CHEESE FLAVOR AND NATIVE PASTURE

Transcription

1 CHEESE FLAVOR AND NATIVE PASTURE P. J. Van Soest, S. Carpino* and G. Licitra* Department of Animal Science, Cornell Universitiy and CoRFiLaC, Regione Siciliana, Ragusa, Italy* INTRODUCTION Feed has long been known to influence the flavor of milk and cheese, but relatively little attention has been paid to its positive effects. This paper presents the results of a study in Sicily on flavor of Ragusano cheese as affected by grazing. (PhD thesis of Stefania Carpino, Cornell University, 2003). The implications of this study are relevant to the unique flavor of Northeastern cheddar cheeses from Northern New York and Vermont. Cheese is a major industry in Southeastern Sicily and a major dairying area of Southern Italy. This region, the Hyblean plateau, is a limestone uplift highly dissected and eroded, feet elevation, and has an ancient history. This area was settled by the Greeks before 1000 BC and by 400 BC was deforested and put under the plow. Sicily was the major granary of the Roman Empire from 200 BC to 400 AD. Eventually, in the middle ages erosion caused the loss of topsoil and exposed bedrock rendering fields untillable, so that grazing with cattle, sheep and goats became the major agricultural activity, cheese being a major product. This situation continued up to the 20 th Century. Introduction to modern American farming methods in the late 20 th century resulted in clearing fields of stone, planting of corn, ryegrass, alfalfa, and the introduction of complete TMR rations for feeding of dairy cattle. The result has been a decline in cheese quality and a drop in cheese price of 30 percent or more relative to traditionally produced cheese. EXPERIMENTAL The grazing study was conducted on a farm near Ragusa on the Hyblean upland plateau (Carpino et al.2003). Two different fields were used in winters of 1999 and Forage is available from February to May and determined by the mild Mediterranean climate with wet winters and hot dry summers. Cows were in late lactation milking 25 kg (63.5 lbs) per day and had a body condition score of Field composition was measured by wooden square meter frames placed randomly in the fields. Forage within the frame was cut, sorted into species in the laboratory. On the same day forage selected by the cows was measured by following the cows for minutes, the number of bouts of eating, bites per feeding bout, species and plant parts were noted. Forage samples were hand plucked imitating the behavior of the cows. A video camera was used in 2000 to facilitate observations. Collected samples were botanically separated and weighed. Field composition and plant selection were

2 calculated by dividing the weight of each species by the net weight collected (wet basis). Sampling was conducted every two weeks. Hyblean native pastures are very complex, comprising as many as forty species of a dozen plant families in any one pasture. Collectively, over 100 hundred species from more than 25 plant families have been identified. Because of this complex diversity, it was not possible to chemically analyze all species. Instead 14 species were chosen from a preliminary survey. These are shown in Table 1. Table 1. Species analysed for flavor compounds, parts per 1000 Family Genus Year 1999 Year 2000 Obs.-Eaten Obs.-Eaten Asteraceae Anthemis Calendula Cerinthe Chenopodiaceae Beta Cruciferareae Diplotaxis Sinapis Euforbiaceae Euphorbia Fabaceae Medicago Scorpiurus 5-3 Geraniaceae Erodium Geranium 1-1 Malvaceae Malva Not observed Fumaraceae Fumaria Not observed Not observed Polygonaceae Rumex

3 Values for observed and eaten forage are given on a proportional basis of the total weights of forage observed and eaten in parts per thousand. The fields in years 1999 and 2000 have different compositions. Comparison of portions eaten relative to observed indicate relative selection or rejection by the cows. Spice girl Diplotaxis is an interesting case eaten in 1999, but somewhat avoided in This was because of one heifer called the Spice girl who ate 20% of her forage intake of this species in 1999 which has a fiery hot taste. Other cows tended to avoid it. This cow was not used in the year Another interesting species is Euphorbia heliscopia, known to contain poisonous alkaloids. Yet it was eaten consistently in small amounts. Cows were observed to search out particularly at the end of grazing, perhaps for its medicinal value. This species contained most of the compounds identified in the pasture cheese. Species eaten but not analyzed are shown in Tables 2 and 3. Table 2: Species not chemically analyzed for flavor compounds Family Genus Year 1999 Year 2000 Obs.-Eaten Obs.-Eaten Asteraceae Centaurea Chicorium Hedypnois Not observed 8-10 Hyoseris Senecio Sonchus Unspecified Caryophyllaceae Silene 9-25 Not observed Poaceae Unspecified Primulaceae Anagallis Ranunculaceae Adonis Umbelliferae Unspecified

4 Table 3: Minor and avoided species (unanalyzed) for flavor compounds Family Genus Year 1999 Year 2000 Obs.-Eaten Obs.-Eaten Asteraceae Carduus Chrisanthemum Capsella Not observed Cruciferae Biscutella Raphanus Not observed Liliaceae Unspecified Papaveraceae Unspecified Not observed Many of these have low occurrences but are collectively significant. Carduus (Thistle) occurred in larger portions but was highly avoided. Of the total species observed about 60% were of the species selected for analysis (Table 4). Table 4: Total species collected parts per 1000 Total species Year 1999 Year 2000 Obs.-Eaten Obs.-Eaten Analyzed Unanalyzed Total identified Unidentified Over 90% of plants were botanically identified. Cows tended to eat what was available as indicated by the high correlations between observed and selected (Table 5). Table 5: Correlations between occurrence and selection Year n Linear Logarithmic All species Combined Eaten species (omitting Carduus and Biscutella) Combined

5 Occurrence and selected values are highly skewed toward a logarithmic distribution. However, this did not disturb statistics. There was no interaction between the years. Quality remained high over the period February to May, but species distribution changed as a succession of species went through their flowering stages. Average composition is given in Table 6. Table 6: Composition of pasture selected by cows (g/kg DM or as stated) Composition Year 1999 Year 2000 Mean S.D. Mean S.D. DM (g/kg) Crude protein NDF AD lignin Ash Dig.DM Cheese preparation and analysis Milks from the two groups of cows were made into Ragusano cheese every 15 days at four intervals. Two blocks from each vat resulting in eight blocks for each type of milk. A block weighed about 14kg and 16 blocks were made. Half were aged 4 months and half 7 months. Cheese was cut into cubes for sensory and chemical analysis. Chemical analyses for flavor compounds were conducted on the selected forages and respective cheeses. Volatiles from the forages were isolated by steam distillation (Figure 1). Volatiles from the cheeses were extracted using both steam distillation and solid-phase microextraction (SPME) apparatus. SPME is an extraction using a 1 to 2 centimeter of fiber coated with a stationary phase that extracts volatiles from liquids and headspace gases. Gas chromatography olfactometry (GCO) was conducted on extracts from plant material and cheeses using charm analysis (Acree et al, 1976, 1984). These methods were developed at the Geneva Experiment Station, Cornell University for identifying volatile flavor compounds in food and wine. GCO involves a technician (sniffer) trained in recognizing odors according to a standard set of descriptive terms. Samples injected into the gas chromatograph and the sniffer detects the respective odors and their retention times. The samples are separately chromatographed in combination with a mass spectrometer.

6 Figure 1. Steam distillation apparatus Results of cheese analyses Sixteen flavor compounds in were found in pasture-derived cheese that were not present in TMR-derived cheese (Tables 7 and 8). Of these ten were identified in pasture plants (Table 7) and compounds not found in the pasture plants in Table 8. Table 7: Compounds in pasture-derived cheeses found in analyzed plants that were eaten (Carpino et al. 2004b). Odor Compound Plants Peach Decalactone 13/13 (all species) Orange Methionol 13/13 (all species) Green hay Nonenal 12/13 Rose Citrunellol 11/13 Vanilla Vanillin 8/13 Daisy Phenylacetaldehyde 5/13 Herbaceous Carvone 4/13 (Eup. Scor. Erod. Ger.) Green flower Octadienal 2/13 (Eup. Calendula) Sweet Et Me butyrate 2/13 (Eup. Rum.) Fresh flower Geranyl acetate 1/13 (Anthemis) These may have been missed in the plant analysis, or have come from plants not analyzed. Also they could be metabolites of plant compounds in the rumen fermentation or from the cheese.

7 Table 8: Compounds in pasture cheeses not found in analyzed plants species Odor Compound Possible sources Jasmin Me jasmonate Overlooked species Nutty di Me pyrazine Overlooked species Onion di Me sulfide Overlooked or metabolic Fried Decadienal Overlooked or metabolic Lilly Dodecanal Overlooked or metabolic Buttery Hydroxybutanone Metabolic product Hot milk Nonanone Metabolic product In contrast, only a few compounds were found in TMR-derived cheese that were absent in pasture-derived cheese (Table 9). Their aromas; viz. sweaty, woody, body odor, contrast with those from pasture-derived cheese that were more fragrant. Twenty seven compounds were identified in the pasture-derived cheese, whereas only 13 were identified in the TMR-derived cheese. Nine compounds were present in both cheeses. These included acetic acid, butyric acid, ethyl butyrate, ethyl hexanoate, nonanal, 1 octen 3 one, methional, dimethyl disulfide and dimethyl tetrasulfide. There were also two unknowns. Table 9: Compounds detected in TMR-derived cheeses but not in pasture-derived cheeses Odor Compound Possible sources Sweaty Hexanoic acid Silage? Woody Hexadecane Maize Body odor 8 nonenone Metabolite? Metallic Unknown From these identifications it is clear that complex terpenoids, aromatics, complex ketones and aldehydes pass from feed through the rumen, the animal and pass into milk and into cheese. They probably pass through the lipid phase. In addition the pasture-derived cheeses were yellower and higher in CLA acids. They had a different texture less liable to fracture. The cheeses were evaluated by a trained panel of 11 persons who had to wear special colored spectacles to mask the yellow carotenoid color. The panel easily distinguished the respective cheeses (Carpino et al. 2004a). The amount of grazed forage needed to elicit these responses may not be large. The daily intake of TMR averaged kg (54.7 lb) and the pastured group kg (47.7 lb). Balancing for energy, the calculated intake of pasture forage was 3.07 kg (7.8 lb) per day which is about 14% of the daily feed intake. To make an example: The species Anthemis arvensis was the only observed source of geranyl acetate (Table 7) and was consumed 1.7 to 5.1% of the net forage eaten. This implies that 50 to 150 grams of dry matter intake per day might influence flavor. However, geranyl acetate may have occurred from the unanalyzed species.

8 IMPLICATIONS This is the first study on diet and cheese that identifies compounds present in certain forage species that pass on to cheese. Other studies in Switzerland have also identified grazing as a factor in cheese flavor as reviewed by Carpino et al. 2004b. It follows that in other places grazing will also affect cheese quality. The most relevant area to us is the cheddar cheese in Northern New York and Vermont that is known to produce unique cheddar cheese that cannot be produced in other parts of the United States. A few years ago milk was brought from Wisconsin to a cheese plant in Northern New York. Quality cheddar could not be made from it. The Wisconsin milk probably came from TMR fed dairies. However, this may not be the whole story. Fields in the north of New York and Vermont are notoriously difficult to maintain pure forage stands. Weeds invade fields particularly the composites (Asteraceae) which were also the most abundant dicot in the Hyblean fields in Sicily. The quality of the northeastern cheddars depends on maintaining pasture diversity. The silage or haylage made from respective fields may very well contain compounds that influence cheese flavor in winter months. Reports from various places in the world reporting that cheese came from ryegrass (Netherlands, England or Scotland) or white clover (New Zealand) and assuming grazing never report the weeds that may be in their fields. Cows deprived of variety in their diet likely search out such plants, and as a result the particular flavor of various cheeses are not adequately described. Weeds are a very much overlooked source of forage. Ironically many of them are higher in nutritive value than the crops they are grown with. These are inconvenient plants, out of place, difficult to manage and likely of lower yields (Van Soest, 1994). Cheese manufacturers do not like to admit that they are unable to make quality cheese from just any milk, yet the northern New York and Vermont observations belie this. This study offers a reason for why the problem in cheese quality exists and that they are beyond the control of the cheese plant manager. The uniqueness of many of specialized cheeses may depend in part on the particular botany of the regional pastures they are from. CONCLUSIONS Plant species often regarded as weeds in pastures have been identified as influencing flavor of cheese. These are dicots dominantly from the composites. Other nongrass, non legume species are also important. Flavor compounds in these species have been identified in cheese. Cheese flavor depends on diet diversity and the monoculture of TMR with corn silage, alfalfa promotes a decline in quality. Cows when grazing are able to select for unusual plants. It is possible that small intakes of these species will affect flavor. The quality of northeast cheddars is dependent on maintenance of traditional grazing and feeding systems. If the economic pressure for TMR feeding management continues in the northeast, means will have to be found to identify native weeds species responsible for flavor in

9 northeastern cheddars and find ways to introduce them into the feeding system. Otherwise the industry will be marginalized or die. REFERENCES Acree, T.E., J. Barnard, and D.G. Cunningham A procedure for the sensory analysis of gas effluemts. Food Chem. 14: Acree, T.E., R.M. Butts, R.R. Nelson, and C.Y. Lee Sniffer to determine the odor of gas chromatographic effluents. Analyt. Chem. 48: Carpino, S Selective grazing on Sicilian pasture by dairy cattle and effects on Ragusano cheese. Ph.D. dissertation, Cornell University, Ithaca, NY. Carpino, S., J. Horne, C. Melilli, G. Licitra, D.M Barbano, and P.J. Van Soest. 2004a. Contribution of native pasture to the sensory properties of Ragusano cheese. J. Dairy Sci. 87: Carpino, S., G. Licitra, and P.J. Van Soest Selection of forage by dairy cattle on complex Sicilian pasture. Anim. Feed Sci. Tech. 105: Carpino, S., S. Mallia, S. La Terra, C. Melilli, G. Licitra, T.E. Acree, D.M. Barbano, and P.J. Van Soest. 2004b. Composition and aroma compounds of Ragusano cheese: Native Pasture and Total Mixed Rations. J. Dairy Sci. 87: Van Soest, P.J., Nutritional Ecology of the Ruminant, second ed. Cornell University Press, Ithaca, N.Y.