The value of safflower (Carthamus tinctorius) hay and silage grown under Mediterranean conditions as forage for dairy cattle

Livestock Production Science 88 (2004) 263 – 271 www.elsevier.com/locate/livprodsci

The value of safflower (Carthamus tinctorius) hay and silage grown under Mediterranean conditions as forage for dairy cattle S. Landau a,*, S. Friedman a, S. Brenner a, I. Bruckental b, Z.G. Weinberg c, G. Ashbell c, Y. Hen c, L. Dvash a, Y. Leshem a a

Department of Natural Resources and Agronomy, Institute of Field and Garden Crops, Agricultural Research Organization, The Volcani Center, P.O. Box 6, Bet Dagan 50250, Israel b Department of Cattle Nutrition and Physiology, Institute of Animal Science, Agricultural Research Organization, The Volcani Center, P.O. Box 6, Bet Dagan 50250, Israel c Forage Preservation and By-Products Research Unit, Agricultural Research Organization, The Volcani Center, P.O. Box 6, Bet Dagan 50250, Israel Received 17 April 2003; received in revised form 30 September 2003; accepted 4 November 2003

Abstract The value for dairy cattle of safflower grown under Mediterranean condition was investigated in two experiments. In experiment 1, safflower hay was given ad libitum as sole food to four dry dairy cows. The DM ingested from hay was of medium CP and NDF contents (148 and 406 g kg 1, respectively). Values of in vivo and in vitro Tilley and Terry DM digestibility were 723 and 646 g kg 1 DM, respectively. In the second experiment, 19 cows were fed a total mixed ration (TMR) including 4 kg (as DM) of corn plus wheat (CW) silage, and another 19 received the same TMR, with safflower silage (S) substituted for CW silage, on the same DM basis, for 62 days. Diets were of similar NDF content (314 and 331 g kg 1 DM, for CW and S, respectively), but cows fed S consumed less DM than those fed CW (20.2 and 22.5 kg, P < 0.02). Milk production (30.2 kg day 1), and the contents of fat (35.4 g kg 1), lactose (46.4 g kg 1), and urea (0.32 g kg 1) were similar between groups. Milk CP tended to be lower in S than in CW (31.6 and 33.6 g kg 1, respectively, P = 0.07). Changes in body live-weight and condition score were not affected by diet. Safflower silage has the potential for widespread adoption as a feed in Mediterranean countries, if special characteristics such as protein degradability are taken into account to optimize its inclusion in TMRs. D 2004 Elsevier B.V. All rights reserved. Keywords: Dairy cattle; Nutrition and feeding; Unconventional forages; Compositae

1. Introduction The most widespread cropping system in Eastern Mediterranean basin consists of a rotation of wheat for * Corresponding author. Tel.: +972-39-68-3492; fax: +972-3966-9642. E-mail address: [email protected] (S. Landau). 0301-6226/$ - see front matter D 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.livprodsci.2003.11.011

grain or forage and legumes (vetch and garden pea) for hay. Cereal weeds are easily controlled in crops of dicotyledonous species, among which legumes have well-known advantages as cultivated fallow for cereal grasses. However, the yields of legumes and their return on investment are low, therefore, continuous wheat is widely practiced, with consequent problems of grass weeds becoming increasingly resistant to

264

S. Landau et al. / Livestock Production Science 88 (2004) 263–271

herbicides, and elevated prices of non-grass forages. From thze agronomical point of view, all dicotyledonous species, and not only legumes, are suitable for rotation with cereal grasses. Plants of the Compositae family hold potential as forage for ruminants. The most prominent examples are chicory (Cichorium intybus L.) which is grown as forage in New Zealand (Hume et al., 1995), or Chrysantemium coronarium L. in Italy (Sulas and Caredda, 1997). Safflower (Carthamus tinctorius L.), a strongly tap-rooted annual plant from the same family, is native to the Middle East. It is resistant to saline condition (Francois and Berstein, 1964) and to moisture stress (Bassiri et al., 1977), and reaches deep-lying water (Aase and Pikul, 2000). The in vivo digestibility and the intake of green safflower fodder were similar to those of a vetch –oats mixture (Vonghia et al., 1992). Grazed safflower was shown to support satisfactory growth rates in Australian steers (French et al., 1988) and to improve fertility in Canadian ewes (Stanford et al., 2001). Safflower cropped at the budding stage can be ensiled (Weinberg et al., 2002), but we are not aware of published data on the intake or the nutritional value of safflower grown as forage, hay or silage, and fed to housed dairy cattle. The aim of the present study was to assess the nutritional value for dairy cattle of safflower, cultivated for hay or silage, under Mediterranean conditions.

2. Materials and methods 2.1. Safflower hay and silage Because of the lack of certainty regarding safflower and the potential economic cost of feeding highyielding dairy cows inadequately, we carried out a preliminary experiment (experiment 1) in which hay was fed to dry cows, before growing safflower and feeding it to lactating animals (experiment 2). In experiment 1, spineless safflower (cv. PI251285, originated in Jordan) was sown on 26 January 2001 (25 kg of seeds ha 1) at the Migda farm in the northern Negev Desert of Israel (31j18VN, 34j38VE), in deep loess soil. From the beginning of January there was total rainfall of 176 mm, and, in addition, the crop was irrigated twice with 500 m3 ha 1 of purified sewage water in February. Weeds were con-

trolled by spraying Chlorsulfuron at 20 g ha 1. The crop was harvested on 30 April at the late budding stage (before blooming) and baled after 1 week, and the biomass DM yield was 3.5 tonnes ha 1. The final composition of the hay is presented in Table 1. The degradation kinetics of the hay was evaluated to in situ (Ørskov and McDonald, 1979) in two rumenfistulated sheep. Milled (2 mm) hay samples (5 g DM each) in dacron bags (pore size 45 Am) were placed in the rumen and were incubated for 3, 6, 9, 12, 24, 36, 48, and 72 h in two replicates for each sheep. At the end of the incubation period, all the bags were washed gently in a washing machine. The soluble fraction was determined by placing bags in water at 39 jC for 30 min. Data were fitted to the exponential equation: D = a + b[1 exp( ct)] where D is the mass degraded in the rumen after t hours, a is the mass of soluble material that was degraded immediately, b is the fraction that was degraded slowly, and c is the rate of degradation of fraction b. The effective ruminal degradability (ED) was calculated as ED = a+[(bxc)/ (c + k)],where k is the ruminal outflow rate. The safflower used in experiment 2 was sown in 5 ha of the Bet Dagan experimental farm (32jN, 35jE), on 15 January 2002, after disking and fertilizing (140 kg N, 50 kg P ha 1). The seeding rate was 25 kg ha 1. A pre-emergence seed control treatment (trifuralin at 2000 ml ha 1) was applied 14 days before seeding. An additional 60 kg of N and 25 kg of P were applied as top fertilization in March, after it appeared that rainfall was favourable (611 mm). The crop was Table 1 Composition of feedstuffs in experiments 1 and 2 DM basis (g kg

1

)

Ash

CP

NDF

ADF

In vitro digestibility

Experiment 1 Safflower hay

108a

134

429

298

625

Experiment 2 Wheat silage Corn silage Safflower silage

103 45 125

91 72 156

563 448 461

339 233 338

607 707 652

93 77

177 163

334 319

177 157

759 780

TMRs Safflower group (S) Corn and wheat (CW) a

Ash composition: Ca, 12 g kg

1

; P, 1.7 g kg

1

DM.

S. Landau et al. / Livestock Production Science 88 (2004) 263–271

cut (cutting height 18 cm) on 25 April, when the DM content was 190 g kg 1; it was wilted for 2 days, and chopped-harvested to a fiber length of 1.5 cm. The total biomass DM yield was 10 tonnes ha 1, of which 7 tonnes ha 1 were harvested, and piled on a concrete double-walled silage clamp 8 m in width and 1.8 m in height. The DM content in the ensiled safflower was 298 g kg 1. The safflower crop was compacted by running a bulldozer over it, and covered with a 1.5 mm polyethylene sheet. Paper mill waste was spread on top of the plastic to ensure anaerobic condition—a routine in commercial dairy operations. At time of ensiling, pH, DM content (g kg 1), and the logarithms of the counts (per gram DM) of lactic bacteria, yeast, and moulds, were 5.47, 292, 5.7, 5.7, and 4.9. The silage clamp was opened at the beginning of August. Silage characteristics at time of clamp opening, obtained as described by Weinberg et al. (2002), were: pH, 4.67 F 0.03 (S.E.); DM content, 296 F 0.2 (S.E.) g kg 1; water-soluble carbohydrates, 17 F 1.7 g kg 1; log (lactic bacteria count), 8.2; lactic acid, 56 g kg 1; ash, 125 g kg 1; ethanol, 1.7 g kg 1, acetic acid, 20.2 g kg 1. No trace of yeasts or moulds could be found. Silage composition is presented in Table 1. 2.2. Animals and management Experiment 1 was carried out at the Metabolic Cow Unit of the Agricultural Research Organization (ARO) at Bet Dagan. Four mature, rumen-fistulated, dry, 7month-pregnant Holstein Israeli cows, were managed in restrained stabulation, but were released to an openair yard for several hours daily. Experiment 2 was carried out at the ‘‘individual cow-shed’’ unit of the ARO at Bet Dagan with 42 multiparous cows aged 4.4 F 0.3 years, weighing 599 F 12 kg, 205 F 18 days in milk (DIM), that yielded 31.7 F 12 kg milk. They were housed on concrete floor in a roofed corral. The cows were equipped with a passive transponder (Texas Instruments, Dallas, TX) that enabled each cow to open only her own feeding station. The feeding stations were troughs mounted on digital scales (Merav 2001; Shekel Balances, Rosh Ha-Ayin, Israel) by which the feed was continuously weighed with an accuracy of F 20 g, and that were permanently connected to a computer. Individual intake was recorded for all days except for one, when there was a power failure.

265

Each cow was also equipped with an electronic active bracelet (Afifarm version 2.03, Afikim, Israel). Milking was carried out three times daily, the cows were identified at each milking, and the milk yield was automatically matched with cow identity. The cows walked into the milking parlour over a digital scale (Merav 2001; Shekel Balances) equipped with a special algorithm for weighing walking animals, and body-weight records were automatically stored. Official ‘‘A1’’ milking measurements were carried out every 2 weeks, when the milk yield was weighed and milk was sampled for chemical analysis at each milking by a technician from the Israeli Association of Cattle Breeders. Body condition was assessed weekly on a 1 –5 scale (Edmonson et al., 1989). In both experiments, the cows had free access to feed and water at all times and were managed according to the Israel Council on Animal Care Guidelines (1994). 2.3. In vivo digestibility of safflower hay (experiment 1) In experiment 1 safflower hay was the sole food, of which the cows were provided with 13.5 kg DM, i.e., ad libitum, daily at 0900 h. The leaves and stems contributed DM at 430 F 22 and 570 F 24 g kg 1, respectively; the composition is shown in Table 1. The experiment comprised two 10-day periods. Each day during period 1, cows were given two 125 g-intra-ruminal doses of polyethylene glycol (PEG, MW 6000, Renex, ICI CC&P, Chocques, France), as indigestible marker, at 0800 and 1600 h, via the rumen fistulas. During period 2, 250-g PEG doses were mixed in 1 l of water, and these PEG solutions were then poured into 5 l of water in the drinking trough. In order to ensure that PEG was totally consumed, water was added only after animals had drunk all these 6 l. Feed intake was recorded every day and orts were collected for the last 3 days of each period, for each cow separately. Samples of hay and orts were dried at 60 jC for 3 days in a forced-ventilation oven and passed through a sieve 1-mm sieve in a cutting mill (Retsch, Germany). Grab samples of faeces were taken at 0000, 0700, 1230, and 1800 h on the last 3 days of each period; they were dried (3 days, 60 jC), and ground through a sieve 1-mm sieve. Composite

266

S. Landau et al. / Livestock Production Science 88 (2004) 263–271

samples of the orts and faeces samples were then kept frozen pending analysis. 2.4. Safflower silage as forage for dairy cows (experiment 2) Forty-two cows were initially fed with a commercial total mixed ration (TMR) formulated to provide 20.2 kg of DM daily and to contain, on a DM basis, 7.28 MJ of Net Energy for lactation, 167 g kg 1 crude protein (CP), 109 g kg 1 rumen degradable protein, 5.5 g kg 1 non-protein nitrogen, and 323 g kg 1 neutral detergent fiber (NDF) of which 51% originated from roughage. Concentrates comprised 4.5 kg of ground corn grain, 1.58 kg of ground barley, 1.66 kg of soybean meal, 0.19 kg of gluten meal, 0.40 kg of sunflower meal, and 0.53 kg of whole Pima cotton seeds. Roughage formed 4.3 kg DM and consisted of 1.65 kg wheat silage, 2.32 kg corn silage, 0.27 kg wheat straw, 1.41 kg wheaten hay, and 0.54 kg vetch hay. The rest of diet consisted of mineral and vitamin mixtures. The cows were sorted into two groups (n = 21), according to days in milk and milk yield. The ration of one group contained corn plus wheat (CW) silage throughout 62 days of experiment, whereas safflower (S) silage (see composition in Table 1) replaced the corn and wheat silage in the TMR consumed by the other group. The two diets were of similar NDF contents. The replacement of CW silage with S silage was done gradually (7 days), and the experiment began (day 0) on August 24, 3 days after completion of the substitution. Each cow’s orts were pooled within for two periods (days 7 –10 and days 37– 40 of the experiment). 2.5. Chemical analyses 2.5.1. Hay and orts analyses Feeds were assayed for DM, ash, CP, ADF, and NDF using the AOAC (1984) procedures, and for in vitro DM digestibility according to Tilley and Terry (1963). 2.5.2. Analysis of PEG in faeces In experiment 1, the PEG content of faeces was analyzed by Near Infrared Spectroscopy (NIRS), as described by Landau et al. (2002). A calibration curve

was constructed by mixing faeces and PEG in which the concentration of PEG ranged between 0 and 100 g kg 1 DM. The faeces used to make these mixtures were collected for 2 days from the experimental cows, when they were fed safflower but not PEG. The samples were pooled, and dried for 3 days at 60 jC. Before NIRS scanning, the faeces samples were placed in an oven at 50 jC for 2 h to stabilize their moisture contents, and were placed in a dessicator for 1 h to cool to ambient temperature. Spectra were taken with a Foss NIRSystems 5500 reflectance (R) apparatus (Foss Tecator, Hoganas, Sweden). NIR spectra of log (1/R) were related to PEG concentration by means of the WinISI II software (ISI, 1999). All wavelengths in the 1108– 2492 nm range were combined in increments of 4 nm, and the modified partial least square method (MPLS) method was used to establish a relationship between the first derivative of log (1/R) and the PEG concentrations of 84 faeces – PEG mixtures. The NIR spectra were corrected for particle size by the Standard Normal Variance (SNV) and de-trend procedure (Barnes et al., 1989). The coefficient of determination (R2) of calibration was 0.99, and the S.E. of cross-validation was 3.8 g kg 1 of DM. 2.5.3. Milk analyses Milk was analyzed for CP (N*6.38), fat, lactose, somatic cell count (SCC), and urea content at the laboratories of the Israeli Cattle Breeders Association (Caesarea, Israel) by Medium Infrared Spectrometry. 2.6. Statistics In experiment 1, the effects on faecal PEG concentration, of period (confounded with method of PEG dosing), time of sampling, and the period  time time of sampling interaction, were analyzed, by means of a repeated measurements procedure, with cow (period) as the term of error, before in vivo digestibility was calculated. In experiment 2, the effects of diet on group averages for milk yield, milk composition, yield of milk constituents, live-weight and body composition score were assessed by using a repeated measurement procedure of analysis of variance, with DIM as covariate and cow (treatment) as the repeated measure (SAS, 1989).

S. Landau et al. / Livestock Production Science 88 (2004) 263–271

Table 3 In vivo and in vitro DM digestibility and chemical composition (g kg 1) of components in safflower hay, when offered to dry, pregnant Holstein cows (experiment 1) and in orts

3. Results and discussion 3.1. Safflower hay for pregnant dry cows The safflower hay used in experiment 1 was harvested at the late budding stage. It was, therefore, of higher CP content (Table 1) than the full-bloom hay used by Stanford et al. (2001) (134 and 97 g kg 1 DM, respectively). However, the ADF and NDF contents were higher than in the Canadian study (298 vs. 230 g kg 1 and 429 vs. 320 g kg 1, respectively, in DM). This was probably because of higher temperatures during the growing period in Israel than in Canada, or may have been due to differences between safflower cultivars. The ruminal solubility of safflower hay DM was 518 g kg 1 (Table 2), which is higher than values reported by others for temperate perennial grass forages (Hoffman et al., 1993) or for wheat silage in Israel (Ashbell et al., 1997). The b fraction of DM from safflower hay was comparable with that of wheat silage, but the rate of degradation of this fraction was more than twice as great for safflower as for wheat. The high solubility of the DM was not an artifact that arose from high ash content, because OM also exhibited high ruminal solubility. The ruminal solubility of protein was very high, and its rate of degradation (0.21 h 1) was considerably higher than in all conventional roughages fed to dairy cattle; as a result, safflower hay exhibited very high effective degradability of DM and CP (Table 2). Cows ingested 11.5 F 0.19 kg of safflower DM hay in experiment 1. The CP and ADF content of the orts DM were 53 and 404 g kg 1 DM in the present study, compared with 44 and 330 g kg 1 in the Canadian study, which involved sheep. The in vitro DM digestibilities of ingested hay, leaves, Table 2 In situ kinetics and the effective ruminal degradability, assuming a ruminal outflow rate of 0.04 h 1 (ED0.04), of the safflower hay used in experiment 1: a, b, and c, respectively, represent the soluble, and slowly degradable fractions, and the hourly rate at which fraction b is degraded

DM OM CP

267

a

b

c

ED0.04

0.518 0.481 0.538

0.299 0.323 0.373

0.08 0.08 0.21

0.712 0.691 0.851

In vitro DMD Hay (in the trough)* Leaves Stems Orts* Ingested hay components*

CP

NDF

ADF

625 F 13 134 F 12 429 F 7 298 F 6 729 231 546 61 505 F 10 53 F 3 646 F 3 148 F 3

339 497 558 F 4 406 F 5

In vivo DMD –

225 – 353 – 404 F 5 – 279 F 3 723 F 5

* Average F S.E. of 5 days within four cows. Amounts of hay on offer and orts were 13.5 kg DM and 2.0 (S.E. 0.19) kg DM, respectively.

stems, and orts were 646, 729, 546, and 505 g kg 1, respectively (Table 3). The composition of orts (very low CP, very high ADF; Table 3) clearly shows that they consisted of stems, and a comparison of the chemical composition (NDF, ADF) of stems with that of orts shows that within stems, the coarsest parts were avoided. As a result of this selective feeding behaviour, the ingested hay components were, compared with the hay offered at the trough, richer in CP (148 vs. 134 g kg 1 DM), poorer in NDF (406 vs. 429 g kg 1 DM) and ADF (279 vs. 298 g kg 1 DM), and more digestible (646 vs. 625 g kg 1 DM). Feeding to excess, i.e., allowing for orts may, therefore, have upgraded the value of safflower hay for cattle, as it did for sheep (Stanford et al., 2001). Such selectivity at the trough, well known in goats (Morand-Fehr et al., 1991), is not commonly described in cattle. Shortening the fibers by chopping prior to feeding, as done before ensiling in experiment 2, would probably have reduced the relatively high residue (146 F 14 g kg 1 DM of the offered hay), associated with selection of the most nutritive components. The average PEG concentrations in faeces were 84 F 5 and 78 F 5 g kg 1 DM in periods 1 and 2, respectively, and these figures between these two figures did not was not significant ( P = 0.18). Faecal PEG concentrations were in a range narrow enough to justify calculations based on steady state condition (85, 79, 74, and 80 g kg 1 in DM for faeces sampled at 0000, 0700, 1230, and 1800 h, respectively). No effect of sampling time or the period x

268

S. Landau et al. / Livestock Production Science 88 (2004) 263–271

sampling time interaction was found on fecal PEG concentration but a significant effect of the individual cow was noted ( P < 0.01). This justified arithmetical averaging of PEG concentration for each cow, over all sampling times. This average was used to calculate DM digestibility in vivo. The in vivo digestibility of DM from safflower hay (Table 3) was 723 F 5 g kg 1. This figure is greater than the in vitro value (646 g kg 1) calculated from the digestibility of hay components selected by the animals. The intake of metabolizable energy from safflower hay by pregnant cows in experiment 1, calculated according to the in vitro (114 MJ day 1) or in vivo values (122 MJ day 1), on the assumption that 0.839 of digested energy is metabolizable (Vermorel, 1978), exceeded the NRC (1988) recommendation for 6-month-pregnant cows. Overall, our data for pregnant cows confirmed the previous finding based on sheep (Vonghia et al., 1992; Stanford et al., 2001), that safflower has the potential to be a valuable fodder for productive ruminants, and these findings justified the implementation of experiment 2. 3.2. Safflower silage for lactating cows Two cows in each group (n = 21) had to be culled because of mastitis, and data based on 19 cows per

group are presented. After a short decline in milk yield, cows in the S group adapted to their new diet, and milk yields were similar in S and CW at the start of experiment (32.0 and 31.5 kg, S.E.M. 1.25 kg; Fig. 1). However, the average level of orts was 0.213 of he distributed ration in S, compared with 0.150 in CW ( P < 0.001). The feed intake of S cows, which had decreased following the inclusion of safflower silage in their diets, did not recover to initial values, and remained lower ( P < 0.02) than in CW throughout the 62 days of experiment (20.2 vs. 22.5 kg DM, S.E.M. 0.4 kg; Fig. 1). The two experimental diets were formulated to include similar NDF total contents, and similar NDF contributions from roughage, and this strategy should have resulted in similar DM intakes (Mertens, 1997). There are several possible reasons for the depressed DM intake in cows of the S group. First, the nutritional values of leaves and stems differ much more in safflower than in cereal silages. Table 4 shows that despite their higher level of ort, cows fed S left residues of very low nutritional value, i.e., higher ( P < 0.001) NDF, ADF, and ash, compared with CW. This is clearly because of the avoidance of coarse stems noted also in experiment 1. Therefore, although chopping at the time of harvest and in the TMR mixing wagon probably decreased the fiber length, and consequently mitigated the selecting

Fig. 1. Individual daily feed intake (kg DM, circles) and milk (squares) in cows fed total mixed rations containing corn and wheat silage (CW, empty symbols) or safflower silage (full symbols).

S. Landau et al. / Livestock Production Science 88 (2004) 263–271 Table 4 Chemical composition of daily individual rations of TMR and orts in cows fed a total mixed ration including corn and wheat (CW) or safflower (S) silage (n = 19) DM basis (g kg

Ration S TMR Orts

1

)

CP

NDF***

ADF***

Ash***

171 F 3 154 F 1

331 F 6 451 F 2

175 F 5 326 F 2

92 F 3 128 F 6

314 F 14 363 F 4

154 F 6 196 F 3

77 F 4 84 F 2

Ration CW TMR 165 F 3 Orts 155 F 1

*** Orts from S and CW rations differ ( P < 0.001).

behaviour of cows at the trough in experiment 2, this behaviour was still very effective. Flowering of safflower is day-length dependent, and since the nutritional quality is negatively correlated with the date of sowing and the biomass accumulation, this quality can be manipulated (Leshem et al., 2001). It is anticipated that by compromising on biomass yield, e.g., by sowing at later date, the quality of stems will be improved, leading to less selective behaviour and increased intake of safflower forage by cattle. Second, it is possible that a secondary compound was present that negatively affected the DM intake in cows fed the S diet. Plants grown in Mediterranean environment must frequently cope with drought conditions, and safflower adaptation to drought has been hypothesized to be associated with polyphenol accumulation in the foliage (Yaginuma et al., 2002). No evidence was found to support this hypothesis in the present study, and Friedman (2002) found only 0.3 –0.4% of condensed tannins in rain-fed safflower grown in the same location in the Negev desert of Israel. In addition, safflower leaves were consumed well in both experiments, and the high ruminal degradability of safflower CP found in this study (Table 1) is not consistent with high polyphenol content (Aerts et al., 1999). Another possible cause of DM intake depression could be terpene. Terpenes are deterrents of insect predation on wild sunflowers (Rogers et al., 1987), and are found in many Compositae. Unfortunately, we have no data on terpene concentration in safflower, and research on secondary compounds is needed before safflower can be adopted as forage for cattle. However, from a more

269

global point of view, even though depression in DM intake is a negative attribute, natural adaptation to drought and predation can be considered as assets, especially in organic agriculture. Milk yields of the S and CW groups were similar: 30.5 and 30.0 kg day 1, respectively, according to the Afimilk procedure or 29.8 and 29.2 kg day 1, according to the measurements on milk control days (Table 5; Fig. 1). Milk from the two groups did not differ in its content of fat, urea, and lactose, or in somatic cell count. However, milk from cows in the S group tended ( P = 0.07) to contain less CP than milk from CW counterparts (31.6 and 33.6 g kg 1, S.E.M. 0.7 g kg 1). This tendency is consistent with the finding that CP in safflower is very degradable. No attempt was made to counterbalance this degradability of CP by augmenting the sources of degradable OM, which possibly resulted in suboptimal use of dietary protein, possibly leading to a reduction in milk CP. Excess degradable CP, relative to degradable OM, in the rumen, has been associated with higher fat concentration in the milk, possibly related with increased NDF digestibility (Witt et al., 2000). However, the difference in milk fat content between the S and CW groups was not significant (Table 5). Body weight increased 16.7 and 15.1 kg (S.E.M. 4.2 kg) in CW and S cows, during the 62 days of experiment, respectively, and group BWs did not differ at any time. Body condition score was not affected by dietary treatments, and averaged 2.92 and 2.76 (S.E.M. 0.12) in CW and S cows, respectively. The question arises therefore, of how cows in the S group ate less than those in the CW group, and still produced similar amounts of milk, while

Table 5 Average milk yield and contents of crude protein (CP), fat, lactose (g kg 1), urea (mg kg 1), and somatic cells (SCC, millions kg 1), in cows fed a total mixed ration including corn and wheat (CW) or safflower (S) silage (n = 19) Daily milk yield

Milk composition

Afimilk A1 control CP CW 30.0 S 30.5 S.E.M. 1.0 P 0.75

29.2 29.8 1.0 0.68

Fat

Lactose Urea

33.6 34.9 46.8 31.6 35.8 46.0 0.7 1.0 0.5 0.07 0.54 0.25

SCC

310 338 326 461 16 100 0.51 0.42

270

S. Landau et al. / Livestock Production Science 88 (2004) 263–271

apparently accreting their fat reserves to the same extent. A first possible explanation is that safflower silage modified the rumen environment conditions and changed the balances between micro-organisms present there, which resulted in improved digestion efficiency or in alteration of the composition of volatile fatty acids released in the rumen. Such an effect could arise from the high ruminal degradability and higher pH in safflower silage than in the corn – wheat silage mixture, probably because of the buffering effect of ash (Table 1). This hypothesis needs to be verified by chemical analysis of rumen fluid. A second explanation is that the ingested safflower components contained more energy than predicted by the in vitro assessment, as evidenced in experiment 1, in which DM digestibility was greater when measured in vivo than in vitro (Table 3). This is because, at the end of lactation, cows that have achieved their milk potential and enjoy positive energy balance, as did cows in the present study (205 DIM at the start of experiment), will probably not attempt to maximize DM intake, especially in the hot Israeli summer. Therefore, our data suggest that 20.2 kg and 22.5 kg DM of the S and CW rations, respectively, provided similar arrays of nutrients, sufficient for milk production and body accretion functions. This hypothesis needs to be ascertained by in vivo measurements of the digestibility of DM from the S and CW diets. We are not aware of any other attempt to feed cows with silage manufactured with a plant of the Compositae family. The first data that we present here indicate that safflower, and, possibly, other Compositae, grown under Mediterranean conditions, hold potential to serve as silage for high-producing dairy cows.

Acknowledgements We are indebted to Mr. M. Nikbahat and Mrs. H. Lehrer for skilled care of animals at Bet Dagan. Contribution from the Agricultural Research Organization, Institute of Field and Garden Crops, Bet Dagan, Israel, No. 119/01. This research was supported by Research Grant No. 257-0203 from the Chief Scientist, Ministry of Agriculture, and the Israel Dairy Board.

References Aase, J.K., Pikul Jr., J.L., 2000. Water use in a modified summer fallow system on semiarid northern Great Plains. Agric. Water Manag. 43, 345 – 357. Aerts, R.J., Barry, T.N., McNabb, W.C., 1999. Review paper: polyphenols and agriculture: beneficial effects of proanthocyanidins in forages. Agric. Ecosyst. Environ. 75, 1 – 12. AOAC, W.C., 1984. Official Methods of Analysis, 14th ed. AOAC Publications, Washington, DC. Ashbell, G., Weinberg, Z.G., Bruckental, I., Tavori, K., Sharet, N., 1997. Wheat silage: effect of cultivar and stage of maturity on yield and degradability in situ. J. Agric. Food Chem. 45, 709 – 712. Barnes, R.J., Dhanoa, M.S., Lister, S.J., 1989. Standard normal variance transformation and de-trending of near-infrared diffuse reflectance spectra. Appl. Spectros. 43, 772 – 777. Bassiri, A., Khosh-Khui, M., Rouhani, I., 1977. The influences of simulated moisture stress conditions and osmotic substrates on germination and growth of cultivated and wild safflowers. J. Agric. Sci. (Camb.) 88, 95 – 100. Edmonson, A.J., Lean, I.J., Weaver, L.D., Farver, T., Webster, G., 1989. A body condition scoring chart for Holstein dairy cows. J. Dairy Sci. 72, 68 – 78. Francois, L.E., Berstein, L., 1964. Salt tolerance of safflower. Agron. J. 56, 38 – 40. French, A.V., O’Rourke, P.K., Cameron, D.G., 1988. Beef production from forage crops in the Brigalow region of Central Queensland: 2. Winter forage crops. Trop. Grassl. 22, 85 – 90. Friedman, S., 2002. Monitoring the value of sown pastures for ruminants in the Northern Negev by NIRS-aided methodology. MSc Thesis, the Hebrew University of Jerusalem (Rehovot, Israel). Hoffman, P.C., Sievert, S.J., Shaver, R.D., Welch, D.A., Combs, D.K., 1993. In situ dry matter, protein and fiber degradation of perennial forages. J. Dairy Sci. 76, 2632 – 2642. Hume, D.E., Lyons, T.B., Hay, R.J.M., 1995. Evaluation of grasslands Puna chickory (Cichorium intybus L.) in various grass mixtures under sheep grazing. N. Z. J. Agric. Res. 38, 317 – 328. ISI, 1999. WinISI II, the Complete Software Solution for Routine Analysis, Robust Calibrations and Networking. Version 1.02A. Infrasoft International, Port Matilda PA, USA. Israel Council on Animal Care Guidelines, 1994. Legislation on Animal Welfare (defending animal rights). Paragraph 14. Kenesset Law Pub., Jerusalem, Israel. In Hebrew. Landau, S., Friedman, S., Devash, L., Mabjeesh, S.J., 2002. Polyethylene glycol, determined by near-infrared reflectance spectroscopy, as a marker of fecal output in goats. J. Agric. Food Chem. 50, 1374 – 1378. Leshem, Y., Bruckental, I., Landau, S., Ashbell, G., Weinberg, Z.G., 2001. Safflower: a promising forage crop for semi-arid regions. Proc. 19th Intl. Grassland Congress, Piracicaba (Brasil), 2 – 6 July 2001. Mertens, D.R., 1997. Creating a system for meeting the fiber requirements of dairy cows. J. Dairy Sci. 80, 1463 – 1481. Morand-Fehr, P., Owen, E., Giger-Reverdin, S., 1991. Feeding behaviour of goats at the trough. In: Morand-Fehr, P. (Ed.), Goat Nutrition, pp. 3 – 12 PUDOC, Wageningen.

S. Landau et al. / Livestock Production Science 88 (2004) 263–271 Ørskov, E.R., McDonald, I., 1979. The estimation of protein degradability in the rumen from incubation measurements weighted according to rate of passage. J. Agric. Sci. (Camb) 92, 499 – 503. NRC, 1988. Nutrient Requirements for Dairy Cattle, Sixth Revised Edition. National Academy Press, Washington, DC. SAS, 1989. SAS/STAT guide. Version 6.12. SAS Institute, Cary, NC, USA. Rogers, C.E., Gershenzon, J., Ohno, N., Mabry, T.J., Stipanovic, R.D., Kreitner, G.L., 1987. Terpenes of wild sunflowers (Helianthus): an effective mechanism against seed predation by larvae of the sunflower moth. Homoesoma electellum. Environ. Entomol. 16, 586 – 592. Stanford, K., Wallins, G.L., Lees, B.M., Mundel, H.H., 2001. Feeding value of immature safflower forage for dry ewes. Can. J. Anim. Sci. 81, 289 – 292. Sulas, L., Caredda, S., 1997. Introduzione in coltura di nuove specie foraggere: produttivita e composizione bromatologica di Chrysanthemium coronarium L. (crisantemo) sottoposto a pascolamento simulato. Introduction of a new species as fodder crop: forage production and quality of a Chrysanthemium coronarium L. population under simulating grazing. Riv. Di. Agrom. 31, 1019 – 1026 (in Italian).

271

Tilley, J.M.A., Terry, R.A., 1963. A two-stage technique for the in-vitro digestion of forage crops. J. Br. Grassl. Soc. 18, 104 – 111. Vermorel, M., 1978. Energie: utilisation energetique des produits de la digestion. In: Jarrige, R. (Ed.), Alimentation des ruminants. INRA Publications, Versailles, France, pp. 47 – 88. In French. Vonghia, G., Pinto, F., Ciruzzi, B., Montemurro, O., 1992. ‘‘In vivo’’ digestibility and nutritive value of safflower utilized as fodder crop cultivated in Southern Italy. In: Guessous, F., Kabbali, A., Narjisse, H. (Eds.), Livestock in the Mediterranean Cereal Production Systems. PUDOC, Wageningen, pp. 127 – 129. Weinberg, Z.G., Ashbell, G., Hen, Y., Leshem, Y., Landau, S., Bruckental, I., 2002. A note on ensiling safflower forage. Grass For. Sci. 57, 184 – 187. Witt, M.W., Sinclair, L.A., Wilkinson, R.G., Buttery, P.J., 2000. The effects of synchronizing the rate of dietary energy and nitrogen supply to the rumen on milk production and metabolism of ewes offered grass silage based diets. Anim. Sci. 71, 187 – 195. Yaginuma, S., Shiraishi, T., Ohya, H., Igarashi, K., 2002. Polyphenol increases in safflower and cucumber seedlings exposed to strong visible light with limited water. Biosci. Biotechol. Biochem. 66, 65 – 72.