.004). Fetal age can be estimated with acceptable accuracy in hair sheep breeds, regardless of fetal number, using existing methods that were developed using other breeds of sheep. Key Words: Sheep; Fetus; Ultra- sonography; Fetal Heart Rate Acknowledgment The authors thank William Gonzales and Melinda Loewer for assistance with animal handling and data collection. This project was a contribution to the North Central Extension and Research Activity NCERA-190 Increased Efficiency of Sheep Production. Introduction Pregnancy diagnosis in sheep usu- ally is done using some form of ultra- sonography, since rectal palpation is impractical due to the size of the ani- mals. Linear-array ultrasonography is a common method because it allows the operator to view the fetus, determine fetal number and collect measurements of the fetus to determine fetal age (Grif- fin and Ginther, 1992; Karen et al. 2004; Romano and Christians, 2008). One measurement that is commonly used to estimate fetal age is crown-rump length (CRL). In addition to CRL, other meas- urements used to estimate fetal age in sheep include head width and thoracic depth (Sergeev, et al., 1990) and bipari- etal skull and body trunk diameter (Aiumlamai, et al., 1992). Neither of these studies indicated that there was an effect of fetal number on the estimated age of the fetus based on the measure- ments collected. Audible Doppler ultra- sonography also has been used for preg- nancy detection in the ewe (Lindahl, 1971), sow (Too et al., 1974; Pierce et al., 1976), mare (Mitchell, 1973) and cow (Mitchell, 1973). Because hair sheep breeds in the tropics are smaller than most wool breeds in temperate areas and lamb birth weights are smaller (Godfrey et al. 2004; Godfrey, 2005), it is unknown if CRL measurements can accurately pre- dict fetal age in these breeds using exist- ing criteria. In addition, there is little information available on the effect of multiple fetuses on the accuracy of fetal age predictions. The objectives of this study were to compare measurements of CRL and fetal heart rate (FHR) throughout gestation using transrectal B-mode and Doppler ultrasonography in hair sheep. The effect of multiple fetuses on the accuracy of the measurements was evaluated. Materials and Methods The Institutional Animal Care and Use Committee of the Agricultural Experiment Station of the University of the Virgin Islands approved all proce- dures prior to commencement of this study (FASS, 1999). Hair sheep ewes (n = 54) were bred during a 35-d period in October 2003, and the number of ewes of each breed type that was carrying sin- 61 Sheep & Goat Research Journal, Volume 25, 2010 ©2010, Sheep & Goat Research Journal Table 1. Number of ewes of each breed type with either single, twin or triplet fetuses. Breed type Sire breed Barbados Dorper x St. Croix Blackbelly Dorper St. Croix White White Dam breed Barbados St. Croix Dorper x St. Croix Blackbelly White St. Croix White White Fetal number Number of ewes 1 12 2 5 1 2 10 4 4 10 3 0 6 0 0 Number of fetuses 32 28 13 21 Figure 1. Crown-rump length (CRL) of single and twin hair sheep fetuses measured between d 28 and 42 of gestation. Figure 2. Audible and visual fetal heart rate (FHR) of hair sheep fetuses measured using Doppler or B-mode ultrasonography, respectively. ©2010, Sheep & Goat Research Journal Sheep & Goat Research Journal, Volume 25, 2010 62 gle, twin or triplet fetuses is shown in Table 1. Rams were equipped with mark- ing harnesses during the breeding period, and ewes were checked twice daily for crayon marks to determine actual breed- ing dates of the ewes. Fetal measurements were collected weekly beginning on d 28 after a suc- cessful mating (d 0), based on crayon marks and non-return to estrus. Ewes were placed in dorsal recumbency in a flip cradle for the transrectal ultrasound examinations. Linear-array, B-mode ultrasonography was used to measure CRL and visually measure FHR using a Pie Medical Scanner 480 with a 5/7.5 MHz linear-array, switchable transducer set to 5 MHz for all examinations or a 5 MHz linear-array intracavity probe (Classic Medical, Tequesta, Fla., USA). Doppler ultrasound coupled to a battery- operated headset was used to measure audible FHR using a rectal probe (Medata Systems Ltd., West Sussex, UK). Two trained technicians con- ducted all ultrasound exams distributed uniformly across days and breeds. Visual FHR was measured on the image of the fetus using the linear-array ultrasound by counting the heartbeats on the screen for 10 s. The CRL was measured on a still image of the fetus on the screen using electronic calipers pro- vided through the software on the machine. The cursor was placed on the top of the head and the second point was placed at the most distant point over the rump of the fetus. Audible FHR was measured using the Doppler ultrasound probe by count- ing the heartbeats through the ear- phones for 10 s. All FHR were adjusted to beats/min (BPM). In ewes with multi- ple fetuses, measurements were collected on each fetus at each time point. This was accomplished by rotating the head of the transducer and monitoring the sound as one fetal heartbeat faded out and the transducer picked up the next fetal heartbeat. Because there was no way to determine which fetal heartbeat was being measured using the Doppler ultrasound, the results were coded by fetus number (1, 2 or 3). Repeatability of fetal-heart-rate measurements was not determined in this study, so the error associated with making the FHR meas- urements cannot be partitioned from the variance associated with the trait. Data were analyzed using GLM procedures for repeated measures (SAS, 1999). Fetal breed type, number of fetuses and day of gestation were the main effects used in the model for CRL and audible and visual FHR. Because there were only two sets of triplets, fetuses were classified as either singles or multiples for the final analysis. There was no effect (P > 0.10) of fetal breed type on CRL or FHR (audible and visual), so all data were pooled across breed types for subsequent analysis. Relationships between fetal age and CRL and FHR measurements were determined using stepwise regression analysis with linear, quadratic and cubic terms included in the models within lit- ter size. To compare fetal ages produced by the machine software and the generated regression equations for CRL and fetal age, a second set of St. Croix White, Barbados Blackbelly and Dorper x St. Croix White ewes (n = 51) with known breeding dates, carrying single (n = 21) or multiple (n=30) fetuses were scanned for CRL between 26 d and 45 d of gesta- tion (validation group). The age of the fetus determined by the software in the ultrasound machine and the regression equations developed for single or multi- ple fetuses were compared to the actual fetal age based on breeding dates. The Figure 3. Regression of crown-rump length (CRL) on days of gestation in single or multiple hair sheep fetuses. The dots represent a measurement of CRL on day 28, 35 or 42 of gestation. specific algorithm used in the ultrasound machine to estimate fetal age was not specified in the operators’ manual. The differences were calculated as (actual age – estimated age), where estimated age was produced either by the software in the ultrasound machine or by the regres- sion equations generated in this study. General Linear Models procedures of SAS were used to compare the differ- ences using number of fetuses (single or multiple) in the model as the main effect. The Student’s t-test was used to determine if the differences were equal to zero within fetal number groups. Results and Discussion Due to the size of the fetus and 5 MHz transducer depth of penetration, CRL was not consistently measurable before 28 d and after 42 d of age, and visual FHR was not consistently detectable before 28 d and after 70 d of age. Audible FHR was not consistently detectable before 35 d of age but could be measured through 140 d of age using the Doppler ultrasound. Romano and Christians (2008) reported that preg- nancy could be detected in ewes as early as d 16 using a 7.5 MHz transducer, but the authors did not conduct any meas- urements of the fetus in the study. Other studies have shown that fetal measure- ments, such as biparietal-skull diameter and thoracic diameter, can be measured between 49 d and 140 d of gestation (Sergeev et al., 1990, Aiumlamai et al., 1992). In both of these studies the ultra- sound was conducted transabdominally using either a 7.5, 5 or 3.5 MHz trans- ducer, as opposed to transrectally using a 5 MHz transducer as in the present study. There was no difference (P > 0.10) in CRL between single and multiple fetuses at 28 d of age, but single fetuses had a greater CRL than multiple fetuses at 35 d (P < 0.0008) and 42 d (P < 0.02) (Figure 1). In contrast, Sergeev et al. (1990) reported that there were no dif- ferences in head length among litter sizes of 1, 2 or 3 lambs. The difference in results between the two studies may be due to the type of fetal measurements being collected. This highlights the need to validate whatever method is chosen to estimate fetal age in sheep, whether it is CRL or head length or diameter. Visual FHR was similar to audible FHR at all times except for 49 d of age (P < 0.0001, Figure 2). There was an increase in FHR using both methods on d 49 but it was unclear what might have caused this. The values reported for FHR in the current study were higher than those reported by Aiumla- mai et al. (1992) during the entire ges- tation. The decrease in FHR as fetal age increases was consistent with reports in sheep (Aiumlamai et al., 1992) and cat- tle (Breukelman et al., 2004). Crown-rump length had a linear relationship (P < 0.0001) with age (Fig- ure 3) in single and multiple fetuses. The brief time period (14 d) when CRL could be measured in the current study limits the amount of information that can be obtained about fetal age using the meth- ods described. The use of other fetal measurements, such as skull or thoracic diameter, may expand the portion of ges- tation during which fetal age can be esti- mated in hair sheep (Sergeev et al., 1990; Aiumlamai et al., 1992). In cattle biparietal diameter of the cranium was measureable between d 35 and d 100 of gestation, and the relationship was lin- ear (Breukelman et al., 2004). In rein- 63 Sheep & Goat Research Journal, Volume 25, 2010 ©2010, Sheep & Goat Research Journal Figure 4. Regression of audible fetal heart rate (FHR) on days of gestation in single or multiple hair sheep fetuses. The dots represent a measurement of audible FHR between days 35 and 140 of gestation. deer CRL was measurable in week 3 through week 17 of gestation, and chest width and depth was measurable through week 19 of gestation, after which the fetus became too large to measure (Vah- tiala et al., 2004). The authors also reported that CRL and chest width and depth had quadratic relationships with gestational age in reindeer. Audible FHR had a linear relation- ship (P < 0.0001) with day of age (Figure 4) in single and multiple fetuses. Visual FHR had a curvilinear relationship (P < 0.0001) with day of age, that was best described by a cubic equation in single fetuses and a quadratic equation in mul- tiple fetuses (Figure 5). In contrast, Aiumlamai et al. (1992) reported that visual FHR had a linear relationship with fetal age during the second half of gestation, but they did not distinguish between single or twin fetuses. In the current study the visual FHR is lowest at d 28, before audible FHR was consis- tently measureable, and this may con- tribute to the curvilinear relationship being reported. The fetal age of the validation group that was scanned to evaluate the accuracy of the machine software and the generated regression equations for CRL ranged from 26 d to 45 d of age, based on breeding dates. On d 26, d 29 and d 38 fetal ages estimated by measur- ing CRL and using the regression equa- tions developed in the current study (Figure 3) were greater (P < 0.05) than those estimated using CRL and the soft- ware in the ultrasound machine (Figure 6). For single fetuses, both the equation and the software underestimated actual age by 2 d to 3 d based on CRL meas- urements (Table 2). For multiple fetuses, the equation overestimated the age by 1 d and the software underesti- mated age by more than 2 d (P < 0.0001). Overall, the equations gener- ated in the current study underesti- mated fetal age by 1 d, and the software underestimated fetal age by more than 2 d (P < 0.04). This small discrepancy between actual age and estimated age may not be important enough to impact decisions regarding management or pre- dicting lambing time of ewes carrying single or multiple fetuses. Conclusions Fetal age can be estimated with acceptable accuracy in hair sheep breeds, regardless of fetal number, using existing methods that were developed using other breeds of sheep. It may be benefi- cial to develop systems that utilize meas- urements besides crown-rump length, such as cranial or thoracic diameter, to expand the number of days during gesta- tion during which fetal age can be esti- mated. Because audible and visible fetal heart rate can be monitored for more days during gestation, it has the poten- tial to be developed more fully as a tool for estimating fetal age. Literature Cited Aiumlamai, S, G. Fredriksson, and L. Nilsfors. 1992. Real-time ultra- sonography for determining the ges- tational age of ewes. Vet. Rec. 131:560-562. Breukelman, S.P., J.M.C. Reinders, F.H. Jonker, L. de Ruigh, L.M.T.E. Kaal, A.M. van Wagtendonk-de Leeuw, P.L.A.M. Vos, S.J. Dieleman, J.F. ©2010, Sheep & Goat Research Journal Sheep & Goat Research Journal, Volume 25, 2010 64 Figure 5. Regression of visual fetal heart rate (FHR) on days of gestation in single or multiple hair sheep fetuses. The dots represent a measurement of visual FHR between days 28 and 70 of gestation. Beckers, Zs. Perenyi, and M.A.M. Taverne. 2004. Fetomoetry and fetal heart rates between day 35 and 108 in bovine pregnancies resulting from transfer of either MOET, IVP- co-culture or IVP-SOF embryos. Theriogenology 61:867-882. Godfrey, R.W., J.R Collins, E.L. Hensley, H.A. Buroker, J.K. Bultman, and A.J Weis. 2004. Production of hair sheep using accelerated lambing and an extensive management system in the tropics. Proc. Caribb. Food Crop Soc., 40:129-136. Godfrey, R.W. Hair Sheep Production in the Tropics: A Caribbean Per- spective. 2005. Proceedings of the Hair Sheep Workshop, Virginia State University, Petersburg. http://www.sheepandgoat.com/Hair S h e e p Wo r k s h o p / t r o p i c s . h t m l Accessed August 11, 2010. Griffin, P.G. and O.J. Ginther. 1992. Research applications of ultrasonic imaging in reproductive biology. J. Anim. Sci. 70:953-972. Guide for the Care and Use of Agricul- tural Animals in Agricultural Research and Teaching. 1999. FASS, Savoy, IL. Karen, A., K. Szabados, J. Reiczigel, J.F. Becker, and O. Szenci. 2004. Accu- racy of transrectal ultrasonography for determination of pregnancy in sheep: effect of fasting and handling of the animals. Theriogenology 61:1291-1298. Lindahl, I.L. 1971. Pregnancy diagnosis in the ewe by intrarectal doppler. J. Anim. Sci. 32:922-925. Mitchell, D. 1973. Detection of foetal circulation in the mare and cow by Doppler ultra-sound. Vet. Rec. 93:365-368. Pierce, J.E., C.C. Middleton, and J.M. Phillips. 1976. Early pregnancy diagnosis in swine using Doppler ultrasound. International Pig Vet- erinary Society – Proceedings of the 4th International Congress: D3. Romano, J.E., and C.J. Christians. 2008. Early pregnancy diagnosis by tran- srectal ultrasonography in ewes. Small Ruminant Res. 77:51-57. SAS Institute Inc., The SAS System for Windows (v 8.0) Cary, NC; 1999. Sergeev, L., D.D. Kleeman, S.K. Walker, D.H. Smith, T.I. Grosser, T. Mann, and R.F. Seamark. 1990. Real-time ultrasound imaging for predicting ovine fetal age. Theriogenology 34:593-601. Too, K., K. Kawata, Y. Fukui, K. Sato, K. Kagota, and K. Kawabe. 1974. Stud- ies on pregnancy diagnosis in domestic animals by ultrasonic doppler method. I. Pregnancy diag- nosis in the pig and fetal heart rate changes during pregnancy. Jpn. J. Vet. Res. 22(3): 61-71. Vahtiala, S., H. Sakkinen, E. Dahl, E. Eloranta, J.F. Beckers, and E. Rop- stad. 2004. Ultrasonography in early pregnancy diagnosis and measure- ments of fetal size in reindeer (Rangifer tarandus tarandus). Theri- ogenology 61:785-795. 65 Sheep & Goat Research Journal, Volume 25, 2010 ©2010, Sheep & Goat Research Journal Figure 6. Relationship between actual fetal age and fetal age estimated by software included with the ultrasound equipment and the equations generated in this study (Figure 3). Table 2. Differences between actual and estimated fetal ages using CRL measurements based on fetal number in the validation group of ewes. Number of fetuses Comparison Differencea Single Actual - Equation 2.6 ± 0.5b Actual - Software 2.4 ± 0.5b Multiple Actual - Equation -0.4 ± 0.5c Actual - Software 2.4 ± 0.5b Pooled Actual - Equation 1.1 ± 0.4d Actual - Software 2.4 ± 0.4e a Difference was calculated by subtracting fetal age estimated by equations generated in this study or fetal age estimated by ultrasound machine software from the actual fetal age, based on day of breeding. The equation used for a single fetus was y = 2.7366(x) + 26.0229 and for multiple fetuses was y = 4.3519(x) + 24.7716 (Fig. 3). b,c Values within fetal number group or within comparison are different (P < 0.0001). d,e Pooled values are different (P < 0.02).