The present invention relates to a new composition for a coating or packaging for food and food ingredients, in particular cheese and sausage and agricultural products such as seeds, nuts and bulbs, and the use of such a composition. During ripening, cheeses and sausages are coated to prevent physical damage, to reduce loss of moisture, to maintain a desired shape or to protect against undesired fungi including yeasts and bacteria. Hard, semi-hard or soft cheeses and various sausages such as matured sausages like salami or chorizo can be protected in this way by a coating. The coating generally allows natural ripening of the food, implying that it allows respiration. The coating consists in general of a plastic or polymer coating (for example based on polyvinyl acetate) that can be applied in a single layer or multiple layers. For some types of food, the coating needs to be flexible because of volume changes of the food during ripening. A skilled person will be able to select a proper polymer, depending on the type of food to be coated. An outer layer of wax, generally on basis of paraffin, can be applied on the polymer coating to block evaporative water loss and to block further ripening. According to the present invention a, preferably plasticiser free, coating dispersion is provided that has significantly low water permeability without fully blocking water evaporation. The coating results in higher production yields through the direct effect of reduced evaporative weight loss. Moreover the low water permeability allows that the food can be stored under conditions of lower relative humidity. These conditions reduce the chance of growth of yeasts and moulds on the outside of the cheese while still keeping a satisfactory amount of water in the food. Growth of yeast and moulds is very unfavourable from health point of view as well as from possible production losses. The present invention relates to a coating, which is obtained by coating onto a food or agricultural product, a water-borne dispersion comprising a copolymer made of vinyl acetate and one or more vinyl esters of a carboxylic acid with 5 to 22 carbon atoms, preferably an ester of a carboxylic acid with more than 8 carbon atoms, most preferably with more than 10 carbon atoms, and/or one or more acrylic acid esters with an alkyl ester with 5 to 15 carbon atoms, and/or one or more maleic and/or fumaric acid esters with one or two alkyl esters with 5 to 22, preferably 5 to 15 carbon atoms. We have found that maleic acid and fumaric acid having 3 or 4 carbon atoms are not suitable for the composition of the present invention due to their high vapour transmission. Preferably the carboxylic acid is a branched carboxylic acid. The alkyl ester is preferably a branched alkyl ester. By water-borne is meant a dispersion of polymer (and optionally other ingredients) in water. Preferably less than 3% w/w, more preferably less than 1% w/w and most preferably less than 0.1 w/w% of solvents other than the water, are present (% is based on the total solvent content). Preferably the water-borne dispersion comprises a copolymer of vinylacetate and one or more vinyl esters of a carboxylic acid with 5 to 22 carbon atoms, preferably an ester of a carboxylic acid with more than 8 carbon atoms, most preferably with more than 10 carbon atoms. The amount of vinyl acetate in the copolymer may vary, depending on the type of co-monomer(s) used, depending on the desired water vapour permeability, and depending on the desired polymer flexibility, but is preferably less than 85% and is preferably more than 50% (on weight basis). The described co-monomer esters with carbon groups of 5 to 22 carbon atoms are advantageous for providing the reduced moisture barrier as compared with prior art coatings for respiring food, without fully blocking water evaporation. Although the use of other co-monomers is possible, in general this is not needed. The dispersion can be obtained by co-polymerising the described monomers in an emulsion, using for example surfactant to emulsify the monomer droplets in water, and using catalyst or initiator to initiate the polymerisation reaction. To prevent sagging of the polymer particles, the dispersion can be stabilised electrostatically or by the use of stabilising colloids, for example one polyvinyl alcohol and / or cellulose derivatives. Preferably, the dispersion is stabilised by cellulose derivatives, since their use enhances moisture barrier properties as compared with polyvinyl alcohol. Cellulose is a carbohydrate that contains many alcohol groups, of which many of the derivatives of alcohols can be formed, such as esters or ethers. Examples of derivatives, are carboxymethyl cellulose, methyl ethyl cellulose, hydroxypropyl cellulose, hydroxyethyl cellulose or methyl hydroxyethyl cellulose, see for example "Handbook of Hydrocolloids", Editors G. O. Philips and P.A. Williams, Woodhead Publishing Limited (Abington Hall, Cambridge, UK), 2000. ISBN 1 85573 501 6. Preferably 0.1 to 5wt%, more preferably 0.3 to 3% cellulose or cellulose derivative is used. Electrostatically stabilized can be done by electrostatic repulsion (Young R.G. and Lovell P.A., 1991 , Introduction of polymers, second edition pages 65 - 68 Chapman & Hall). Co-polymerisation is a process known in the art and the skilled person will easily select the proper conditions and if necessary the proper additions to obtain the desired copolymer. See for example G. Challa, Polymer Chemistry: An Introduction, Ellis Horwood Limited, Herts, UK (1993), ISBN 0-13-489691-2, pages 98-101. Further ingredients for the coating according to the present invention are essentially similar to all the ingredients which are applied for prior art coatings including dispersions of polymers such as copolymers of vinyl acetate- dibutylmaleate. A person skilled in the art will have no difficulty in selecting these further ingredients. When applied on food, agricultural product and food or agricultural ingredients, the coating has a water vapour permeability as described in Example 1 of less than 20 mg/cm224h, more preferably less than 15 mg/cm224h, even more preferably less than 10 mg/cm224h, and most preferably less than 6 mg/cm224h. Preferably the water vapour permeability is more than 0.5 mg/cm224h, more preferably more than 2 mg/cm2 24h. The water vapour permeability of the coating is measured as described in Example 1 The coating can be applied in one or multiple layers. A multi- layered coating can be composed of two or more layers of coating whereby one or more layers of coating according to the present invention is used, possibly combined with one or more layers of coatings of different composition. A similar one- or multiple-layered coating can be applied on a different substrate, and used as a packaging material for food and food ingredients. In general the thickness of the coating layer of the present invention will be between 0.03 and 0.3 mm. Preferably the thickness will be less than 0.15mm. The present invention also relates to food and food ingredients coated with the coating as defined above. In general, the present invention relates to a coating for water containing products, preferably for food products or product of agricultural origine that are respiring, implying that these products respire gasses like CO2 due to natural ripening and / or water vapour that evaporates from the food. Respiration depends on the speed of ripening, which is related to the temperature, moisture content and acidity. Respiration also depends on the use of ripening bacteria. Respiration of water is related to tie water contents of the food, the shape of the food or agricultural product (surface area related to volume), and the relative humidity, temperature and air circulation of the storage facilities. Although the coating reduces moisture evaporation of the food, the ripening process is usually not negatively influenced. The present coating has significantly low water permeability without fully blocking water evaporation, therefore allowing respiration, and consequently allowing natural ripening when coated on ripening food or food ingredient. Application can be manual (e.g. with sponge or brush), semi-automatic or automatic (e.g. by brushing, dipping or spraying) depending on the number of foodstuffs to be treated. Specialised cheese coating and sausage coating machines are commercially available. Application can be done according to a procedure similar to the coating protocol for commercially available coating products such as Plasticoat®, Superdex®, Delvocoat® or Premicoat®. The skilled person in the art will appreciate to select the proper number / frequency of coating treatments. Machine settings can be modified in an empirical way in order to produce a specific end result in terms of coating application. In general coating viscosity is regulated and the coating is applied in such way that uniform coating takes place. In order to allow a regular and even distribution of the coating, it is required to individually fix treatment schemes for each kind of food.

EXAMPLES

Example 1 In the following experiment the reduced water vapour permeability of the water- borne dispersions comprising a copolymer made of vinyl acetate and a mixture of vinyls esters of branched carboxylic acids with 11 carbon atoms (Veovo 11™ obtained from Resolution Performance Products), was demonstrated. The dispersion was stabilised with hydroxy ethyl cellulose. Polymer dispersion comprising a copolymer of vinyl acetate and di-(n-butyl)maleate, Mowilith SDM 4230 KL acquired from Celanese (Germany), was used as a reference material. Coating films were prepared in such a way that the dried films have the same thickness of 0.2 mm. The films were attached on top of a holder that contained water. The holders were conditioned at 20"C and 60% RH. Water vapour permeability was determined by measuring weight loss caused by evaporation of water through a total coating surface of 12cm2, and is expressed in mg/cm224h. Water vapour permeability was determined after 24 hours and after 48 or 72 hours, all giving the same result in mg/cm^h. The results are given in Table 1. Table 1 : water vapour permeability [mg/cm224h]

Coatings made of water-borne dispersions comprising a copolymer substantially made of vinyl acetate and one or more vinyl esters of carboxylic acids with 2 to 18 carbon atoms, showed a significant reduction in water vapour permeability.

Example 2 In the following experiment the positive effect of coatings of the present invention on water loss reduction of food is demonstrated on two types of cheese. Test cheese coating was prepared based on a polymer dispersion of copolymer made of 70% vinyl acetate and a mixture of vinyl esters of branched carboxylic acids with 11 carbon atoms, stabilised with hydroxy ethyl cellulose Reference cheesecoating was prepared based on a polymer dispersion of a copolymer of vinyl acetate and di-(n-butyl)maleate, Mowilith SDM 4230 KL acquired from Celanese (Germany). Semi-hard Maasdammer-type 12 kg eyehole cheeses and semi-hard Gouda-type 12 kg cheeses were treated manually with cheesecoating from brining (day 4) until the age of 25 days. They were treated every 2 to 3 days from day 4 until day 14 and every 3 to 4 days from day 14 until day 25, resulting in eight times single sided treatment (4 times all around). /Ml coatings were applied at same thickness, which results from the identical coating process and using the same coating viscosity (Brookfield viscosity at 23eC, 20rpm, of 3500 - 45000 mPas). Cheeses were stored under regular warehousing conditions for these types of cheese. Loss of moisture was determined as the weight loss as a percentage of total weight.

The results are given in Table 2, using the following sample coding. A1 : 8 treatments with reference coating, 11 g coating per treatment. A2: 8 treatments with reference coating, 20 g coating per treatment. A3: 8 treatments with reference coating, 15 g coating per treatment. B1 : 8 treatments with test coating, 11 g coating per treatment. B2: 8 treatments with test coating, 20 g coating per treatment. C3: 4 treatments with test coating, 4 treatments with reference coating, 15 g coating per treatment.

Tests were performed on semi-hard Gouda-type 12 kg cheese, from two cheese factories (10 cheeses per series, series numbers 1 and 2) or on semi-hard Maasdammer-type 12 kg eyehole cheese (5 cheeses per series, series number 3).

Table 2: weiht loss %

Example 3 In the following experiment the positive effect of coatings of the present invention on water loss reduction of food is demonstrated on semi-hard Gouda-type 12 kg cheese. Test cheese coating was prepared based on a polymer dispersion of copolymer made of 60% vinyl acetate and a mixture of vinyl esters of branched carboxylic acids with 11 carbon atoms, stabilised with hydrolxy ethyl cellulose. Reference cheesecoating was prepared based on a polymer dispersion of a copolymer of vinyl acetate and di-(n- butyl)maleate, Mowilith SDM 4230 KL acquired from Celanese (Germany). Semi-hard Gouda-type 12 kg cheeses from 2 cheese factories (A and B) were treated manually with 15 gram cheesecoating per side, per treatment, from brining (day 4) until the age of 29 days. They were treated every 2 to 3 days from day 4 until day 14 and every 3 to 4 days from day 14 until day 25. This resulted in four times single sided treatment (2 times all around) until day 14, and five times single sided treatment from day 14 until day 29. Series of 10 cheeses were either coated with test coating or reference coating, or a combination of the two, as can be seen in Table 3. All coatings were applied at same thickness, which results from the identical coating process and using the same coating viscosity (Brookfield viscosity at 23eC, 20rpm, of 3500 - 45000 mPas). Cheeses were stored under regular warehousing conditions for these types of cheese (12 - 17"C, 78 - 88% RH). Loss of moisture was determined as the weight loss as a percentage of total weight. Results are given in Table 3.

Example 4 In the following experiment the positive effect of coatings of the present invention on water loss reduction of food is demonstrated on semi-hard Gouda-type 12 kg cheese. Test cheese coating was prepared based on a polymer dispersion of copolymer made of 62,5% vinyl acetate and a mixture of vinyl esters of branched carboxylic acids with 11 carbon atoms, stabilised with hydroxy ethyl cellulose. Reference cheesecoating was prepared based on a polymer dispersion of a copolymer of vinyl acetate and di-(n- butyl)maleate, Mowilith SDM 4230 KL acquired from Celanese (Germany). Semi-hard Gouda-type 12 kg cheeses were treated manually with cheesecoating from brining (day 4) until the age of 19 or 26 days. They were treated every 2 to 3 days from day 4 until day 14 and every 3 to 4 days from day 14 until day 19 or 26, resulting in six or eight times single sided treatment, respectively (3 or 4 times all around). Series of 5 cheeses were either coated with test coating or with reference coating. All coatings were applied at same thickness, which results from the identical coating process and using the same coating viscosity (Brookfield viscosity at 23QC, 20rpm, of 3500 - 45000 mPas). Cheeses were stored under regular warehousing conditions for these types of cheese. Loss of moisture was determined as the weight loss as a percentage of total weight. Results are given in Table 4.

Table 4: weight loss [%]

Example 5 In the following experiment the positive effect of coatings of the present invention on water loss reduction of food is demonstrated on semi-hard Gouda-type 12 kg cheese. Test cheese coating was prepared based on a polymer dispersion of copolymer made of 62,5% vinyl acetate and a mixture of vinyl esters of branched carboxylic acids with 11 carbon atoms, stabilised with hydroxy ethel cellulose. Reference cheesecoating was prepared based on a polymer dispersion of a copolymer of vinyl acetate and di-(n- butyl)maleate, Mowilith SDM 4230 KL acquired from Celanese (Germany). Semi-hard Gouda-type 12 kg cheeses were treated manually with cheesecoating from brining (day 4) until the age of 27 days. They were treated every 2 to 3 days from day 6 until day 14 and every 4 days from day 14 until day 27, resulting in six times single sided treatment, respectively (3 times all around). Series of 4 cheeses were either coated with test coating or with reference coating, a combination of the two or a mixture of the two, as can be seen in Table 5. Al coatings were applied at same thickness, which results from the identical coating process and using the same coating viscosity (Brookfield viscosity at 23SC, 20rpm, of 3500 - 45000 mPas). Cheeses were stored under regular warehousing conditions for these types of cheese. Loss of moisture was determined as the weight loss as a percentage of total weight. Results are given in Table 5. Table 5: weight loss [%]

Example 6 In the following experiment the positive effect of coatings of the present invention on water loss reduction of food is demonstrated on semi-hard Gouda-type 12 kg cheese. Test cheese coating was prepared based on a polymer dispersion of copolymer made of 62,5% vinyl acetate and a mixture of vinyl esters of branched carboxylic acids with 11 carbon atoms, stabilised with hydroxy ethyl cellulose. Reference cheesecoating was prepared based on a polymer dispersion of a copolymer of vinyl acetate and di-(n- butyl)maleate, Mowilith SDM 4230 KL acquired from Celanese (Germany). Semi-hard Gouda-type 12 kg cheeses were treated by machine with cheesecoating from brining (day 4) until the age of 27 days. They were treated every 2 to 3 days from day 4 until day 14, resulting in 4 times single sided treatment (2 times all around), and every 3 to 4 days from day 14 until day 28. Six series of 200 to 300 cheeses were either coated with test coating or with reference coating until day 14, and were treated with reference coating after 14 days. Test and reference series were from the same cheese production batch. All coatings were applied at same thickness, which results from the identical coating process and using the same coating viscosity (Brookfield viscosity at 23eC, 20rpm, of 3500 - 45000 mPas). Cheeses were stored under regular warehousing conditions for these types of cheese. Loss of moisture at day 15 was determined as the weight loss during brining and ripening as a percentage of total weight. Weight loss during brining is approximately 2.5%. Loss of moisture between days 15 and 28 was determined in relation to the standard norm friat was indexed at 100. This is the lowest achievable weight loss as determined by experience, which is related to storage conditions (temperature and humidity) and treatment frequency. Results are given in Table 6.

Table 6: moisture loss

By using this coating of the present invention until day 14, the weight loss between days 14 and 28 was 0.7% less (absolute) than when using the reference coating.

Example 7 In the following experiment the positive effect of coatings of the present invention on water loss reduction of food is demonstrated on semi-hard Gouda-type 12 kg cheese. Test cheese coating was prepared based on a polymer dispersion of copolymer made of 62,5% vinyl acetate and a mixture of vinyl esters of branched carboxylic acids with 11 carbon atoms, stabilised with hydroxy ethyl cellulose derivatives and surfactants. Reference cheesecoating was prepared based on a polymer dispersion of a copolymer of vinyl acetate and di-(n-butyl)maleate, Mowilith SDM 4230 KL acquired from Celanese (Germany). Semi-hard Gouda-type 12 kg cheeses were treated by machine with cheesecoating from brining (day 4) until the age of 25 days and longer. They were treated every 2 to 3 days from day 4 intil day 14, resulting in 4 times single sided treatment (2 times all around), and every 3 to 4 days from day 14 until day 24 or longer. Series of 300 to 1600 cheeses were either coated with test coating or with reference coating until day 14, and were treated with reference coating after 14 days. Test and reference series were from the same cheese production batch. All coatings were applied at same thickness, which results from the identical coating process and using the same coating viscosity (Brookfield viscosity at 23eC, 20rpm, of 3500 - 45000 mPas). Cheeses were stored under regular warehousing conditions for these types of cheese, and one test series and reference series (series B) were stored at lower relative humidity (70 - 75 % RH instead of 80 - 88%). Loss of moisture was determined as the weight loss during ripening from day 15 until day 25 as a percentage of total weight. And for one test series and reference series (series C) weight loss was determined from day 15 until days 43 or 38, respectively. Loss of moisture after day 15 was also determined in relation to the standard norm that was indexed at 100. This is the lowest achievable weight loss as determined by experience, which is related to storage conditions (temperature and humidity) and treatment frequency. Results are given in Table 7.

Table 7: moisture loss

By using this coating of the present invention until day 14, the weight loss between days 15 and 25 was 0.7% less (absolute) than when using the reference coating.

Example 8 In the following experiment the positive effect of coatings of the present invention on water loss reduction of food is demonstrated on semi-hard Gouda-type 12 kg cheese. Test cheese coating was prepared based on a polymer dispersion of copolymer made of 65% vinyl acetate and a mixture of vinyl esters of branched carboxylic acids with 11 carbon atoms, stabilised with hydroxy ethyl cellulose. Reference cheesecoating was prepared based on a polymer dispersion of a copolymer of vinyl acetate and di-(n- butyl)maleate, Mowilith SDM 4230 KL acquired from Celanese (Germany). Semi-hard Gouda-type 12 kg cheeses were treated manually with cheesecoating from brining (day 4) until the age of 25 days. They were treated every 2 to 3 days from day 4 until day 14, resulting in 4 times single sided treatment (2 times all around), with either test coating or with reference coating (8 cheeses per series). From day 14 until day 25 they were treated every 3 to 4 days with reference coating. Al coatings were applied at same thickness, which results from the identical coating process and using the same coating viscosity (Brookfield viscosity at 23SC, 20rpm, of 3500 - 45000 mPas). Cheeses were stored under regular warehousing conditions for these types of cheese. Loss of moisture was determined as the weight loss during ripening from day 4 until day 15 and from day 15 until day 25, as a percentage of total weight. Results are given in Table 8.

Table 8: weight loss [%]

By using this coating of the present invention until day 14, the weight loss between days 4 and 25 was 0.7% less (absolute) than when using the reference coating.

Example 9 In the following experiment the positive effect of coatings of the present invention on water loss reduction of food is demonstrated on semi-hard Gouda-type 12 kg cheese. Test cheese coating was prepared based on a polymer dispersion of copolymer made of 65% vinyl acetate and a mixture of vinyl esters of branched carboxylic acids with 11 carbon atoms, stabilised with hydroxy ethyl cellulose. Reference cheesecoating was prepared based on a polymer dispersion of a copolymer of vinyl acetate and di-(n- butyl)maleate, Mowilith SDM 4230 KL acquired from Celanese (Germany). Semi-hard Gouda-type 12 kg cheeses were treated by machine with cheesecoating from brining (day 4) until the age of 26 days. They were treated every 2 to 3 days from day 4 until day 14, resulting in 4 times single sided treatment (2 times all around), with either test coating or with reference coating (400 - 500 cheeses per series). From day 14 until day 25 they were treated every 3 to 4 days with reference coating. Test and reference series were from the same cheese production batch. All coatings were applied at same thickness, which results from the identical coating process and using the same coating viscosity (Brookfield viscosity at 23eC, 20rpm, of 3500 -45000 mPas). Cheeses were stored under regular warehousing conditions for these types of cheese. Loss of moisture was determined as the weight loss during ripening from day 15 until day 26, as a percentage of total weight. Results are given in Table 9.

Table 9: weight loss [%]

Example 10 In the following experiment the increase in water vapour permeability when using polyvinyl alcohol stabiliser in the water-borne dispersions, was demonstrated. A water- borne dispersion was used, comprising a copolymer made of 65% vinyl acetate and a mixture of vinyl esters of branched carboxylic acids with 11 carbon atoms (Veovo 11™ , obtained from Resolution Performance Products), stabilised with hydroxy ethyl cellulose. To this dispersion 1%wt of polyvinyl alcohol was added, with a degree of hydrolysis of 87,7% (Mowiol 40-88 obtained from Clariant). Coating films were prepared and water vapour permeability was determined as described in Example 1. The results are given in Table 10.

Table 10: water vaour ermeabilit m/cm224h