In the first year, four farms were involved in producing a trench silo of biofertilized whole-plant maize, in a quantity considered sufficient to feed milking Holstein Italian cows over a period of almost two months in the second year. Four farms in the Po valley, in the Piedmont Region (Carmagnola), and a farm in Soncino in the Cremona Region were involved in the ProLacte Project. MICOSAT F wp® seed for tanning cereals was used for biofertilization purposes at a dose of 1 kg/ ha. This is a somewhat complex consortium in which 100 g contains symbiotic fungi (32 g of sorghum root inoculum, spores and hypha of the Glomus coronatus GO01 and GU53 species, Glomus caledonium GM24, Glomus intraradices GB67 and GG32, Glomus mosseae GP11 and GC11, and Glomus viscosum GC41), bacteria (Bacillus subtilis BR48), saprophytic fungi (Beauveria spp. BB48, Trichoderma harzianum TH01, Trichoderma atroviride TA28) and inert media ama 100.

Details of the methods used for cropping and monitoring the growth and yield were published in a study by Masoero and Giovannetti 22, which focussed on the variations of the in-vivo stem pH at a cereus status. Each plant was cut at 2/3 of the internode below the ear, and a measurement was made of the in vivo stem pH at the up position (pH_up). A second cut was then made at ½ of the 2^nd internode, and the stem pH was measured at the bottom position (pH_bot). The pH measurements were conducted using a BORMAC XSpH 70 pH meter, provided with a combined plastic-glass electrode Hamilton PEEK Double-Pore F, with 35×6 mm dimensions (LxØ), and two decimals. All the on-plant measurements were conducted at the center of the cut sections. One leaf was retained from each plant and the thus obtained 1278 leaves were then examined in a laboratory, by means of vibrational spectroscopy, using a portable LabSpec 4 Standard-Res Lab UV-Vis-NIR Analyzer fiber optic diode array spectrophotometer (ASD, Analytical Spectral Device Inc., Boulder, CO) to scan the lower pages of the leaves over a 350-1025 nm range (676 absorbance points).

The spectra were imported in a format that was compatible with the WinISI II v1.04 software (FOSS NIRSystem / Tecator, Infrasoft International, LLC) for chemometric processing, by means of a Partial Least Squares (PLS) method, to calibrate the binary condition C vs. B, as the 1,2 dummies. No leaf quality prediction model was available at that time for the ASD instrument. Therefore, in order to consider the modifications of the biofertilizer on the leaves and to extend and validate the features of the present study, we included a collection of the NIR spectra of 1627 leaves obtained from a series of later carried out biofertilization published experiments 23.

A low cost NIR device was used for this purpose; the SCiO^{TM 23} and the spectra were downloaded, imported in WinISI and the quality traits were predicted using models validated in a previous sorghum study 25 and further improved in a vine study 26.

Analyses were carried out on chopped green whole maize plants, by means of a NIRSystem 5000, which is currently used by the Regional Breeder Association.

In the first year, one farm changed its crop organisation from intensive maize feeding and dairy production to meadow feeding and hay -haylage production, using the aforementioned biofertilizer over the entire crop area. The meadows included Trifolium pratensis, Medicago sativa, Dactilys glomerata, Lolium multiflorum, Festulolium and Festuca pratense. No comparative analyses were carried out on the grasses or hay.

In the transition period, the ration of the milking was gradually enlarged for the forage-to-concentrate ratio.

Several kinds of measurements involved the controls on milk. Individual dairy surveys were conducted on the milking cow farms by means of a functional control conducted by the Italian Herd Breeder Association. A total of 14833 yield records (kg) were collected and several paired standard analyses (fat%, protein%, somatic cell count 000, linear score, dressing% (fat% * 0.9979 + protein% * 4.19 -5.24) and cheese yield (kg*dressing%) were carried out in the two C and B maize silage feeding phases. Over the whole 2013-14 years, about 90% of the controls were conducted during the conventional feeding period and the remaining 10% during the biofertilized feeding period.

Bulk milk samples (No. 98) were regularly examined to establish their quality and cheesemaking properties using a MilkoScan FT120 27 for routine analyses of the fat, protein and lactose and for the registered interferograms (n=1154 digits), and a Fossamatic 90 for the somatic cell count, which was transformed into a Linear Score. Special laboratory analyses were conducted concerning the functional milk properties, namely the fatty acid profile 28,29 and the vitamin A and vitamin E contents 30.

A significant rise in yield was obtained as a result of the biofertilizer treatment, more so for the stem system (+15%) than for the ears (+2%), and an increased S to H ratio (+10%) was observed (Table 1).

The pH of the stem was lowered significantly by the biofertilizer, more so at the bottom measuring point (-2.9%) than at the top point (-0,8%), although the up-bot difference was more pronounced (-144%) (Table 2).

As far as the putative leaf composition is concerned, as fingerprinted in the 1278 NIR spectra, the difference between the two theses was extremely high, with an R² of 0.82 (Figure 1).

Where considering the 1627 extra leaves (Table 3), the foliar pH varied in parallel as the stem pH varied, thus confirming a 3% lower value in the biofertilized maize.

As far as the foliar composition, predicted by means of SCiO^TM NIRS, is concerned (Table 3), the general feature that was observed is a delay in maturity after the treatment, as represented synthetically by the -4% in the crop maturity index. The more juvenile status was testified by more protein % (+6.4), ash (+7), hemicellulose (+13), cellulose (+3) and, conversely, by less lignin (ADL, -5) indigestible NDF (-3) and ADF (-2). The whole-plant composition reflected the consequences of the preponderant part of the stem vs. the ears, and especially in the biofertilized plants, which exceeded the average stem/ear ratio by 10%. The biomass was obviously increased much more in the stem system (+15) than in the ears (+2). Thus, the whole plant reflected the leaf composition as being mixed with the stem and ears parts, and the result was a relative reduction in the fibers, compensated by a rise in protein (+3), but mainly in total sugars (+11). In short, a more palatable and nutritive silage food was obtained.

All items unitless are % dry matter; differences in ln(B/C)% are shown in parentheses; * P<0.05

In the context of the ProLacte project, the results pertaining to the quantity of milk produced by individual cows, carried out by the AIA on 13171 cases of feeding with conventional silomais (C) and 1662 of biofertilized feeding (B), indicated a slightly negative production trend (-2.7%) after a comparative analysis. This is mainly due to the voluntary reduction of the feeding level verified on farm #2 and to sanitary reasons on farm #3 (Table 4).

On the other hand, the results on the composition of the milk are interesting. Overall, the fat content increased by +5.5%, while the protein content increased by +3.3%. This has led to an increase in the dairy dressing percentage of 5.3%, and +2.3% in cheesemaking. In parallel, the results of the bulk-milk analyses (Table 5) pointed out even higher increases in fat (+5.7), protein (+4.6) and in the dressing percentage (+6.9).

A clear characterization of the differences is presented in Figure 2, where the regression of the milk protein on the daily yield in the biofertilized (B) and in the control (C) periods highlights two nearly parallel trends, but a dramatic upper value of 3.3% (in relative log dimension). It should be noted that the B rhombus point in this figure is the actual situation on a farm that has applied a systematic biofertilization for ten years, while the C rhombus point represents the breed average from 682 conventional herds.

Functional results of considerable dietary interest were obtained concerning the vitamins and the composition of the fatty acids. In fact, significant increases were observed for the Vitamin E (+25%) and Vitamin A (+27%) contents on 3 out of the 4 farms (Table 6), with a more favorable response for the lower levels of biofertilization in the conventional period (Figure 3). The two functional properties are correlated (r =0.81).

The amount of saturated fatty acids decreased significantly (-6%): myristic acid (-18); stearic acid -32%, with a balanced increase in unsaturated acids (+15%) and a decrease in the saturated / unsaturated ratio (-25%) (Table 7).

The transition from conventional maize to biofertilized meadow was accompanied by a reduction in the individual milk yield of about 8% (Table 8), in part due to a reduction from a triple milking to the normal double one. Importantly, the living conditions of the cows were dramatically improved, as testified by the calving order, which passed from 1.72 to 2.02 (+18%), together with a reduced need for medical treatments. The reduction in energy in the ration was compensated for by a good roughage, and this was reflected in a superior quality of the milk, not of the fat content, but of the protein content (+7% and +9% respectively for T and B) which influenced the dressing% to a great extent (+8% and +10%), while the cheese yield was penalized by the minor yield.

The value chain examined in the unique farm and pertaining to the biofertilized meadow (B) stage was compared with results obtained for the conventional stage (Table 9). A reduction in the operative cost was relevant, as a result of the reduction in chemicals (fertilizers, herbicides, pesticides) and in mechanical processing in the field operations. Moreover, a major advantage was observed for the dairy system in the stable, as a result of the lower feeding costs, but mainly because of the superior quality of the milk parameters, which compensated for the lower quantity, thus resulting in a positive balance, that is, of about 17% of the yearly milk income.

The value chain pertaining to the biofertilized maize was compared with the results verified in the conventional stage for the average of the four farms (Table 10). A certain advantage was observed for the field operations after the reduction in phosphatic fertilizers, as well as an improved yield, for constant mechanical processing costs. However, as before for the meadows study, a major advantage was observed for the stable and dairy system, as a result of the superior quality of the milk parameters, which compensated for the lower quantity, thus resulting in a positive balance, that is, of about 9% of the yearly milk income.

The pivotal role of the foliar pH as a marker of mycorrhization - as previously published 22 has been revealed in different crops 31 and confirmed in further experiments on olive trees in Apulia 32, but mainly in maize experiments 23, where the foliar pH in a set of 13 pairwise inoculation comparisons on average acidified (-3.7%), while the yield responses ranged from +25.2% to -9.2%. However, the relationship was not confirmed in one experiment with biofertilized tomato 33, but the examined samples were not in a good state of conservation. This multifaceted parameter had already been revealed as being important for the predictability of grape yield variations, almost important as the whole foliar NIR spectra 34. In short, we confirm here a general relevance since the foliar pH has been shown to be a parameter that is inversely correlated with the temperature and with a direct UV-AB exposition in Lepidium sativum35, but is also directly correlated with the soil moisture 36and with the UV in the 24^th solar cycle for grapevine leaves 37.

According to Sabia et al 10, the stalks of plants biofertilized with Micosat F showed significantly lower levels of crude fiber, NDF, ADF and ash (but the P% was raised) than the control plants, according to the presented data, although a greater yield increase (+18%) was observed. Pinar et al. 15, using an AM spore-based biofertilizer, observed a +20% increase and + 39% in a synchronous experiment with sorghum silage, but observed no difference in maize silage composition, while a reduction in NDF (-5%) and an increase in the protein content (+21%) were observed in sorghum, as in the present case.

It seems absurd that corn, the main bioenergy crop throughout the world, is not affected by the results in biofertilization studies. Literature comparisons, limited to the early stages in pot experiments ^{38, 39} were mostly focused on physiological agronomic traits and showed evidence of a higher leaf-supply of N, P, Mg and Ca, but not of K.

A different consideration derives from an expansion of the category of biofertilizers to the biofertilizers issued from biogas. Such products cannot be defined exactly as bio-inoculants since they are derived from anaerobic fermentation, while the useful microflora in the soil, which represent the core of a biofertilizer, are aerobic 40. These products are simply amendments, that is, products used at exponential doses compared to real inoculant biofertilizers, which provide the soil with a mainly organic-N supply. Studies conducted on juvenile plants at the Embrapa Dairy Cattle research unit 41, witnessed a significant increase in the quantity of fresh and dry matter in plants receiving consistent ferti-biofertirrigation through sprinkling, at around 800-1600 mc ha^-1, in comparison with the controls, but without any variations in the dry matter, NDF, ADF, CP or N levels.

Corn silage, despite being scarce in protein, is the most relevant feed for intensive ruminant production. Other important pawns on the cultural chessboard are needed in the sustainable farm, especially in cropping systems. In this strategy, chemical and chicken manure fertilizers, biofertilizers based on N-fixing bacteria and phosphate-solubilizing bacteria have shown profitability in intercropping corn-soybean systems 42and have ranked high in NDF and ADF contents but less in protein and mass yield, while corn-soybean intercrops could slightly increase forage yields and quality, and produce 2% more total protein yields, but also overall produce 39% more than zero-fertilized crops. However, the chemical-organic treatments elevated the zero-fertilizer by about 118%. This is exactly the overly critical point of a biofertilizer advance in real fields: a dose \ response should be adequate for medium - high levels of production, and the results must be confirmed over years and not months. The litterbag-NIRS33,34 is a recommended indirect simple method apt to monitor the evolution of the soil biota, especially after the use of biofertilizers.

In stable practice, the difference between a conventional feeding and a biofertilized one resides in several aspects that are derived from a better palatability and lower level of mycotoxins in the ration, as shown in the Amico project in which the maize grain was measured directly, as well as after poultry feeding experiments 12. The only references in literature for dairy cows are two experiments conducted by the CREA group. According to Chiariotti et al 14, an improvement in the milk protein content (+6% relative) and in the overall animal condition was found, as confirmed by a greater feed intake (+6%) and by an accentuated weight gain during lactation (+118%). A microscopic investigation of the rumen microflora revealed an excess of protozoa (+15%) in a wide variety of shapes and multicolored microbial species, because of the feeding with corn grains treated with Microbial fertilizer microbial biota Micosat F. In the second replicate experiment 16, the AA speculated that a treatment, even under sub-optimal conditions, could affect some peculiar features of the maize and ration, favouring appetite and digestibility. In fact, the group fed biofertilized maize showed a significantly higher and smoother feed intake over the first period of the trial, thus suggesting that the cows were less subject to the physiological disorders caused by a lack of energy during the first part of lactation. The average daily milk yield was higher in the cows in the mycorrhized group (+5.4%), although there were no notable qualitative differences in the bulk milk, except for a favorable reduction of 16% in amyloid A, a marker used to indicate mastitis, and an unfavorable 1% reduction in lactose. Moreover, the cheesemaking features were significantly improved, with -12% in the coagulation time, a sign of a positive interference on the casein features.

The energy balance of dairy cows must be analytically measured from input-output measures and changes in the body reserves, and as this requires the measurement of all the energetic inputs (feed intake) and outputs (milk, fetus, growth), but this it is not feasible under the current commercial conditions. An indirect option that has been put forward for measuring the energy balance is to consider the changes in the milk composition, for example, the changes in the fat and protein contents of the milk. According to the proposal of Friggens et al 43, a tentative PLS relationship:

Energy Balance (Mj d^-1 ) = 82.4 + 5* Fat% - 80.5 (Fat/Protein), has provided an average comparison of ln(B/C) negative on farms 1 (-17%) and 2 (-26%), null on farm 3 and positive on farm 4 (+15%), but +55% on the farm with meadows, where the real livable conditions were improved, as pointed out by the lower elimination rate of the cows that raised the order of the calving.

Milk producers are concerned about the raw commercial features, but consumers may have different opinions, and some are willing to pay for top quality products. Different Spectroscopy and Electronic Nose Techniques can be used to objectively characterize milk features 44build blueprints of different milk systems 45 and track feeding regimes and geographic origins 46. Vibrational spectroscopy can be used to distinguish mountain milk from pasture milk or from the milk of cows fed maize on farms on the plain 47. A quick down-up approach, which can be used to organize the variability of the commercial, nutritional, and aromatic properties of cow milk and grade them into multifaceted feature quality classes, which would be useful for producers, transformers, and consumers, was described 48.

In that survey, which involved a total of 106 dairy farms in Puglia, the farms were classified as belonging to four characteristic dairy Types: 3-SCH (Silage Concentrate-High, n=25); 4-HCH (Hay Concentrate-High, n=33); 5-HCL (Hay Concentrate- Low, n=32); 6-PCZ (Pasture Concentrate Zero, with Podolic cows, n=16).

The comparisons in highlight that the saturated FAs decreased in parallel to a lower intensity of the feeding-cow system, and, conversely, the unsaturated FAs grew. Vitamin A appeared slightly different, for the systems in the two experiments.

According to Bernardini et al. 49, with reference to Holstein dairy cows fed two isoenergetic diets, based on either grass hay or maize silage, the milk from animals fed the green diet contained lower concentrations of saturated FAs and higher levels of polyunsaturated FAs.

Confirmation of the high correlation of the two functional properties (r =0.81) observed in the present work can be found in the work of Strusińska et al 50, who observed that the inclusion of pasture swards in the feed rations raised the contents of liposoluble vitamin A, vitamin E and β-carotene in the milk. A higher concentrate share in the diet - corresponding to raise in yield - also determined a raise in both vitamins but not in the β-carotene.

As the health of a dairy cow depends on the quality of the feed, how much of this health will be transferred to the biologically active components (particularly vitamins and trace elements) that are fundamental for promoting the neonatal and adult health of humans? Previous experiments with mycorrhized corn grain 12 showed that antioxidant qualities were conferred to the animals and transformed into products. In the case of milk, it appears that a better fatty acid profile, less saturated acids, and a greater antioxidant power is conferred to milk for greater vitamin A and E contents. Therefore, the presence of a greater amount of vitamin E in the raw milk analyzed in this study could also have a positive effect on cows and therefore could avoid the onset of infertility phenomena or problems related to disorders of the musculoskeletal system.

However, it is the protein in milk that appears to be a kind of cornerstone of the construction that starts from the mycorrhizal symbiosis, runs through energized metabolic pathways, highlighted by a lower foliar pH in the stalk, and finally determines a better nitrogen efficiency of the entire soil-root-plant-cow system. The experimental evidence obtained from the quantitative analysis of vitamin A and vitamin E in the raw milk of cows fed mycorrhized corn silage could be related to a greater availability of nitrogen present in the mycorrhized cultivar, compared to the untreated cultivar. The results of this study have in fact confirmed many findings in the literature. Some studies have shown how the availability of nitrogen determines a variation in the photosynthetic pigment content, for example, in the chlorophyll a, chlorophyll b, lutein, β-carotene, neoxanthin, xanthine violet, zeaxanthin and xanthine anther contents. A study by Lipppert et al 51, conducted on Phragmites autralis (common reed), highlighted that the photosynthetic pigment content depends on the position of the leaves along the stem. It also stated that the leaves containing a greater quantity of nitrogen also have a greater pigment content; the synthesis of chlorophylls, proteins and amino acids in fact depends on the availability of nitrogen.

Kopsell et al. 52 showed that increasing the total nitrogen concentration, causes a linear increase in carotenoids in the dry plant. Furthermore, the molecular form of nitrogen can alter the accumulation of pigments; an increase in carotenoids was observed in cabbage cultivars where the NO₃-N titer was increased in the "nutrient solution".

Another study 53, conducted on the pea plant (Petroselinum crispum), showed that a higher concentration of nitrogen in the nutrient solution leads to a significant increase in the carotenoid content, in particular lutein, zeaxanthin and β-carotene, and to an increase in the biomass of the plant.

A further study 54 conducted on AM symbiosis in Zea mays, revealed that an accumulation of isoprenoids takes place in the cortical cells of the plant root, in particular of micorradicin and of cyclohexenone derivatives. These two groups of compounds can be synthesized through two routes: a direct route, in which the IPP coming from the MEP route is directly converted into these two compounds, and an indirect route, in which the IPP is initially converted into carotenoids and subsequently, through their oxidative degradation, into micorradicin and cyclohexenone derivatives. Furthermore, it has been seen that the mycorrhized roots of Zea mays contain a greater quantity of carotenoids than the control , especially of violaxanthin and neoxanthin esters. Furthermore, by treating a mycorrhizal Zea mays plant with a herbicide (norflurazon), which inhibits the enzyme phytoene desaturase (PDS) present in the biosynthesis of carotenoids, an accumulation of phytoene was found, which was not present in the non-mycorrhizal corn roots. This indicates that the biosynthesis of phytoene is more active in mycorrhized roots than in non-mycorrhized roots. In fact, it was found that there is a greater transcription of the gene of the enzyme phytoene desaturase in mycorrhizal roots, which was not found in the non-mycorrhized roots. It should be pointed out that the pH showed a basipetal trend, as can be seen in Table 2, a sign of energetic mycorrhizal reactions, even from the roots.

The better protein yield of cows fed with biofertilized forages can be an indicator of a lower dispersion of methane into the environment by lactating cows fed biofertilized maize. Methane is formed from a bacterial biotransformation of hydrogen, as generated by rumen protozoa 55. The defaunation of the rumen from the protozoa is a way of trying to reduce the emissions of ruminants. Various products (bentonite, tannins, yeasts) have been tested in vitro, but the results cannot be measured in practice, except indirectly from the acid profile of the milk 19. In an in-vivo experiment on cows in individual head chambers 56, an Italian product, supplemented at 8 g head d^-1 Enteric, showed a significant reduction in methane emissions of about 20%, while the protein % of the milk increased parallelly by around 5%.

As a corollary, we here report the experience of two colleagues 14,16 who, on examining the rumen fluid of cows fed mycorrhized and non-mycorrhized corn under a microscope, discovered a multiform and multicolored collection of microflora in the former, that is, a colored photograph vs. a black and white one.