Art14_Golier.indd Journal of Applied Botany and Food Quality 85, 216 -223 (2012) Department of Agriculture, Forest and Food Sciences, University of Torino, Grugliasco (TO), Italy Changes in pasture and cow milk compositions during a summer transhumance in the western Italian Alps A. Gorlier, M. Lonati, M. Renna, C. Lussiana, G. Lombardi, L.M. Battaglini (Received July 22, 2012) Summary The changes occurring in pasture and milk compositions during summer grazing were studied following a transhumance of a dairy cattle herd in the western Italian Alps. During three consecutive grazing periods (P1, P2, and P3) the cows exploited, in sequence, mountain pastures located at 1200-1260 m a.s.l. (A1), alpine pas- tures at 2000-2200 m a.s.l. (A2), and then returned to A1 pastures. The botanical and nutritional compositions of pastures, as well as cow milk yield, gross composition and fatty acid (FA) profi le were assessed during the transhumance. Within the pastures, a cluster analysis allowed the recognition of fi ve vegetation types and seven vegetation sub-types; their allocation and plant species composition differed among the exploited grazing areas. The average Pastoral Values were signifi cantly higher in the mountain (A1P1, A1P3) than in the alpine pastures (A2P2) due to the abundance of high- and medium-quality forage species such as Dactylis glomerata L., Polygonum bistorta L., and Festuca rubra s.l.. Nevertheless, the nutritional quality of the herbage offered to the animals did not differ between A1P1 and A2P2, while it was signifi cantly higher in A1P3 due to a younger vegetation phenolo- gical stage. The nutritional parameters were found to be correlated to the pasture botanical composition and phenology: organic matter digestibility and net energy for lactation were correlated negative- ly to the phenological stage and the Specifi c Contribution (SC) of Poaceae and positively to the SC of Fabaceae and Asteraceae. Milk yield signifi cantly declined while milk protein increased during the grazing season, following the advance of cows’ stage of lactation. Milk fat and lactose percentages did not vary signifi cantly among the grazing periods. The same was also observed for milk FA, with the exception of palmitic acid, whose level was lower in A2P2 if com- pared to the other two periods. Signifi cant correlations were found between the percentages of some FA in milk and the SC of the main botanical families of the grazed pastures. In particular, linoleic acid was negatively correlated with the SC of Poaceae and positively cor- related with the SC of Fabaceae. Results showed that the changes in the nutritional composition of pastures depended on variations in pasture botanical composition and phenology at the time of grazing, and that such factors con- curred with animal-related factors in affecting milk quality during the grazing season. Introduction In the Alps, semi-natural pastures are the main feed sources for ruminants during summer. They are traditionally exploited accord- ing to a vertical transhumance, following their gradual availability at growing altitudes (DODGSHON and OLSSON, 2007). In recent years, semi-natural pastures have also become a key factor for the trace- ability and the quality of dairy products (MARTIN et al., 2005). This added value arises from the consideration of the multiple effects of grazing on animal performance. Milk yield and composition are af- fected by a combination of factors including diet (SUTTON, 1989) and animal-related aspects (COULON et al., 1991). Nevertheless, when animals graze, additional variables infl uence milk characteristics. Botanically diverse pastures are known to confer peculiar organolep- tic and nutritional qualities to milk compared with either conserved forages or cereal-based concentrate feeds. In particular, dairy pro- ducts obtained from grass-fed ruminants are very rich in benefi cial fatty acids (FA), such as polyunsaturated fatty acids (PUFA) which have some interest for their positive effects on human health (VAN DORLAND et al., 2006). Furthermore, there is evidence that grazing in alpine areas per se can affect milk fat composition as well. Severe environmental conditions due to altitude, topography, and climate, along with the frequently unstable nutritional quality of herbage, lead to animals’ increased energy requirements and physiological changes (i.e., increased body fat mobilisation, ruminal ecosystem alterations) which have some bearing on milk production (ZEMP et al., 1989; LEIBER et al., 2004). Alpine pasture types (JOUGLET et al., 1992; CAVALLERO et al., 2007) and the factors determining changes in their nutritional and chemi- cal compositions (JEANGROS et al., 1999; BOVOLENTA et al., 2008) have been described in literature. Similarly, the nutritional aspects and the chemical composition of milk produced in alpine areas have been well deepened (COLLOMB et al., 2001; COLLOMB et al., 2002a). Several studies also aimed at understanding the relationships ex- isting between alpine pastures and milk quality during grazing (JEANGROS et al., 1997; BUCHIN et al., 1999; COLLOMB et al., 2002b; DE NONI and BATTELLI, 2008). However, to date, only few of these studies investigated the compositional changes in both pastures and milk following a dairy herd under the usual farming conditions of a transhumance and providing at the same time a full description of pastures’ botanical composition (BUCHIN et al., 1999; DE NONI and BATTELLI, 2008). By contrast, given that such movements of herds are widespread in all European mountain areas and considered the increasing consumers’ interest in traditional dairy products, it should be important to determine which factors usually affect pasture and milk quality in natura. In particular, understanding the effects of changes in pasture botanical composition and timing of grazing on pasture nutritional composition could help in improving grazing management during summer and, consequently, animal perform- ance and the quality of milk and derived products (ELGERSMA et al., 2006). Therefore, the goals of this study were: i) to assess the botanical and nutritional compositions of semi-natural pastures grazed during a transhumance in the western Italian Alps, ii) to assess yield, gross composition, and FA profi le of the milk produced by the herd, and iii) to determine which factors led to changes in pasture and milk compositions during the transhumance. Materials and methods Study sites and experimental design This trial was carried out in the Mont Avic Natural Park (Aosta Valley, Northwest Italy) during an entire summer grazing season. A dairy farm breeding a herd of thirty-six lactating Aosta Red Pied cows was selected as representative of the traditional pasture-based Pasture and cow milk quality changes during a summer transhumance 217 farming system typical of this alpine area. All cows were multi- parous, in mid lactation (176 ± 30 days in milk at the beginning of the trial), and adapted to the alpine environment. Their feed source was limited to fresh grass from pastures. From the beginning of June to the end of July (P1), the herd exploi- ted mountain pastures located at 1200-1260 m a.s.l. (A1). In August (P2), the cows moved to alpine pastures located at 2000-2200 m a.s.l. (A2), while in September (P3) they returned to the mountain pastures already exploited at the beginning of the trial (A1). Thus, during the grazing season three experimental periods were identifi ed (A1P1, A2P2, and A1P3) (Tab. 1). The cows were managed accor- ding to a rotational grazing system, with paddocks grazed in sequence both in the mountain and alpine areas. The surface of paddocks and the stocking periods ranged respectively from 0.4 to 1.6 hectares and from 3 to 8 days depending on vegetation composition, herbage bio- mass, and site conditions (e.g., topography, stone cover). Vegetation surveys Pasture botanical composition was surveyed using a vertical point- quadrat method (DAGET and POISSONET, 1969) along 25 m transects laid out on representative and homogeneous areas. Twenty-two surveys were performed before exploitation (A1P1 and A2P2). Five surveys were repeated in A1 during the regrowth period (P3). All plant species were identifi ed (PIGNATTI, 1982) and their rela- tive cover was computed as Specifi c Contribution (SC) percentage (DAGET and POISSONET, 1969). Vegetation Pastoral Values (PV) were determined according to DAGET and POISSONET (1972). Vegetation sampling and analysis Representative mixed grass samples (500 g) were collected at each grazing area to determine the herbage nutritional composition (Tab. 1). Sampling was carried out two days before exploitation following herd movements within the grazing areas. The pheno- logical stage of the most abundant species was recorded using the Lambertin’s schedule (LAMBERTIN, 1990) and 1×1 m2 area of each pasture was harvested to assess the herbage biomass (t DM ha-1). After collection, the samples were immediately oven-dried at 60°C for 48 hours and ground to pass a 1-mm screen using a Cyclotech mill (Tecator, Höganäs, Sweden). The samples were analyzed for dry matter (DM) (930.15; AOAC, 2000), crude protein (CP) (978.04; AOAC, 2000), neutral detergent fi bre (NDF), acid detergent fi bre (ADF), acid detergent lignin (ADL) (VAN SOEST et al., 1991), or- ganic matter digestibility (OMD) (AUFRÈRE, 1982), and ash (930.05; AOAC, 2000). The net energy for lactation (NEl) was calculated on the basis of the INRA energy system of feedstuffs evaluation (JARRIGE, 1988). Milk sampling and analysis Milk yield recording and milk sampling started in A1P1 after a three- week period provided to allow cow rumen adaptation to summer grazing. In A2P2 and A1P3 milk yield recording and milk sampling started fi ve days after the entrance in each grazing area to support rumen adaptation to vegetation types (Tab. 1). Milk samples to be analyzed for their gross composition were immediately transported to the laboratory in a portable refrigerator at 4°C, while the samples destined to the assessment of the FA composition were frozen until analyzed. Milk fat, protein, and lactose percentages were assessed by infrared spectroscopy (MilkoScan 4000, Foss Electric, Hillerød, Denmark). For FA analysis, milk fat extraction was obtained by centrifugation at 8000 g for fi ve minutes at -4°C. The fatty acid methyl esters (FAME) were prepared by esterifi cation (AOAC, 2000) and determined by us- ing a gaschromatograph (Perkin-Elmer P-E 8700, Perkin-Elmer Co., Norwalk, CT, USA) equipped with a DB-WAX (J&W Scientifi c Inc., Folsom, CA, USA) capillary column (60 m × 0.53 mm ID, 1.0 µm Tab. 1: Features of the grazing season and experimental design. Dairy herd 36 Aosta Red Pied cows Grazing periods June 1st - July 27th July 28th - August 31st September 1st - October 6th codes P1 P2 P3 Grazing areas Mountain pastures Alpine pastures Mountain pastures codes A1 A2 A1 coordinates 45°41’N, 7°36’E 45°39’N, 7°35’E 45°41’N, 7°36’E altitude (m a.s.l.) 1200-1260 2000-2200 1200-1260 surface (ha) 9.0 9.2 6.9 notes fi rst grazing event fi rst grazing event second grazing event Grazing management system Rotational grazing (paddocks) Fodder Fresh grass from pasture Milking system By hand (shed) Surveys and samplings A1P1 A2P2 A1P3 Botanical surveys on the fi rst grass growth, before the fi rst exploitation on the regrowth, before the second exploitation number 6 16 5 Vegetation sampling 2 days before grazing number 7 5 4 Milk sampling 3 weeks after moving 5 days after moving into the grazing area into the grazing area number 7 7 7 218 A. Gorlier, M. Lonati, M. Renna, C. Lussiana, G. Lombardi, L.M. Battaglini fi lm thickness). The column temperature was held at 180°C for one minute and then raised 4°C min-1 to a fi nal temperature of 225°C, where it remained for 45 minutes. The injector was maintained at 250°C and the fl ame ionization detector (FID) at 270°C. Peaks were identifi ed by comparing their retention times with pure FAME standards (Restek Corporation, Bellefonte, PA, USA; Matreya Inc., Pleasant Gap, PA, USA). Milk FA composition was expressed as a percentage of each individual FA per total FA detected. Statistical analysis A two-level classifi cation system, based on pasture types and sub- types, was used to describe pasture vegetation (CAVALLERO et al., 2007). To identify homogeneous vegetation groups – in terms of species composition – botanical data from the fi rst grass growth (A1P1 and A2P2) were classifi ed by cluster analysis performed using the SC (©Clustan Graphics 5.27 software; WISHART, 1987). The similarity matrix was calculated using Pearson’s correlation, while Between-groups linkage was selected as agglomeration method. One-way ANOVA was used to compare the characteristics of pas- tures (i.e., SC of the main botanical families, phenological stage, PV, herbage biomass), the nutritional composition of herbage, yield, gross composition, and FA profi le of milk among the three experi- mental periods. The assumption of equal variances was assessed by Levene’s homogeneity of variance test. The Bonferroni post-hoc test was performed to test the difference between each pair of means. Correlations between the herbage nutritional parameters, milk qua- lity, and vegetation attributes (i.e., SC of the main botanical families, phenological stage) were calculated using the Pearson’s correla- tion test. Statistical analyses were performed using ©SPSS (2007). Signifi cance was declared at P<0.05. Results Botanical composition and attributes of pastures A total of 145 plant species belonging to 38 botanical families were identifi ed in the study area. Five vegetation types and seven vegeta- tion sub-types were recognized (Fig. 1). The botanical composition of types and sub-types is shown in Tab. 2. The average herbage biomass, PV, and SC of the main botanical families in the three ex- perimental periods are given in Tab. 3. Poaceae, Fabaceae, Asteraceae, Polygonaceae, and Cyperaceae were the most widespread families in the pastures (78% of SC). The abundance of Poaceae and Cyperaceae did not differ signifi cantly among the grazing areas. Fabaceae and Polygonaceae species were signifi cantly more abundant in the mountain (A1P1 and A1P3) than in the alpine pastures (A2P2), while Asteraceae species were sig- nifi cantly more abundant in A1P3 with respect to both A1P1 and A2P2 pastures (Tab. 3). Overall, dicotyledonous species SC were signifi cantly higher (P<0.05) in the mountain regrowth (60.9% in A1P3) than in the alpine pastures (37.0% in A2P2). The mountain pastures had signifi cantly higher PV than the alpine pastures. In fact, in the A1P1 area, vegetation sub-types with higher PV ascribable to type Polygonum bistorta and type Festuca rubra s.l. were dominant (about 77% of total A1 surface), while sub-types with lower PV ascribable to type Bromus erectus were confi ned to steep heat slopes covering few areas (about 23% of surface). By con- trast, in the A2P2 pastures, sub-types with lower PV (type Nardus stricta and type Carex sempervirens) were more widespread (about 80% of total A2 surface) than those with higher PV (type Festuca rubra s.l.; about 20% of surface). By comparing A1P1 and A1P3, the mountain pastures were shown not to have undergone signifi cant changes in their PV and their bo- tanical composition did not differ with the exception of Asteraceae abundance. However, the amount of herbage biomass was sig- nifi cantly lower at the regrowth stage than at the beginning of the season. Nutritional values of pastures The phenological stage and nutritional values of pastures during the three experimental periods are shown in Tab. 3. The growth stage of pastures differed signifi cantly among the three periods. On average, pastures were grazed by the cows at the full fl owering stage in A1P1, at the end of the fl owering stage in A2P2, and at the beginning of the heading stage in A1P3. DM percentages did not differ among the experimental periods; Fig. 1: Dendrogram of the vegetation data obtained through cluster analysis. For the botanical composition of vegetation types and sub-types refer to Tab. 2. The survey codes refer to areas (A1 and A2) and to periods (P1 and P2) of the study. Pasture and cow milk quality changes during a summer transhumance 219 T ab . 2 : B ot an ic al c om po si ti on o f th e ve ge ta ti on t yp es a nd s ub -t yp es i n th e m ou nt ai n (A 1P 1) a nd a lp in e (A 2P 2) p as tu re s, a nd b ot an ic al c om po si ti on o f A 1P 3 m ou nt ai n pa st ur es a ft er t he fi r st e xp lo it at io n. T he m ea n (x ) an d S ta nd ar d D ev ia ti on ( S D ) of S pe ci fi c C on tr ib ut io ns a re g iv en f or t he t en m os t ab un da nt s pe ci es o f su b- ty pe s (t he s pe ci es h av in g cu m ul at iv e S C > 25 % a re i n bo ld f on t) . F lo ri st ic n om en cl at ur e fo ll ow s P ig na tt i (1 98 2) . A re a /P er io d M ou n ta in P as tu re s – A 1P 1 (1 20 0- 12 60 m a .s .l .) M ou n ta in P as tu re s – A 1P 3 (1 20 0- 12 60 m a .s .l .) of g ra zi n g ty p e 1 - P ol yg on u m b is to rt a 2 - B ro m u s er ec tu s 3 - F es tu ca r u br a s. l. s u b -t yp e co d e 1. 1 2. 1 3. 1 x S D x S D x S D † x S D x S D P ol yg on u m b is to rt a 12 .4 1. 4 B ro m u s er ec tu s 26 .7 9. 0 T ri se tu m fl a ve sc en s 13 .8 - T ar ax ac u m o ffi c in al e 12 .9 3. 4 T ar ax ac u m o ffi c in al e 23 .9 6. 4 A gr os ti s te n u is 11 .3 7. 3 F es tu ca r u b ra s .l . 9. 7 3. 6 F es tu ca r u br a s. l. 13 .4 - T ri se tu m fl a ve sc en s 12 .0 5. 3 F es tu ca r u br a s. l. 9. 5 7. 2 T ar ax ac u m o ffi c in al e 9. 6 4. 1 C a re x p a ll es ce n s 5. 9 8. 4 C a re x ca ry o p h yl le a 7. 7 - P ol yg on u m b is to rt a 11 .0 2. 3 D a ct yl is g lo m er a ta 6. 7 5. 1 D a ct yl is g lo m er a ta 7. 2 3. 3 L o tu s co rn ic u la tu s 4. 5 0. 8 B ro m u s er ec tu s 6. 9 - T ri fo li u m r ep en s 8. 4 1. 9 P o ly g o n u m b is to rt a 6. 7 7. 6 T ri fo li u m p ra te n se 7. 0 9. 7 T ri se tu m fl a ve sc en s 3. 8 5. 4 L u zu la c a m p es tr is s .l . 6. 1 - D a ct yl is g lo m er a ta 7. 9 4. 6 H o lc u s la n a tu s 6. 4 4. 5 R u m ex a ce to sa 5. 1 3. 9 A g ro p yr o n r ep en s 3. 1 4. 5 A ch il le a m il le fo li u m 5. 7 - A g ro st is t en u is 7. 0 4. 9 A ch il le a m il le fo li u m 6. 3 2. 5 T ri se tu m fl a ve sc en s 5. 0 2. 3 A ch il le a m il le fo li u m 3. 1 4. 3 L a th yr u s p ra te n si s 3. 7 - R a n u n cu lu s a cr is 6. 7 1. 1 T ri fo li u m p ra te n se 6. 2 6. 6 A n th o xa n th u m o d o ra tu m 4. 9 7. 8 O n o n is s p in o sa 2. 8 3. 9 R a n u n cu lu s bu lb o su s 3. 7 - A ch il le a m il le fo li u m 5. 6 3. 7 L o li u m p er en n e 5. 2 3. 8 A ch il le a m il le fo li u m 4. 8 3. 6 H ie ra ci u m p il o se ll a 2. 8 3. 9 L o tu s co rn ic u la tu s 3. 3 - G a li u m m o ll u g o 4. 3 3. 9 P la n ta g o l a n ce o la ta 3. 7 1. 5 R a n u n cu lu s a cr is 3. 7 1. 3 B ra ch yp o d iu m r u p es tr e 2. 7 1. 7 P la n ta g o l a n ce o la ta 2. 4 - L a th yr u s p ra te n si s 3. 9 4. 5 B ra ch yp o d iu m r u p es tr e 3. 5 3. 0 A re a / P er io d of g ra zi n g A lp in e P as tu re s – A 2P 2 (2 00 0- 22 00 m a .s .l .) ty p e 3 - F es tu ca r u br a s. l. 4 - N ar du s st ri ct a 5 - C ar ex s em pe rv ir en s s u b -t yp e co d e 3. 1 3. 2 4. 1 4. 2 5. 1 x S D † x S D x S D x S D x S D F es tu ca r u br a s. l. 22 .0 - F es tu ca r u br a s. l. 18 .2 4. 5 N ar du s st ri ct a 22 .4 7. 8 A n th ox an th u m a lp in u m 11 .8 3. 6 C ar ex s em pe rv ir en s 16 .1 6. 2 T ri se tu m fl a ve sc en s 13 .0 - N ar du s st ri ct a 15 .3 6. 1 C ar ex f u sc a 22 .0 3. 9 N ar du s st ri ct a 11 .6 5. 9 P la n ta go s er pe n ti n a 15 .0 4. 8 P o a a lp in a 12 .4 - P h le u m a lp in u m 14 .3 9. 8 A n th o xa n th u m a lp in u m 11 .4 7. 1 F es tu ca o vi n a s. l. 10 .3 2. 0 L eo n to d o n h el ve ti cu s 14 .1 3. 8 T ri fo li u m p ra te n se 10 .7 - P o a a lp in a 10 .6 9. 3 L eo n to d o n h el ve ti cu s 6. 3 5. 9 G eu m m o n ta n u m 8. 1 2. 2 N a rd u s st ri ct a 6. 5 5. 8 A g ro p yr o n r ep en s 10 .2 - C a re x fu sc a 3. 8 6. 0 C a re x st el lu la ta 5. 8 4. 7 C a re x se m p er vi re n s 5. 9 6. 6 F es tu ca o vi n a s .l . 5. 3 2. 2 P o ly g o n u m b is to rt a 6. 8 - G eu m m o n ta n u m 3. 6 3. 9 F es tu ca o vi n a s .l . 4. 4 6. 7 P o a a lp in a 5. 8 2. 7 A g ro st is a lp in a 4. 9 5. 7 L eo n to d o n h el ve ti cu s 4. 0 - A n th o xa n th u m a lp in u m 3. 5 4. 5 V io la b ifl o ra 3. 9 2. 6 P o te n ti ll a g ra n d ifl o ra 4. 7 5. 6 P o a a lp in a 3. 2 2. 5 A n th o xa n th u m a lp in u m 3. 4 - T ri fo li u m p ra te n se 2. 7 3. 4 P o a a lp in a 3. 1 4. 0 P la n ta g o s er p en ti n a 4. 6 3. 3 S o ld a n el la a lp in a 3. 0 3. 0 P o te n ti ll a g ra n d ifl o ra 3. 4 - S o ld a n el la a lp in a 2. 7 3. 2 S o ld a n el la a lp in a 2. 6 3. 5 L eo n to d o n h el ve ti cu s 4. 0 4. 2 C a m p a n u la s ch eu ch ze ri 2. 7 2. 4 P h le u m a lp in u m 2. 3 - P o te n ti ll a g ra n d ifl o ra 2. 5 2. 4 T ri ch o p h o ru m c a es p it o su m 2 .4 3. 0 A g ro st is t en u is 2. 7 3. 3 A n th o xa n th u m a lp in u m 2. 5 1. 4 † S ub -t yp es i nc lu di ng o nl y on e su rv ey . 220 A. Gorlier, M. Lonati, M. Renna, C. Lussiana, G. Lombardi, L.M. Battaglini however, all other nutritional parameters varied signifi cantly during the season. Most of them did not differ between A1P1 and A2P2, but the pasture quality was signifi cantly higher in A1P3, as demonstrated by the higher CP, OMD, and NEl and the lower NDF. The phenolo- gical phase of plants was signifi cantly correlated with all considered nutritional parameters, with the exception of the ADF and ADL percentages: it was positively correlated to DM and NDF percen- tages and negatively to CP, NEl, and OMD (Tab. 4). Many nutritional parameters were also correlated to the SC of Poaceae, Fabaceae, and Asteraceae. In particular, NEl and OMD were found to be correlated negatively to Poaceae SC and positively to Asteraceae and Fabaceae SC (Tab. 4). Milk yield, gross composition and fatty acid profi le Average bulk milk yield, gross composition and FA profi le in the three grazing periods are shown in Tab. 5. During the transhumance, a decrease in milk yield was observed, specifi cally from 380 kg herd-1 day-1 at the fi rst sampling in June to 60 kg herd-1 day-1 at the last sampling in October. Average milk yield signifi cantly differed among the three experimental periods, being about 50% lower both in A2P2 compared to A1P1 and in A1P3 compared to A2P2. The protein percentage was signifi cantly lower in A1P1 with respect to A2P2 and A1P3. The percentage of fat showed higher values in A2P2 and A1P3 if compared to A1P1 but such variations were not signifi cant. The lactose percentage did not vary signifi cantly Tab. 3: Attributes and nutritional values of the pastures grazed by the herd during the three experimental periods. Area / Period of grazing A1P1 A2P2 A1P3 SEM Inter-area / period Bonferroni post-hoc signifi cant comparisons differences† at P < 0.05‡ Pasture attributes Herbage biomass (t DM ha-1) 3.3 3.0 1.8 0.167 F = 27.605*** A1P1 vs A1P3 A2P2 vs A1P3 Phenological stage§ 400.7 497.2 212.5 29.713 F = 42.310*** A1P1 vs A2P2 A1P1 vs A1P3 A2P2 vs A1P3 Pastoral Value 33.3 20.4 45.4 2.614 F = 15.708*** A1P1 vs A2P2 A2P2 vs A1P3 Poaceae (%) 43.2 47.2 38.3 2.593 F = 0.905 Fabaceae (%) 11.1 3.0 11.9 1.376 F = 6.892** A1P1 vs A2P2 A2P2 vs A1P3 Asteraceae (%) 11.3 7.8 25.6 1.862 F = 12.944*** A1P1 vs A1P3 A2P2 vs A1P3 Polygonaceae (%) 9.7 1.7 9.2 1.254 F = 7.066** A1P1 vs A2P2 A2P2 vs A1P3 Cyperaceae (%) 5.4 15.6 0.0 2.106 F = 7.108 Nutritional values DM (%) 25.0 29.4 21.8 1.417 F = 2.464 CP (%DM) 10.1 10.9 13.3 0.416 F = 12.129** A1P1 vs A1P3 A2P2 vs A1P3 NDF (%DM) 58.1 59.4 41.6 2.115 F = 29.233*** A1P1 vs A1P3 A2P2 vs A1P3 ADF (%DM) 37.2 33.2 29.3 1.026 F = 11.991** A1P1 vs A1P3 ADL (%DM) 6.5 4.4 6.5 0.374 F = 5.613* A1P1 vs A2P2 Ash (%DM) 5.9 4.8 9.1 0.466 F = 35.942*** A1P1 vs A1P3 A2P2 vs A1P3 OMD (%DM) 51.1 50.5 68.0 2.184 F = 21.870*** A1P1 vs A1P3 A2P2 vs A1P3 NEl (MJ kg DM-1) 3.9 3.8 5.2 0.165 F = 19.822*** A1P1 vs A1P3 A2P2 vs A1P3 Abbreviations: DM, dry matter; CP, crude protein; NDF, neutral detergent fi bre; ADF, acid detergent fi bre; ADL, acid detergent lignin; OMD, organic matter digestibility; NEl, net energy for lactation. †Signifi cance: *P<0.05; **P<0.01; ***P<0.001. ‡vs = versus. §Lambertin phenology scale (P = Poaceae and Cyperaceae; O = Other families): 100 = P/O: vegetative stage; 200 = P: 70% of spikes in stems, O: 70% of fl owering bottoms; 300 = P: 70% of spikes out of stems (spikes close to stems), O: 70% of fl owering bottoms opened; 400 = P/O: full fl owering stage; 500 = P: lactic corns, O: fl owers withered; 600 = P: doughy corns, O: starting fructifi cation (fruits formed); 700 = P: hard corns, O: fruits fully ripened; 800 = P/O: end of vegetation. Tab. 4: Pearson’s correlation coeffi cients† between the nutritional values of the grazed herbage, the phenological stages, and the Specifi c Contributions of the main botanical families. DM NDF ADF ADL CP NEl OMD Phenological stage 0.625** 0.784** 0.353 −0.474 −0.548* −0.752** −0.747** Poaceae 0.555* 0.460 −0.139 −0.857** −0.391 −0.504* −0.503* Fabaceae −0.408 −0.531* 0.033 0.580* 0.367 0.547* 0.534* Asteraceae −0.540* −0.549* −0.111 0.699** 0.454 0.642** 0.630** Polygonaceae −0.356 −0.410 0.100 0.470 0.225 0.433 0.410 Cyperaceae 0.440 0.391 0.103 −0.434 −0.160 −0.448 −0.418 Abbreviations: DM, dry matter; NDF, neutral detergent fi bre; ADF, acid detergent fi bre; ADL, acid detergent lignin; CP, crude protein; NEl, net energy for lacta- tion; OMD, organic matter digestibility. †Signifi cance: *P<0.05; **P<0.01. Pasture and cow milk quality changes during a summer transhumance 221 throughout the grazing season. The FA percentages of milk did not show signifi cant differences among periods, except in the case of palmitic acid (C16:0), whose level was signifi cantly lower in alpine (A2P2) than in mountain milk (A1P1 and A1P3). Signifi cant correlations were found between the percentages of some FA in milk and the SC of the main botanical families of the grazed pastures (Tab. 6). Palmitic acid was positively correlated with the SC of Polygonaceae. Oleic acid (C18:1 c9) was negatively correlated with the SC of Polygonaceae and positively correlated with the SC of Cyperaceae. Linoleic acid (C18:2 c9c12) was negatively corre- lated with the SC of Poaceae and positively correlated with the SC of Fabaceae. The total monounsaturated fatty acid (MUFA) percentage was negatively correlated with the SC of Polygonaceae and posi- tively correlated with the SC of Cyperaceae. The total PUFA per- centage was instead positively correlated with the SC of Fabaceae. No signifi cant correlations were observed between milk FA and the SC of Asteraceae. Discussion In the alpine farming systems, dominant plant species and available herbage biomass at the time of grazing, along with the topographic features of pastures (e.g., altitude, slope), are usually the main factors taken into account to set grazing management strategies (JOUGLET et al., 1992). PV, as deduced from vegetation composition, is parti- cularly considered by pastoralists the most effective tool to assess pasture carrying capacity (DAGET and POISSONET, 1972). Instead, other relevant factors such as plant phenology hardly support ma- nagement decisions because of logistic constraints, with consequen- ces on the quality of herbage actually offered to ruminants during the grazing seasons. In the study area, mountain and alpine pastures, whose vegetation types and sub-types can be considered representa- tive of those usually grazed in the western Alps (CAVALLERO et al., 2007), showed signifi cant differences in their botanical composition and PV. In particular, the average PV were higher in the mountain than in the alpine pastures due to the abundance of high- and me- dium-quality forage species such as Dactylis glomerata, Polygonum bistorta, and Festuca rubra s.l.. Nevertheless, mountain and alpine pasture fi rst growth did not display, as expected, signifi cant dif- Tab. 6: Signifi cant Pearson’s correlation coeffi cients between milk fatty acids and the Specifi c Contributions of the main botanical families in pastures. Milk fatty acids Botanical families (Pearson’s correlation coeffi cient†) C16:0 Polygonaceae (r = 0.614*) C18:1 c9 Polygonaceae (r = −0.587*); Cyperaceae (r = 0.643*) C18:2 c9c12 Poaceae (r = −0.610*); Fabaceae (r = 0.697*) ΣMUFA Polygonaceae (r = −0.589*); Cyperaceae (r = 0.603*) ΣPUFA Fabaceae (r = 0.653*) Abbreviations: MUFA, monounsaturated fatty acids; PUFA, polyunsaturated fatty acids. †Signifi cance: *P<0.05. Tab. 5: Bulk milk yield, gross composition and fatty acid (FA) profi le during the three experimental periods. Area / Period of grazing A1P1 A2P2 A1P3 SEM Inter-area / period Bonferroni post-hoc signifi cant comparisons differences† at P < 0.05‡ Milk yield (kg herd-1 day-1) 325.8 160.0 86.0 27.546 F = 77.209*** A1P1 vs A2P2 A1P1 vs A1P3 A2P2 vs A1P3 Milk gross composition (%) Fat 3.87 4.23 4.14 0.077 F = 1.952 Protein 3.21 3.63 3.75 0.060 F = 26.255*** A1P1 vs A2P2 A1P1 vs A1P3 Lactose 4.82 4.72 4.76 0.019 F = 3.020 Fatty Acids (% of total FA) C10:0 1.97 2.38 2.26 0.110 F = 1.164 C12:0 2.26 2.98 2.75 0.161 F = 1.796 C14:0 9.70 8.67 9.77 0.283 F = 1.948 C14:1 c9 1.27 1.23 1.44 0.057 F = 1.297 C15 aiso 0.92 0.97 0.90 0.028 F = 0.514 C15:0 1.62 1.74 1.97 0.065 F = 3.088 C16:0 25.40 23.62 25.55 0.357 F = 5.305* A1P1 vs A2P2 A2P2 vs A1P3 C16:1 c9 1.35 1.47 1.37 0.105 F = 0.121 C18:0 17.63 17.27 16.15 0.397 F = 1.213 C18:1 c9 34.59 36.98 34.45 0.567 F = 2.880 C18:2 c9c12 1.79 1.48 1.88 0.146 F = 0.723 C18:3 c9c12c15 1.52 1.23 1.52 0.123 F = 0.649 ΣSFA 59.49 57.61 59.35 0.630 F = 0.981 ΣMUFA 37.21 39.68 37.26 0.557 F = 2.955 ΣPUFA 3.30 2.71 3.40 0.263 F = 0.716 ΣUFA 40.52 42.39 40.66 0.629 F = 0.974 Abbreviations: SFA, saturated fatty acids; MUFA, monounsaturated fatty acids; PUFA, polyunsaturated fatty acids; UFA, unsaturated fatty acids. †Signifi cance: *P<0.05; ***P<0.001. ‡vs = versus. 222 A. Gorlier, M. Lonati, M. Renna, C. Lussiana, G. Lombardi, L.M. Battaglini ferences in herbage allowance and quality, as they were both grazed at advanced phenological stages with digestibility levels extremely low for dairy cows (JARRIGE, 1988). On the contrary, the regrowth of mountain pastures in September, which did not differ from the fi rst growth for its PV, was found to have the highest nutritional quality. During the entire grazing season plant phenology concurred strongly with plant species composition in determining such changes in pas- ture nutritional profi les. Plant growth normally results in increased fi bre content and decreased energy values (SCHUBIGER et al., 2001), but it is known that the extent to which nutritional values shift and their rapidity depend on the species, being generally greater in grass- lands rich in Poaceae than in those rich in dicotyledonous (JARRIGE, 1988; FARRUGGIA et al., 2008). Moreover, regrowths are known to have a higher leaf:stem ratio and anticipated development stages, which usually determine higher digestibility values and lower pro- ductivity relative to the fi rst growth (SCHUBIGER et al., 2001). The differences among pastures and the correlations between nutritional values, phenological stages and the SC of Poaceae, Asteraceae, and Fabaceae confi rmed such tendencies. Although plant senescence effects are usually well known by farmers, in alpine environments short vegetative season normally determines rapid plant develop- ment and decline in herbage quality (VAN DORLAND et al., 2006), while differences in site conditions and plant species composition lead to a high variability of growth rates among pastures (FARRUGGIA et al., 2008), which altogether do not allow to easily manage the timing of grazing during the transhumances. In the western Alps, pasture quality changes such as those observed in this study are commonly considered to affect milk yield and composition during the summer grazing season (BATTAGLINI et al., 2003). Given that in the study area calving traditionally occurs in December (BARMAZ, 1992), a drop in milk yield and some changes in milk nutritional composition were expected due to the advance in the cows’ stage of lactation. However, many other factors affected milk yield and composition throughout the season, especially as a consequence of herd transfers between mountain and alpine areas. The general grazing conditions in alpine pastures (i.e., walking, altitude) coupled with low herbage quality, likely concurred in reducing milk yield by 50% in a short period as A2P2 (about one month). In fact, high altitude grazing adaptation can be generally achieved by the animals within a few weeks, but it is associated with a concomitant decline in herbage intake and milk yield due to animal stress (VAN DORLAND et al., 2006). Instead, further decrease in milk yield after return to A1P3 occurred mainly as cows were at the end of the lactation. Since herbage allowance was lower compared to the previous considered periods, regrowth quality probably allowed sup- porting lower animal requirements at the end of the grazing season (BOVOLENTA et al., 2008). Concerning milk gross composition, fat was expected to increase signifi cantly during lactation as it was observed for protein. Never- theless, fat is known as the most sensitive parameter in milk to dietary infl uences (SUTTON, 1989) and percentages varied notably during the entire season probably because of variations in herbage quality. In particular, although changes were not signifi cant, milk fat values increased in A2P2 compared to A1P1 in correspondence to the drop in milk yield and the supposed increased body fat mobili- sation due to energy defi cits (LEIBER et al., 2006). No signifi cant variations were also observed for lactose, but such result was ex- pected as this parameter is known to be approximately constant in milk (SUTTON, 1989). In this study, the FA levels in milk were generally consistent with those of previous reports for cows grazing natural pastures (COLLOMB et al., 1999, 2002a; DE NONI and BATTELLI, 2008). However, FA percentages did not vary as expected in correspon- dence of variations in pasture botanical composition and quality. In fact, when cows graze grasslands rich in dicotyledonous plant species or at young growth stages, as in A1P3, higher concentrations of long-chain MUFA and PUFA (i.e. oleic, linoleic, and α-linolenic acids), along with lower concentrations of saturated fatty acids (SFA) (i.e., myristic and palmitic acids) are expected in milk fat be- cause of the higher availability of long-chain unsaturated fatty acid (UFA) precursors in the diet (COLLOMB et al., 2001). Instead, only palmitic acid percentage varied signifi cantly during the season, be- ing lower in the alpine area in correspondence of lower abundances of dicotyledonous such as Fabaceae and Polygonaceae species. Since it is well known that feeding has a dominant role on milk FA composition if compared to animal-related factors (PALMQUIST et al., 1993) and that other factors (e.g., altitude, botanical diversity) can affect milk FA as well (LEIBER et al., 2005), it is hypothesised that the overall conditions of the grazing areas coupled with their botani- cal and nutritional composition concurred to the values observed in milk FA. In particular, the synthesis of short- and medium-chain FA by the mammary gland usually declines notably when cows are in negative energy balance and when a proportionately high dietary fi bre causes digestive modifi cations in the animals, as during alpine grazing, confi rming that the A2P2 grass NEl was likely insuffi cient to cover animal energy needs (PALMQUIST et al., 1993; KHANAL et al., 2008). Although animal intake was not taken into account in the present study, it is also reasonable to hypothesise that herbage intake varied during the grazing season, being lower as expected in A2P2 because of the effect of alpine grazing conditions (LEIBER et al., 2006). Finally, the correlations found between milk FA compo- sition and the SC of some botanical families support the role played by these families in the synthesis of some FA pools (COLLOMB et al., 2002b). In particular, Fabaceae and Cyperaceae abundances con- fi rmed to be positively correlated, while Poaceae and Polygonaceae were negatively correlated, with some representative individual FA as well as MUFA and PUFA in milk. By contrast, no signifi cant cor- relations involved other dicotyledonous families (e.g., Asteraceae, Apiaceae, and Rosaceae) whose presence has been reported to in- crease the level of PUFA in other studies (COLLOMB et al., 2002b; DE NONI and BATTELLI, 2008). In conclusion, considering pasture and milk compositions together allowed understanding which factors normally affect animal per- formance during traditional transhumances of dairy cow herds. In mountain and alpine areas with high-diverse pasture types, botani- cal and PV evaluations could be not suffi cient, per se, to assure a constant herbage quality to dairy cows. Managing the botanical composition along with the timing of grazing can be an effective strategy for optimizing the quality of the herbage offered to rumi- nants throughout a grazing season, in particular for vegetation types dominated by Poaceae species. Early exploitation in both mountain and alpine areas would generally lead to higher herbage energy values, improving animal performances and the quality of milk and derived dairy products. Acknowledgements The authors gratefully acknowledge Prof. Andrea Cavallero for su- pervision of the project. The authors wish also to thank Massimo Bocca and the Mont Avic Natural Park for support and technical as- sistance in the fi eldwork, Sergio Valsoaney and Jean Paul Allasinaz for fi eldwork, Claudio and Roberto Bosonin for allowing the study to be performed in their farm. The research was supported by a grant from the Région Autonome Vallée d’Aoste and the Mont Avic Natural Park. References AOAC, 2000: Offi cial methods of analysis of AOAC International. 17th Edition. Association of Offi cial Analytical Chemists, Gaithersburg, MD, Pasture and cow milk quality changes during a summer transhumance 223 USA. AUFRÈRE, J., 1982: Etude de la prévision de la digestibilité des fourrages par une méthode enzymatique. Ann. Zootech. 31, 111-130. BARMAZ, A., 1992: La qualità del latte durante il periodo di monticazione estiva in alpeggio in Valle d’Aosta. Annales de l’Institut Agricole Régional 1, 537-552. BATTAGLINI, L.M., MIMOSI, A., MALFATTO, V., ROBIN PREILLAN, A., BIANCHI, M., 2003: Management and productive aspects of a dairy herd in a Valle d’Aosta alpine pasture. Ital. J. Anim. Sci. 2, 293-295. BOVOLENTA, S., SPANGHERO, M., DOVIER, S., ORLANDI, D., CLEMENTEL, F., 2008: Chemical composition and net energy content of alpine pasture species during the grazing season. Anim. Feed Sci. Tech. 146, 178-191. BUCHIN, S., MARTIN, B., DUPONT, D., BORNARD, A., ACHILLEOS, C., 1999: Infl uence of the composition of Alpine highland pasture on the chemical, rheological and sensory properties of cheese. J. Dairy Res. 66, 579-588. CAVALLERO, A., ACETO, P., GORLIER, A., LONATI, M., LOMBARDI, G., MARTINASSO, B., TAGLIATORI, C., 2007: I tipi pastorali delle Alpi piemontesi. Alberto Perdisa Editore, Bologna, Italy. COLLOMB, M., BUTIKOFER, U., SIEBER, R., BOSSET, J.O., JEANGROS, B., 2001: Conjugated linoleic acid and trans fatty acid composition of cows’ milk fat produced in lowlands and highlands. J. Dairy Res. 68, 519-523. COLLOMB, M., BUTIKOFER, U., SIEBER, R., JEANGROS, B., BOSSET, J.O., 2002a: Composition of fatty acids in cow’s milk fat produced in the Lowlands, Mountains and Highlands using high resolution gas chroma- tography. Int. Dairy J. 12, 649-659. COLLOMB, M., BUTIKOFER, U., SIEBER, R., JEANGROS, B., BOSSET, J.O., 2002b: Correlation between fatty acids in cows’ milk fat produced in the Lowlands, Mountains and Highlands of Switzerland and botanical composition of the fodder. Int. Dairy J. 12, 661-666. COLLOMB, M., BÜTIKOFER, U., SPAHNI, M., JEANGROS, B., BOSSET, J.O., 1999: Composition en acides gras et en glycérides de la matière grasse du lait de vache en zones de montagne et de plaine. Sci. Aliment. 19, 97-110. COULON, J.B., CHILLIARD, Y., RÉMOND, B., 1991: Effets du stade physio- logique et de la saison sur la composition chimique du lait de vache et ses caractéristiques technologiques (aptitude à la coagulation, lipolyse). INRA Prod. Anim. 4, 219-228. DAGET, P., POISSONET, J., 1969: Analyse phytologique des prairies – Applications agronomiques, Document 48. Centre National de la Recherche Scientifi que B.P., Montpellier, France. DAGET, P., POISSONET, J., 1972: Un procédé d’estimation de la valeur pastorale des pâturages. Fourrages 49, 31-39. DE NONI, I., BATTELLI, G., 2008: Terpenes and fatty acid profi les of milk fat and ‘‘Bitto” cheese as affected by transhumance of cows on different mountain pastures. Food Chem. 109, 299-309. DODGSHON, R.A., OLSSON, E.G.A., 2007: Seasonality in European mountain areas: A study in human ecology. In: Palang, H., Sooväli, H., Printsmann A. (eds.), Seasonal Landscapes, 85-101. Springer, Dordrecht, Neder- land. ELGERSMA, A., TAMMINGA, S., ELLEN, G., 2006: Modifying milk composi- tion through forage. Anim. Feed Sci. Tech. 131, 207-225. FARRUGGIA, A., MARTIN, B., BAUMONT, R., PRACHE, S., DOREAU, M., HOSTE, H., DURAND, D., 2008: Quels intérêts de la diversité fl oristique des prairies permanentes pour les ruminants et les produits animaux? INRA Prod. Anim. 21, 181-200. JARRIGE, R., 1988: Alimentation des bovins, ovins & caprins. INRA Publications, Paris, France. JEANGROS, B., SCEHOVIC, J., TROXLER, J., BACHMANN, H.J., BOSSET, J.O., 1999: Comparaison de caractéristiques botaniques et chimiques d`herbages pâturés en plaine et en montagne. Fourrages 159, 277-292. JEANGROS, B., TROXLER, J., CONOD, D., SCEHOVIC, J., BOSSET, J.O., BÜTIKOFER, U., GAUCH, R., MARIACA, R., PAUCHARD, J.P., SIEBER, R., 1997: Etude des relations entre les caractéristiques des herbage et celles du lait, de la crème et du fromage de type l’Etivaz ou Gruyère: I. Présentation du projet. Rev. Suisse Agric. 29, 23-34. JOUGLET, J.-P., BORNARD, A., DUBOST, M., 1992: Eléments de pastoralisme montagnard. Tome 1: Végétation. Equipements. Etudes du Cemagref, série Montagne n° 3. Cemagref, Gap, France. KHANAL, R.C., DHIMAN, T.R., BOMAN, R.L., 2008: Changes in fatty acid composition of milk from lactating dairy cows during transition to and from pasture. Livest. Sci. 114, 164-175. LAMBERTIN, M., 1990: Calcul de la charge d’un alpage par une approche phyto-ecologique. Institut Agricole Régional Aoste Publ., Aosta, Italy. LEIBER, F., KREUZER, M., JÖRG, B., LEUENBERGER, H., WETTSTEIN, H.R., 2004: Contribution of altitude and Alpine origin of forage to the infl u- ence of Alpine sojourn of cows on intake, nitrogen conversion, metabolic stress and milk synthesis. Anim. Sci. 78, 451-466. LEIBER, F., KREUZER, M., LEUENBERGER, H., WETTSTEIN, H.R., 2006: Contribution of diet type and pasture conditions to the infl uence of high altitude grazing on intake, performance and composition and renneting properties of the milk of cows. Anim. Res. 55, 37-53. LEIBER, F., KREUZER, M., NIGG, D., WETTSTEIN, H.R., SCHEEDER, M.R.L., 2005: A study on the causes for the elevated n-3 fatty acids in cows’ milk of alpine origin. Lipids 40, 191-201. MARTIN, B., VERDIER-METZ, I., BUCHIN, S., HURTAUD, C., COULON, J.B., 2005: How do the nature of forages and pasture diversity infl uence the sensory quality of dairy livestock products? Anim. Sci. 81, 205-212. PALMQUIST, D.L., BEAULIEU, A.D., BARBANO, D.M., 1993: Feed and animal factors infl uencing milk fat composition. J. Dairy Sci. 76, 1753-1771. PIGNATTI, S., 1982: Flora d’Italia. Edagricole, Bologna, Italy. SCHUBIGER, F.X., LEHMANN, J., DACCORD, R., ARRIGO, Y., JEANGROS, B., SCEHOVIC, J., 2001: Valeur nutritive des plantes de prairies. 5. Digestibilité de la matière organique. Rev. Suisse Agric. 33, 275-279. SPSS, 2007: User’s Guide, Base 16.0 for Windows. Statistical Package for Social Sciences Inc., Chicago, IL, USA. SUTTON, J.D., 1989: Altering milk composition by feeding. J. Dairy Sci. 72, 2801-2814. VAN DORLAND, H.A., WETTSTEIN, H.-R., KREUZER, M., 2006: Species-rich swards of the Alps: constraints and opportunities for dairy production. In: Elgersma, A., Dijkstra, J., Tamminga, S. (eds.), Fresh Herbage for Dairy Cattle, 27-43. Springer, Dordrecht, Nederland. VAN SOEST, P.J., ROBERTSON, J.B., LEWIS, B.A., 1991: Methods for dietary fi ber, neutral detergent fi ber, and nonstarch polysaccharides in relation to animal nutrition. J. Dairy Sci. 74, 3583-3597. WISHART, D., 1987: Clustan User Manual. University of Edinburgh, Edinburgh, UK. ZEMP, M., LEUENBERGER, H., KÜNZI, N., BLUM, J.W., 1989: Infl uence of high altitude grazing on productive and physiological traits of dairy cows. I. Infl uence on milk production and body weight. J. Anim. Breed. Genet. 106, 278-288. Address of the authors: Alessandra Gorlier (corresponding author: alessandra.gorlier@unito.it), Michele Lonati, Manuela Renna, Carola Lussiana, Giampiero Lombardi, and Luca Maria Battaglini, Department of Agriculture, Forest and Food Sciences, University of Torino, Via. L. da Vinci 44, 10095 Grugliasco (TO), Italy.