24th Int. Symp. “Animal Science Days”, Ptuj, Slovenia, Sept. 21st−23rd, 2016. Acta argiculturae Slovenica, Supplement 5, 113–117, Ljubljana 2016 COBISS: 1.08 Agris category code: L01 CORRELATIONS BETWEEN UDDER TRAITS AND THEIR RELATIONSHIP WITH MILK YIELD DURING FIRST LACTATION IN SLOVENIAN ALPINE GOATS Angela CIVIDINI 1, Tina FLISAR 2, Milena KOVAČ 3, Drago KOMPAN 4 Correlations between udder traits and their relationship with milk yield during first lactation in Slovenian Alpine goats 1 University of Ljubljana, Biotechnical Faculty, Department of Animal Science, Groblje 3, SI-1230 Domžale, Slovenia, e-mail: angela.cividini@bf.uni-lj.si 2 Same address as 1, e-mail: tina.flisar@bf.uni-lj.si 3 Same address as 1, e-mail: milena.kovac@bf.uni-lj.si 4 Same address as 1, e-mail: drago.kompan@bf.uni-lj.si ABSTRACT Udder traits were subjectively scored in 36 Slovenian Alpine goats as udder depth (UD), fore (FUA) and rear udder attachment (RUA), medial suspensory ligament (MSL), teat diameter (TD), length (TL) and position (TP), and measured as UDcm, RUAcm, TDcm, TLcm and TPcm. Phenotypic correlations among udder traits were estimated. Standardised regression coefficients were estimated for the effect of days in milk (DIM) on udder traits and daily milk yield (MY) and were tested for slope heterogeneity. Phenotypic correlations between UD and FUA, and UD and RUA were 0.25 and 0.11, respectively. Highly positive correlation was estimated between RUA and RUAcm (0.70), TD and TDcm (0.78), and TL and TLcm (0.85). Moderate correlations were estimated between days in milk and UDcm (−0.39), RUAcm (−0.34), and TDcm (−0.26), suggesting UD, RUA and TD become smaller as lactation progresses. Changes of udder traits over lactation were smaller than changes for daily milk yield. Decrease of TP, RUAcm, UDcm, TDcm, and milk yield throughout lactation was obtained. We found similar trend in changing UDcm and RUAcm with the trend of changing daily milk yield. Key words: goats, udder traits, phenotypic correlation, milk yield, lactation period 1 INTRODUCTION Slovenian Alpine (SA) goat is the most important goat dairy breed. There are also some other breeds used in dairy production, like Slovenian Sannen and autochtho- nous Dreznica goat. Although the SA counts only about 800 purebred animals, the breeding association is running a breeding programme (Kastelic et al., 2010). The selection of does is based on genetic evaluation for milk traits and exterior characteristics. The present selection on dairy pro- duction traits in Slovenia does not include all udder traits, therefore, a linear scoring system (Linear Appraisal System, 2014) has been recently proposed for seven udder traits on a nine-point linear scale. The improvement of udder traits is beneficial to milking ability and animal health. Moreover, deep and well-attached udders are highly correlated with milk production in sheep (Casu et al., 2000 cited in Altin- çekiç and Koyuncu, 2011; Legarra and Ugarte, 2005) and goats (McLaren et al., 2016). Higher and more tightly at- tached udders are associated with lower somatic cell scores, thus, udder depth is considered as an easy-to-collect pre- dictor of udder health in goats (Rupp et al., 2011). Vertical placed teats on the lowest part of the cistern are less subject to teat-cup falls during milking (Casu et al., 2000 cited in Altinçekiç and Koyuncu, 2011). Linear scoring system has to include udder traits which have economic impact and are at least moderately heritable. Heritabilities estimated by Manfredi et al., (2001) were moderate (about 0.30) for udder traits and teat loca- tion traits and exceeds 0.40 for teat length, width, and form. Similar results were obtained by Rupp et al. (2011) where heritabilities ranged from 0.2 for teat angle to 0.5 for teat form and length. Estimates of heritability for rear udder at- tachment were from 0.23 to 0.27. Several studies reported Acta agriculturae Slovenica, Supplement 5 – 2016114 A. CIVIDINI et al. negative genetic correlations between milk yield and ud- der depth in goats (Manfredi et. al., 2001; Rupp et al., 2011; McLaren et al., 2016). McLaren et al. (2016) obtained weaker but still negative correlations between milk yield and udder attachment (−0.28). In addition, the correlations estimated between milk yield and teat angle (−0.40) and milk yield and teat placement (−0.35) remained negative throughout lactation in goats. McLaren et al. (2016) sug- gested that breeding programs for dairy goats would ben- efit by taking udder traits into account in order to prevent deterioration in udder conformation. Rupp et al. (2011) stated that improvement of udder conformation could also reduce lactation somatic cell count in goats. The main goals of our study were to (1) estimate phe- notypic correlations among udder traits, (2) to estimate phenotypic correlations between measured and appraisal udder traits, and (3) to assess phenotypic relationship be- tween udder traits and milk yield during the first lactation. 2 MATERIAL AND METHODS 2.1 DATA The study was conducted at the Educational and Re- search Animal Husbandry Centre Logatec at Biotechnical Faculty, University of Ljubljana. The data set comprised records collected for 36 Slovenian Alpine goats during first lactation for measurements and linear scores for ud- der traits. Rear udder attachment (RUAcm), udder depth (UDcm), teat diameter (TDcm) and teat length (TLcm) were measured using flexible canvas tape. Fore (FUA) and rear (RUA) udder attachment, udder depth (UD), me- dial suspensory ligament (MSL), teat placement-rear view (TP), teat diameter (TD) and length (TL) were scored sub- jectively using a nine-point scale. Altogether, 36 first lactation goats were obtained three times over lactation, each scoring was repeated in five con- secutive days, giving 15 measurements for each udder traits per goat with only four goats one scoring missing. In three traits (TD, TDcm, and RUAcm) one observation per each trait was excluded due to data errors. Milk yield of 34 goats was measured four times through lactation and three times for 2 goats. Therefore, 536 records for udder traits and 142 for milk yield was included. Milk yield was obtained by routine milk recording using AT4 method (ICAR Record- ing Guidelines, 2016). 2.2 UDDER TRAIT DEFINITIONS The linear scoring system for dairy goats from American Dairy Goat Association (Linear Appraisal Sys- tem, 2014) was followed to define udder and teat traits. Averages for twelve udder traits are given in Table 1. The term ’’linear’’ refers to the fact that traits are scored on a linear scale from one (1 point) to the other (9 points) biological extreme. For scoring the FUA, the recorder looked at attachment strength of lateral ligaments as they extend forward and laterally to the abdominal wall. The average of FUA was 2.98 scores and was wide from expected average (5 points). An extremely loose attach- ment was assigned to 1 point, an extremely strong attach- ment to 9 point. RUA stands for width and shape of the attachment of the rear udder. An extremely narrow and pointed rear udder attachment was assigned to 1, while extremely wide to 9. The RUA average was higher (4.42) with larger standard deviation (1.62) as FUA. An udder with a weak medial suspensory ligament (MSL), without any cleft but just negative bulge presented, was assigned to 1 point, an udder with clearly defined halving to 5 points, and an udder with an extreme cleft to 9 points. The overall score of MSL (5.43) in our experiment was close to expected average. Udder depth (UD) was defined as vertical distance of udder floor above hocks, scoring 1 for very deep and 9 for extremely high udder. The average was in the middle of the scale with large standard devia- tion. Teat placement (TP) was viewed from the rear. Teats positioned on the outside third of udder half was assigned from 1 to 3 points. Extremely inside positioned teats were scored from 7 to 9 points while vertical positioned teats were valued between 4 and 6 points. The distribution of TP was right skewed (1.39) and the average was closer to minimum scale – extremely outside positioned teats (2.11). Teat diameter (TD) was evaluated from rear at at- tachment to the udder. TD was scored as very narrow (1, 2 or 3 points) to very wide (7, 8 or 9 points). The average was in the middle of the scale (5.13, Table 1) with larger standard deviation as TP. Teat length (TL) was evaluated from the rear as well and set to 1 point if very short and 9 if extremely long. RUA, UD, TP, TD and TL were also measured by tape and indicated with RUAcm, UDcm, TPcm, TDcm and TLcm, respectively. RUAcm was measured as udder width in place where the udder was attached, as viewed from the rear. UDcm was measured as a distance from the attachment place to udder floor. TPcm was measured as a distance between teats. TPcm was determined by the cen- tre of the teat at the point where the teat attached to the udder. TDcm was measured as the diameter of the teat at its base where it meets the left half of the udder, as viewed from the rear. TLcm was measured as a distance from the base of the teat to the end of the teat on the left half of the udder. The distributions of measured traits were symmet- ric with the exception of distributions of RUAcm, which was right skewed (Table 1). The results might indicate Acta agriculturae Slovenica, Supplement 5 – 2016 115 CORRELATIONS BETWEEN UDDER TRAITS ... WITH MILK YIELD DURING FIRST LACTATION IN SLOVENIAN ALPINE GOATS some difficulties in recording practice as RUA and TPcm were normally distributed and alternatives (RUAcm and TP) were right skewed. On the other hand, all goats as- sessed came from one smaller herd and it is too early to make more certain conclusions. 2.3 STATISTICAL ANALYSIS Pearson correlation coefficients were estimated be- tween udder type traits. As udder traits were not taken pairwise on the same day as daily milk yield but as close as possible. In addition, udder traits were assessed 5-times in consecutive days. Linear regression was used to evalu- ate the effect of DIM on udder traits and MY (Model 1). Test for slope heterogeneity was applied among udder traits and milk yield using standardised regression coef- ficients as applied in Model 2. ( )µ= + − +i i iy b x x e (Model 1) µ= + +ij i i ij ijy b x e (Model 2) Standardized individual observation for dependent variable (udder type traits and daily milk yield) is pre- sented by yij, µi intercept of i-th trait, bi regression coef- ficient, nested within i-th trait, xij standardized covariate (days in milk) and eij residual. Statistical analyses were performed by GLM procedure using SAS/STAT® soft- ware, Version 9.4 of the SAS system for Windows (SAS Inst. Inc., 2014). 3 RESULTS AND DISCUSSION Estimates of the phenotypic correlations (Table 2) of udder depth (UD) with fore (FUA, 0.25) or rear (RUA, 0.11) udder attachment were weak. Correlations between udder depth and udder attachment was higher (0.38) in mixed-breed dairy goats published by McLaren et al. (2016) and in Latxa sheep (−0.58) by Legarra and Ugarte (2005). Negative value in Latxa sheep were due to different definition of udder depth on reverse scale. Correlations were weak and negative varying from –0.11 to −0.33 be- tween UD and teat traits TD, TL, TPcm, TDcm and TLcm, indicating that teats were longer, wider, and positioned far away from each other in deep and poorly attached udders. UD was correlated with measured TPcm (−0.33), although less correlated with subjectively scored TP (0.14). Udder length and teat distance was also uncorrelated in Black and Meriz goats as published by Merkhan and Alkass (2011). The correlation between RUA and TL (−0.14), RUA and TLcm (−0.12) and, RUAcm and TL (−0.14) were weak. TL might be shorter wherever udders were well attached. Heritability estimates of udder traits by Manfredi et al. (2001) were moderate, about 0.3, and teat traits (width and length) even higher (0.4). While udder depth was cor- related to teat measurements (width and length), rear and front udder attachment, the selection against ‘’baggy’’ ud- der could cause correlated genetic response in other teat and udder traits. In addition, Rupp et al. (2011) found the genetic correlations between lactation somatic cell scores (LSCS) and type traits to be the highest for teat length Linear assigned udder traits N Mean value SD skewness kurtosis Min Max Fore udder attachment (FUA)* 536 2.98 0.74 −0.05 2.17 1 6 Rear udder attachment (RUA) 536 4.42 1.62 0.11 0.53 1 9 Medial suspensory ligament (MSL) 536 5.43 1.28 0.58 0.89 2 9 Udder depth (UD) 536 5.52 1.23 0.43 0.25 1 9 Teat position (TP)* 536 2.11 1.52 1.39 1.33 1 8 Teat diameter (TD) 535 5.13 2.24 −0.07 −0.98 1 9 Teat length (TL) 536 5.29 2.22 −0.02 −0.79 1 9 Measured udder traits Rear udder attachment (RUAcm), cm* 535 6.14 1.31 1.63 4.69 3.5 13.0 Udder depth (UDcm), cm 536 15.82 2.06 −0.16 −0.20 10.0 21.0 Teat position (TPcm), cm 536 10.22 2.78 0.03 −0.66 3.0 18.0 Teat diameter (TDcm), cm 535 2.76 1.01 0.57 0.11 0.8 6.7 Teat length (TLcm), cm 536 5.82 1.72 0.46 −0.42 2.5 10.5 Days in milk (DIM), days 678 117 40 −0.01 −1.13 49 206 Milk yield, g 142 2055 1111 0.93 1.01 100 5900 N – number of observations; SD – standard deviation;*non normal distribution Table 1: Descriptive statistics for udder traits, days in milk, and milk yield in Slovenian Alpine goats Acta agriculturae Slovenica, Supplement 5 – 2016116 A. CIVIDINI et al. (0.29), teat width (0.34), and teat form (−0.27) in the Al- pine breeds. They reported that the shortest and tightest teats were associated with the lowest LSCS. When compared linear scored udder traits with the same traits but measured, there was a highly positive cor- relation between RUA and RUAcm (0.74), TD and TDcm (0.78), and TL and TLcm (0.86). If RUA was assigned as extremely wide with 7, 8 or 9 points, RUAcm was ex- tremely high as well. Subjective scores for TD and TL and measurements of TDcm and TLcm agreed very much. UD and UDcm were moderately correlated (−0.60). The correlation was negative because extremely deep udder scored with 1, 2 or 3 points were associated by extremely high udder floor. TP and TPcm were correlated below the expectations (−0.41). Udder traits by linear scoring system are offering similar conclusions like equivalent (counter- part) among measurements taken. Experienced scoring experts could achieve sufficient information about udder and teat type traits. An additional problem by measuring udder traits is to restrain an animal in a natural position long enough to make an accurate measurement, especially when the differences measured were small. Linear regression coefficients for all udder traits and daily milk yield were estimated (Table  3). Standardized regression coefficients for all udder traits and daily milk yield were estimated to compare trends. Changes of udder traits over lactation are smaller than changes for MY. De- crease of TP, RUAcm, UDcm, TDcm, and MY throughout lactation was obtained. RUAcm was smaller for 0.37 SD (standard deviation) when lactation progressed for 1 SD. Distance from the attachment place to the floor of the ud- der (UDcm) decreased for 0.41 SD per 1 SD. Milk yield decreased for 0.62 SD per 1 SD longer lactation. UD and FUA RUA MSL UD TP TD TL RUAcm UDcm TPcm TDcm TLcm FUA - 0.47 −0.26 0.25 0.07 −0.20 −0.18 0.30 −0.29 −0.02 −0.18 −0.19 RUA *** - −0.38 0.11 0.09 −0.15 −0.14 0.74 −0.12 0.24 −0.13 −0.12 MSL *** *** - 0.00 0.19 0.28 0.26 −0.33 0.08 −0.30 0.24 0.26 UD *** * ns - 0.14 −0.16 −0.15 −0.05 −0.60 −0.33 −0.17 −0.11 TP ns * *** ** - 0.12 0.14 0.12 −0.04 −0.41 0.15 0.23 TD *** *** *** *** ** - 0.91 −0.13 −0.15 −0.29 0.78 0.83 TL *** ** *** *** ** *** - −0.14 −0.01 −0.28 0.67 0.86 RUAcm *** *** *** ns ** ** * - 0.15 0.34 0.02 −0.08 UDcm *** * ns *** ** ns ns ** - 0.43 0.16 −0.00 TPcm ns *** *** *** *** *** *** *** *** - −0.24 −0.38 TDcm *** * *** *** ** *** *** ns *** *** - 0.75 TLcm *** * *** * *** *** *** ns ns *** *** - FUA = Fore udder attachment; RUA = Rear udder attachment; MSL = Medial suspensory ligament; UD = Udder depth; TP = Teat position; TD = Teat diameter; TL = Teat length; RUAcm = Rear udder attachment, cm; UDcm = Udder depth, cm; TPcm = Teat position, cm; TDcm = Teat diameter, cm; TLcm = Teat length, cm; ns = not significant; * = p < 0.05; ** = p < 0.01; *** = p < 0.001. Table 2: Pearson correlation coefficients among udder traits (above diagonal) with statistical significance (below diagonal) FUAϮ RUA MSL UD TPϮ TD TL RUAcmϮ UDcm TPcm TDcm TLcm MY b̂ 0.001 0.001 0.001 0.005 −0.008 0.003 0.007 −0.012 −0.021 −0.006 −0.007 0.000 −17.212 SEE 0.001 0.002 0.001 0.001 0.002 0.003 0.003 0.001 0.002 0.003 0.001 0.002 1.961 Sig. ns ns ns ** *** ns ** *** *** ns *** ns *** îb 0.03 0.03 0.03 0.16 −0.22 0.06 0.13 −0.37 −0.41 −0.08 −0.27 0.00 −0.62 SEE 0.04 0.04 0.04 0.04 0.04 0.04 0.04 0.04 0.04 0.04 0.04 0.04 0.09 Sig. ns ns ns ** *** ns ** *** *** ns *** ns *** FUA = Fore udder attachment; RUA = Rear udder attachment; MSL = Medial suspensory ligament; UD = Udder depth; TP = Teat position; TD = Teat diameter; TL = Teat length; RUAcm = Rear udder attachment, cm; UDcm = Udder depth, cm; TPcm = Teat position, cm; TDcm = Teat diam- eter, cm; TLcm = Teat length, cm; MY = Milk Yield; b̂ = estimated regression coefficients; îb = estimated standardized regression coefficients; SEE = standard error; Sig = significance; ns = not signifi- cant; * = p < 0.05; ** = p < 0.01; *** = p < 0.001; Ϯ = non normal distribution. Table 3: Linear regression coefficients of days in milk on udder traits and daily milk yield (MY) Acta agriculturae Slovenica, Supplement 5 – 2016 117 CORRELATIONS BETWEEN UDDER TRAITS ... WITH MILK YIELD DURING FIRST LACTATION IN SLOVENIAN ALPINE GOATS TL showed positive trend, although increase was small (0.16 SD and 0.13 SD, respectively). Differences in slope among MY and udder traits were obtained using test for slope heterogeneity. The smallest difference in trend was observed between UDcm and MY, on the other hand, the highest difference in trend was ob- tained with subjectively scored UD, while UD and UDcm were in negatively correlation. Results confirmed the strongest relationship of daily milk production with ud- der depth predicting the cistern volume. These results are in agreement with literature (Manfredi et. al., 2001; Rupp et al., 2011; McLaren et al., 2016). Moreover, RUAcm was decreasing similarly to decrease of MY, with estimated dif- ference in standardized regression coefficient 0.25. When lactation progressed, daily milk yield was smaller and rear udder attachment was shorter. Other traits showed inde- pendent trend comparing to milk production. 4 CONCLUSION There was weak relationship between udder depth and fore udder attachment and, udder depth and rear ud- der attachment. Stronger correlation was evaluated for days in milk with some measured udder traits, suggesting that with progressed lactation the measurements of udder depth, rear udder attachment, and teat diameter become smaller. Negative values of standardised regression coef- ficient of udder traits in relationship with days in milk confirmed the negative trend. The correlations between linear scored udder traits and their measured counterparts were highly positive, suggesting that linear scoring system could be appropri- ate substitution for measuring system. For most of the udder traits we did not find similari- ties in trends with milk yield, with exception for UDcm and RUAcm. With decreasing of milk yield UDcm and RUAcm decreased similar. The current results are not enough informative for general application of the tested scoring system into se- lection programs. As more data become available, future estimates may become more accurate, particularly in terms of the standard errors associated with the estimates observed and the fact that the data originated from just one farm. 5 REFERENCES Altinçekiç, Ş.Ö., Koyuncu, M. (2011). Relationship between Udder Measurements and The Linear Scores for Udder morphology Traits in Kıvırcık, Tahirova and Karacabey Merino Ewes. Journal of the Faculty of Veterinary Medicine, 17(1), 71–76. ICAR Recording Guidelines. (2016). International agreement of recording practices. International Committee for Animal Recording – ICAR. Retrieved from http://www.icar.org/ wp-content/uploads/2016/03/Guidelines-Edition-2016.pdf Kastelic, M., Birtič, D., Bojkovski, D., Cividini, A., Čepon, M., Drašler, D., Gorjanc, G., Klopčič, M., Kompan, D., Komprej, A., Krsnik,J., Potočnik, K., Simčič, M., Zajc, P., Žan-Lotrič, M. (2010). Rejski program za slovensko srnasto pasmo. Domžale: Biotehniška fakulteta, Oddelek za zoot- ehniko. Legarra, A., Ugarte, E. (2005). Genetic parameters of Udder Traits, Somatic Cell Score, and Milk Yield in Latxa Sheep. Journal of Dairy Science, 88, 2238–2245. Linear Appraisal System. (2014). Linear Appraisal Program, American Dairy Goat Association. Retrieved from http:// adga.org/wp-content/uploads/2015/06/LA_BOOKLET.pdf Manfredi, E., Piacere, A., Lahaye, P., Ducrocq, V. (2001). Genet- ic parameters of type appraisal in Saanen and Alpine goats. Livestock Production Science, 70, 183–189. McLaren, A., Mucha, S., Mrode, R., Coffey, M., Conington, J. (2016). Genetic parameters of linear conformation type traits and their relationship with milk yield throughout lac- tation in mixed-breed dairy goats. Journal of Dairy Science, 99(7), 1–10. Merkhan, K.Y., Alkass, J.A. (2011). Influence of udder and teat size on milk yield in Black and Meriz goats. Research Op- tions in Animal & Veterinary Sciences, 1(9), 601–605. Rupp, R., Clément, V., Piacere, A., Robert-Granié, C., Manfredi, E. (2011). Genetic parameters for milk somatic cell score and relationship with production and udder type traits in dairy Alpine and Saanen primiparous goats. Journal of Dairy Science, 94, 3629–3634. SAS Institute Inc. 2014. SAS/STAT® 13.2 User’s Guide. Cary, NC: SAS Institute Inc. FUAϮ RUA MSL UD TPϮ TD TL RUAcmϮ UDcm TPcm TDcm TLcm diff. 0.65 0.64 0.65 0.78 0.41 0.68 0.75 0.25 0.20 0.54 0.34 0.62 SEE 0.10 0.10 0.10 0.10 0.10 0.10 0.10 0.10 0.10 0.10 0.10 0.10 Sig. *** *** *** *** *** *** *** * * *** ** *** FUA = Fore udder attachment; RUA = Rear udder attachment; MSL = Medial suspensory ligament; UD = Udder depth; TP = Teat position; TD = Teat diameter; TL = Teat length; RUAcm = Rear udder attachment, cm; UDcm = Udder depth, cm; TPcm = Teat position, cm; TDcm = Teat diam- eter, cm; TLcm = Teat length, cm; MY = Milk Yield; diff. = estimated difference in standardized regression coefficients; SEE = standard error; Sig = significance; ns = not significant; * = p < 0.05; ** = p < 0.01; *** = p < 0.001; Ϯ = non normal distribution. Table 4: Estimated differences in standardized trend between udder traits and daily milk yield (MY)