Effects of Hybrid and Maturity Stage on in Vitro Rumen Digestibility of Immature Corn Grain

Abstract

This study aimed to evaluate the influences of hybrids (HYB) and maturity stage (SAMP) on in vitro rumen digestibility of immature corn grain. Four HYB (Gigantic, Y43, Klips and 9575) from the FAO group 700 were grown under identical agronomic conditions. First sampling (T1) was done after 95 days from seedling and then 4, 8, 13, 18 and 27 days later (T2 to T6). In vitro starch digestibility (STD_7h) and gas production (72 h) were measured. Whole plant and grain dry matter (WP_DM and GR_DM, respectively) and zein content were significantly affected (P<0.01) by HYB and SAMP. Starch content was significantly affected by HYB, SAMP and their interaction. It increased from T1 to T4 (from 67.47 to 72.82% of GR_DM) and then tended to plateau. Concurrently, STD_7h significantly decreased with advancing SAMP and was also affected by HYB. With advancing maturity, total volatile fatty acids (VFA) significantly decreased, with an increase of acetate and a decrease of propionate molar proportion (P<0.01). Gas production rate (GP_c) was significantly affected by HYB, SAMP and HYB×SAMP. Whole plant grain DM correlated (P<0.01) positively with grain starch content (r=0.60 and 0.64) but negatively with STD_7h (r=-0.39 and r=-0.63) and VFA concentration (r=-0.59 and -0.75). Zein percentage in crude protein negatively affected (P<0.01) total DM (r=-0.65,), STD_7h (r=-0.73) and GP_c (r=-0.68). Results suggest that genotypes and maturity stages influence DM and rumen starch digestibility of immature corn grain and in this respect zein can play a significant role.

Keywords:

• Corn genotypes
• Maturity
• In vitro
• Starch digestibility

Introduction

Nutritive value of corn silage can be affected by genetic, agronomic, harvesting and preservation practices (Johnson et al., 2002b). In recent years, seed companies have increased the interest in selecting corn hybrids specialised for whole plant corn silage production (Johnson et al., 1997; Kuehn et al., 1999). Rumen starch fermentability can affect the nature and amounts of nutrients delivered to the animal and, in turn, milk yield and composition (Nocek and Tamminga, 1991). The stage of maturity of whole corn plants at harvest may affect starch digestibility (STD_7h) (Harrison et al., 1996) and consequently the energy supply to the animal and its milk yield (Bal et al., 1997).

Ruminal degradability of non-fibrous carbohydrates in corn grain largely determines the energy value of this feed (Allen, 1997). The physical structure of starch granules in corn grain influences digestibility of the stored carbohydrates. The texture of the corn grain is determined by the amount of vitreous endosperm proportionally to the farinaceous endosperm (Watson, 1998). Zein is a major component of the protein matrix that envelops corn starch granules and largely affects its rumen and total tract digestibility (Philippeau et al., 2000; Correa et al., 2002). Some recent researches evaluated the corn germplasms for the differences in ruminal starch degradability (Philippeau and Michalet-Doreau, 1997; Correa et al., 2002; Johnson et al., 2002b; Taylor and Allen, 2005) to improve the grain and silage utilisation in ruminants; nonetheless, ruminal digestion of the immature grain fraction has not been widely investigated yet (Verbi et al., 1995). Thus, the objective of this study was to evaluate the changes in chemical composition of grain from different corn hybrids at different stage of maturity and the effects on in vitro ruminal STD_7h.

Materials and methods

Cultivation of different corn hybrids were carried out at the experimental farm Vittorio Tadini, Podenzano (PC), Italy (44°57’26,28’’ N; 9°41’9,60’’ E). The local climate is temperate or type C according to the Köppen-Geiger climate classification (Peel et al., 2007). In terms of rainfall, the area is characterised by typical regime sublittoral with a total annual rainfall equal to 850 to 900 mm distributed on 80 to 85 rainy days. The soil (sand 31.0%, silt 52% and clay 16%) was classified as silt loam (USDANRCS, 1999).

Four corn (Zea mays L.) hybrids belonging to Food and Agriculture Organization (FAO) maturity class 700, 135 CRM group were used: Gigantic (Monsanto, St. Louis, MO, USA), Y43 (Pioneer, Johnston, IA, USA), Klips and 9575 (KWS Saat, Einbeck, Germany). The hybrids were seeded according to a randomised block design with three replicates per hybrid. Parcels were 8 rows large, at a spacing of 70×19 cm, seeded on May 28, 2010 and harvested at six maturity times.

Top dressing fertilisation was made with 200 kg/ha of N after hoeing. For weed control ACETOCLIC (Sipcam Italia S.p.A., Pero, MI, Italy) and MERLIN (Bayer Cropscience Srl, Hyderabad, India) were used at the dosage of 4 L/ha (Acetochlor 1.148 kg/ha + Terbuthylazine 0.574 kg/ha + Furilazole 0.384 kg/ha) and 1.2 L/ha (Isoxaflutole 52.8 g/ha), respectively.

First harvest time of grain was planned in order to have a whole plant dry matter (DM) content within the optimal range suggested for ensiling in horizontal silos (Mueller and Green, 2001). Whole plant moisture content was monitored collecting random samples from the whole field every other day around the target date. First sampling (T1) was done on August 31, 95 days after the seeding date. This happened four days later than the date foreseen based on field monitoring and due to adverse weather conditions. Subsequent samplings were made after 4, 8, 13, 18 and 27 days (T2, T3, T4, T5 and T6, respectively) from T1.

For each hybrid, at each maturity stage, 4 ears were randomly selected from the central rows excluding the first meter at the front and rear border of each parcel. Grains from the same sampling time were pooled from each parcel and dried at 55°C in a forced air oven until constant weight was achieved. At the same sampling time, 6 whole plants per parcel were also harvested, pooled and cut at a 0.7 cm theoretical length. Dry matter content was measured on representative sample by oven drying at 55°C.

Pre-dried samples were ground to pass through the 1 mm sieve of a knife mill (Wiley mill model 3; Thomas Scientific, Swedesboro, NJ, USA). Ground samples were analysed for: crude protein (CP) (AOAC 990.03), ash (AOAC 942.05), total soluble sugar (phenol-sulfuric acid assay; DuBois et al., 1956) and starch content (AOAC 996.11) (AOAC, 1995). Zein was analysed by a modification of the method of Drochioiu et al. (2002). Briefly, 0.5 g of defatted sample was weighed in a TNT bag (F57 bag; Ankom Technology, Macedon, NY, USA), and extracted twice in the ANKOM²⁰⁰ Fiber Analyzer (Ankom Technology) at room temperature with 75% ethanol containing 5 g/L of sodium acetate. Crude protein was measured from this residue (AOAC 990.03; AOAC, 1995). Zein was calculated as the difference between total CP in the original sample and in the defatted and extracted residue. To evaluate the effectiveness of the solubilisation procedure, 0.1 g of pure zein (Acros Organics, Fair Lawn, NJ, USA) in pre-tared F57 bags were treated as the samples, and efficiency of solubilisation was expressed by the ratio between the solubilised and the initial DM.

In vitro digestibility of dry matter and starch

Samples were fermented by in vitro technique according to Mauricio et al. (1999) with some modification to measure DM, starch disappearance and production of volatile fatty acids (VFA). In brief, 0.5 g of ground grain sample were inoculated in a 125 mL glass bottle with ruminal fluid collected 6 h after morning meal from 3 rumen cannulated dairy heifers. The donor heifers were fed a ration based on grass hay (8 kd/d), a corn-soybean meal (85:15, as fed basis) concentrate (2 kd/d) and a mineral-vitamin supplement. Rumen digesta was hand-squeezed and the fluid collected in pre-warmed thermos flasks, transported to the laboratory, filtered through a synthetic cheese cloth with pores of about 500 μm and kept in a water bath at 39°C. All operations were carried out under a flux of CO₂. Forty mL of medium (Goering and Van Soest, 1970) were added by an automatic dispenser (Fortuna^® Optifix^®, Sigma Aldrich, St. Louis, MO, USA) to each serum bottle that was purged with O₂-free CO₂, sealed with a butyl rubber stopper and aluminum crimp cap and moved to a water bath, set at 39°C. The serum bottles were inoculated with 10 mL of inoculum and incubated for 7 h. Fermentation bottles without substrate (blanks) were also included, and samples and blanks were incubated in triplicate.

At the end of the incubation, the serum bottles were cooled in ice water to stop fermentation. Bottle content was filtered on tared crucibles (P2 porosity), then dried in an air oven at 80°C overnight and weighed. Starch content was analysed on the residue to calculate in vitro STD_7h. Starch digestion rate (Starch_K_d) was calculated according to Sniffen and Ward (2011), without any lag, as follows:

A portion of the filtered liquid was also collected and centrifuged at 3000×g for 10 min at 10°C, and the supernatant was stored at -20°C for VFA analysis, later performed according to Ahmed et al. (2013).

In vitro gas production

Samples were fermented to evaluate entity and pattern of in vitro gas production according to the pressure reading technique of Mauricio et al. (1999), with the above mentioned modifications. The serum bottles were incubated at 39°C for 72 h, with three replicates per sample. Headspace gas pressure was measured at 2, 4, 6, 8, 12, 20, 24, 30, 36, 48 and 72 h post-inoculation using a high precision digital pressure gauge (LabDMM; AEP, Cognento, Italy), and the gas produced was released after each measurement. Gas pressure readings were converted to volumes of gas based on a pre-established pressure/volume calibration curve.

Calculations and statistical analyses

The raw gas production values were corrected for gas released from buffered rumen fluid (blank), and net gas values were expressed as mL/g organic matter (OM). Corrected values were fitted with time to the exponential model (Ørskov et al., 1981):

where Y is the cumulated volume of gas produced (mL/g of OM) at time t, b is the maximum potential gas production, c is the gas production rate (mL/h), and t is the incubation time (h). The parameters b (GP_b) and c (GP_c) were estimated by the Marquardt iterative procedure using the PROC NLIN of SAS (SAS, 2008).

Pearson’s correlations between analytical and in vitro parameters were calculated using Proc CORR procedure of SAS (SAS, 2008). Proc MIXED of SAS was used to perform a 2-way analysis of variance (ANOVA) on chemical and nutritional parameters. The model included the fixed effects of corn hybrids (N=4), sampling date (N=6), and block random effect (N=9). The means differences were separated using the Tukey’s range test. P values less than 0.05 were considered significant in the analyses.

Results and discussion

Whole plant and grain dry matter (WP_DM and GR_DM, respectively) contents are shown in Table 1. The mean WP_DM at T1 was 33.05% and at T6 it was 45.59%. There were significant effects of HYB and SAMP (P<0.001) on WP_DM. Hybrid 9575 had the highest DM throughout all the period, and Gigantic tended to have a lower DM. The interaction HYB×SAMP was not significant, suggesting a similar rate of DM change in all the hybrids. The plant DM increased linearly throughout the experimental period by 0.50 points percentage per day (r=0.88, P<0.01).

The average GR_DM was 47.64% at T1 and 67.12% at T6. Hybrids had a significant effect (P<0.01) on DM. In particular, Gigantic showed significantly lower mean DM (53.66%, P<0.05), whereas Klips and 9575 had the highest DM (59.87%, P<0.05). The interaction HYB×SAMP was not significant, suggesting a similar rate of DM change in all the hybrids. The increase of grain DM was linear and as high as 0.71 point percentage per day (r=0.88, P<0.01). Grain dry matter was higly correlated to WP_DM (r=0.83, P<0.01).

For whole plant silage, the optimal harvest point is established based on the achievement of i) maximum DM yield per hectare, reached between 2/3 milk line and black layer stage of maturity (Ganoe and Roth, 1992), and ii) a DM content high enough to avoid silage effluent production but not too high to allow an efficient packing. These goals are achieved when DM ranges from 30 to 40%, depending on the kind of silo used (Mueller and Green, 2001). The beginning of the controls was conditioned by adverse weather conditions that caused a delay in the date of first control, which had been foreseen on the basis of preliminary tests. Dry matter percentage of whole plant at T1 (33.05%) was almost in the middle of the interval (30 to 35% DM) suggesting that is was suitable for being ensiled in horizontal silos (Mueller and Green, 2001). This also coincides with the time at which the whole corn plant reaches the maximum nutritional value. Extension specialists (Petzen and Bertoldo, 2011) suggested that for proper fermentation of corn silage an optimal DM range is between 32 and 38%. In the last sampling time (T6), 27 days after the T1, DM was on average 45.59%, higher than the upper limit of that range. Controls were then extended to monitor changes in grain quality to almost the black layer stage (not reached in this trial), when a significant decrease of digestibility is reported to occur (Bal et al., 1997). The high correlation between WP_DM and GR_DM is in agreement with the observation of Vattikonda and Hunter (1983). The ratio between the changes of DM content in the grain and in the whole plant (0.71/0.50 points percentage per day) was similar to the value usually reported of 1.4 (Vattikonda and Hunter, 1983). Crude protein, starch, soluble sugars, crude fat, ash and zein contents of the grain of different hybrids are shown in Table 1. The CP content slightly decreased during the sampling period, from 9.23 to 8.81% of DM, and it was significantly influenced by SAMP, HYB and HYB×SAMP. Hybrid Y43 showed the highest (9.62% DM on average) and 9575 the lowest CP content (8.08% DM). Associated with advancing maturation, a slow decline of CP and a much more marked increase of starch contents occurred. Similar decline in CP content is frequently reported in the literature (Zeoula et al., 2003; Xu et al., 2010) and largely attributable to the diluting effect of starch accumulation. On the contrary, Masoero et al. (2011) found an increase of CP at black layer, a stage of maturity.

The crude fat content remained almost constant through maturation, and on average it was 3.80% of grain DM. Significant differences were recorded between hybrids, Gigantic and Y43 having a significantly higher (P<0.01; 3.99% of DM, on average) crude fat compared to Klips and 9575 (3.61% of DM, on average).

The starch level was significantly affected by HYB, SAMP and their interaction. Overall, it increased from T1 to T4 (from 67.47 to 72.82% of DM) and then tended to stabilise. Hybrid 9575 showed a more rapid increase across the first 4 sampling dates and the highest mean content (73.77% of DM; P<0.05), whereas the other three genotypes had similar levels (on average 70.52% of DM). The increase of starch content was pronounced and clearly followed a curvilinear pattern tending to plateau at T5 (115 days after sowing), when WP_DM was on average 41.34% and GR_DM 62.76%. Similar increase of starch content at comparable stages of maturity is reported by other authors for corn grain (Masoero et al., 2011; Thornton et al., 1969) and whole plant (Philippeau and Michalet-Doreau, 1997; Sutton et al., 2000; Andrae et al., 2001; Johnson et al., 2003; Jensen et al., 2005). Differences in chemical composition among corn genotypes are also widely reported in the literature. Similar and even more pronounced differences were found for corn grain obtained from genetically different hybrids (Watson, 2003; Berard et al., 2009), but also for whole corn plant (Johnson et al., 2002a) and its fractions (Masoero et al., 2006).

Soluble sugars were significantly affected by HYB, SAMP and their interaction. Sugar content progressively declined in all HYB through maturity from, on average, 10.70% of DM at T1 to 5.49% at T6, and it was negatively correlated with starch content (r=-0.72, P<0.01; Table 2).

The zein content of corn grain was calculated as zein in % of DM, % of CP and % of starch in the grain. The average zein content was 2.35% GR_DM and it was significantly affected (P<0.01) by both HYB and SAMP. As a percentage of CP, zein increased from 23.77% at T1 to 28.30 at T6 and was also markedly influenced (P<0.01) by both HYB and SAMP. When expressed as percentage of total starch, zein was 3.33% with a slight increase through maturation, from 3.28 (T1) to 3.45% (T6). Among HYB, zein ranged between 2.66% (for 9575) to 3.84% of starch (for Y43) and also differred significanly (P<0.01). Zein content was analysed by an original method that used an extraction procedure derived from the method of Drochioiu et al. (2002). In a preliminary unpublished comparison carried out on 20 samples of immature corn grains, this method appeared to be well correlated (r=0.86) with that of Larson and Hoffman (2008) and similar absolute values (2.24 and 2.31% of DM, for this and the Larson and Hoffman method, respectively) were produced. Giuberti et al. (2011) obtained higher zein values for dry corn grains with the method of Larson and Hoffman (2008) (3.30% of DM) compared with that of Drochioiu et al. (2002), with a regression coefficient (R²) as high as 0.60. The same authors reported 8 different samples of dry corn grain, the average zein values ranging from 1.73 to 3.46% of DM, depending on the analytical method used. During the 4-week sampling period, zein increased and the increase was more evident when zein was expressed as a percentage of CP; on the contrary, when zein was expressed on a DM basis these changes were much less pronounced, as the accumulation of starch in the kernel hid this increase. Little data exists on corn grain zein content at this stage of maturity. Murphy and Dalby (1971) reported an increase in the proportion of zein in corn endosperm protein more marked 30 days after pollination which, in the present research, occurred around T1.

Seven hour in vitro digestibility trial

In Table 3, the results from 7 h in vitro digestibility trial are presented for total dry matter (DMD_7h) and STD_7h of the corn grain. In both cases, digestibility decreased with advancing maturity and was significantly affected (P<0.01) by SAMP and HYB. Interaction HYB×SAMP was not significant for both parameters suggesting that despite different absolute values, all the hybrids had a similar decline in digestibility. The mean DMD_7h at T1 was 45.42% but at T6 it was 35.18%. The STD_7h behaved similarly to DMD_7h, but it exhibited higher levels at any sampling time, revealing that starch was more digestible than the other components of the grain. At T1 a mean value of 55.22% was recorded, whereas at T6 it decreased to 41.85%. Starch digestibility appeared to have a negative correlation (Table 2) with whole plant (r=-0.41; P<0.01) and grain (r=-0.64, P<0.01) DM content. Starch_Kd (Table 3) was mathematically derived from STD_7h, and consequently it was not statistically evaluated. It was 0.0440/h on average, decreasing from 0.0540 at T1 to 0.0364/h at T6. In vitro digestibility of DM and STD_7h followed a similar pattern of variation during maturation among the HYB and these two parameters also highly correlated (r=0.96, P<0.01). This outcome was expected as starch accounted for about 70% of total grain DM and is one of its most digestible components. This results confirms that GR_DM digestibility can represent a good estimate of STD_7h, as already proposed by Correa et al. (2002) for mature grains. The level of STD_7h in this study (47.88% on average) is lower than the values (73.3 to 80.1%) reported for corn silage in the USA by Sniffen et al. (2009). Sniffen and Ward (2011) indicated an average STD_7h of 62.0±9.7% for corn silage. It is known that ensiling increases starch rumen digestibility, and this increase also depends on the length of conservation (Mahanna, 2005). The differences between our results and those reported above can be attributed to using unfermented grain, but laboratory methodological differences may have also played an important role. According to Huhtanen and Sveinbjörnsson (2006), digestibility of starch is a difficult parameter to measure, and every method has pros and cons. In vitro digestibility with enzymes or rumen fluid is suggested to be a valuable method to compare samples of similar origin but not to estimate in vivo ruminal or total tract digestibility (Allen, 2012). Starch digestibility decreased markedly during the 4 week long monitoring period in all the hybrids, but Klips tended to reach the minimum earlier (at T5) and remained almost constant at T6. Similar decreases in STD_7h were reported for whole plant corn, fermented or unfermented, harvested from half milk line to black layer or within a whole plant content range similar to that of the present experiment (Andrae et al., 2001; Johnson et al., 2003; Jensen et al., 2005; Masoero et al., 2011). Others authors, on the contrary, found no significant or very limited effects of maturation on STD_7h. Ngonyamo-Majee et al. (2008) worked on corn grain of 33 inbred lines selected for starch and endosperm characteristics and found a significant but limited decrease from half milk line to black layer. Johnson et al. (2002b) reported no significant changes in starch rumen digestibility on different hybrids from 1/3 milk line to black layer. Correa et al. (2002) found a decrease of ruminal starch availability after black layer stage of maturity but not between half milk line and black layer. Starch_Kd, mathematically linked to STD_7h, obviously presented the same behaviour and differences among HYB. It declined from 0.0553 to 0.0491/h. The low temperature (55°C) adopted to dry the grains in this study should not negatively affect their digestibility. Similar rates of starch digestion in the rumen were reported for unfermented and fermented immature corn grain and whole plant (Philippeau and Michalet-Doreau, 1997; Andrae et al., 2001; Johnson et al., 2003; Jensen et al., 2005). Based on a database of in situ degradation from 48 references Offner et al. (2003) attributed to ground corn a degradation rate of 0.0550/h. Higher digestibility rates of corn silage starch are computed also in the Cornell Net Carbohydrate and Protein System (Lanzas et al., 2007). This discrepancy can partially arise from the techniques adopted to measure the digestibility. Secondary particle loss during the in situ incubation can lead to an overestimation of STD_7h, as already outlined by Huhtanen and Sveinbjörnsson (2006) for neutral detergent fibre.

Concentration of VFA measured in the fermentation liquid at the end of the 7h in vitro digestibility trial (Table 3) decreased with increasing maturity, in agreement with the similar trend of DMD_7h and STD_7h (from 39.60 mmol/L at T1 to 34.27 at T6). Significant differences exist among SAMP (P<0.05), but not in the HYB for VFA. Molar proportions of acetate and propionate (Table 3) in total VFA had opposite trends during maturation, although their variations were quantitatively small. Acetate increased from 64.27% at T1 to 65.53% at T6 and, conversely, propionate reduced from 22.83% at T1 to 21.45% at T6. Significant differences existed among HYB for acetate (P<0.01), but not for propionate. Butyrate was on average 10.63% of total VFA, and this proportion did not change significantly with the advancing stage of maturity, but a significant though quantitatively small effect was evidenced for HYB (P<0.001). Minor VFA isobutyrate, valerate and isovalerate represented on average about 2.5% of total VFA and remained unaffected by SAMP. Valerate concentration was not affected, but isovalerate appeared to be significantly influenced by SAMP (P<0.05) and HYB (P<0.01). The effect of advancing maturity on VFA molar proportions, with the progressive rise of the concentration of acetate and the reduction of propionate is in agreement with the reduction of starch availability and degradation rate. Sutton et al. (2000) reported an almost constant STD_7h for whole plant silages when harvested at 23 or 33% of DM and a marked reduction at 38% of DM. In that case, VFA rumen concentration of dairy cattle fed diets formulated with the different silages showed a non-significant tendency towards an increase of acetate and a decrease of propionate with advancing maturity.

In vitro gas production

In vitro gas production was carried out for 72 h in order to allow a good estimation of the parameters describing the kinetics of fermentation (Table 4). In general, GP_b was similar for all the hybrids and increased progressively from T1 to T5, but at T6 a decrease was recorded. The in vitro potential gas yield increased until T5 and then decreased for all the hybrids. If the initial increase can be explained by the increase of starch in grain, the final decrease is justified by the decline of starch fermentability. The declining pathway of rate of gas production from T1 to T6 clearly reflects a progressively reduced availability, mainly attributable to the starch component of GR_DM as maturation progresses. The potential gas yield was significantly correlated with DMD_7h and STD_7h, which is in agreement with Rymer and Givens (1999) and Chai et al. (2004). The GP_c was on average 0.0515/h, and was significantly affected by HYB, SAMP and their interaction. The Klips variety showed a lower rate of gas yield that was consistent across sampling times. Hybrid Klips showed the lowest fermentation rate, in agreement with the levels of DMD_7h and STD_7h of this hybrid.

With advancing maturity, the rate of gas production slightly but regularly decreased from 0.0553/h at T1 to 0.0491/h at T6. This behaviour was evident in all hybrids, but between T5 and T6 the decline tended to be reduced. Fermentation rates were very similar to the values of Starch_K_d. These results confirm, by a different technique, Starch_K_d calculated from STD_7h. For corn grain, gas production rate can be assumed to largely reflect starch fermentation rate (Chai et al., 2004), as starch represents the majority of fermentable OM. This assumption was confirmed here by the high correlation found between GP_c and STD_7h (r=0.75, P<0.01). In a previous study (Ahmed et al., 2013), we measured the rates of gas production in a similar range for naturally occurring soluble sugars, except glucose, fructose and sucrose that had a k_d of about 0.10/h.

Correlations

Correlations between the parameters of chemical composition and in vitro estimates are presented in Table 2. Whole plant and grain DM appeared correlated to most parameters, pointing out the effect of the maturation process on the chemical composition of the corn grain and its nutritional properties. Among others, zein expressed as a percentage of DM, of CP or of starch appeared negatively correlated to in vitro DM and STD_7h and to GP_c as well. The best correlations obtained were for zein expressed as percentage of CP. Its correlation with STD_7h and Starch_Kd was -0.73 (P<0.01) and with GP_c it was -0.68 (P<0.01). A strong correlation was also evident between Starch_K_d and GP_c.

Table 1. Whole plant and corn grain dry matter and chemical composition of corn grain collected at different sampling times from different hybrids.

	SAMP						HYB				SEM	P
	T1	T2	T3	T4	T5	T6	Gigantic	Klips	Y43	9575	SAMP	HYB	HYB×SAMP
WP_DM, %	33.05^a	34.22^b	33.15^a	37.68^c	41.34^d	45.59^e	34.75^a	38.08^c	36.06^b	41.13^d	0.6159	0.0001	0.0002	0.0676
GR_DM, %	47.64^a	52.33^b	55.49^c	60.25^d	62.76^e	67.12^f	53.66^a	60.06^c	57.00^b	59.67^c	1.0560	0.0001	0.003	0.1840
CP, % DM	9.23^b	9.20^b	8.83^a	8.86^a	8.74^a	8.81^a	9.00^b	9.08^b	9.62^c	8.08^a	0.04197	0.0001	0.0001	0.0001
Starch, % DM	67.47^a	70.35^b	71.69^c	72.82^d	72.94^d	72.70^d	71.08^b	70.23^a	70.24^a	73.77^c	0.2551	0.0001	0.0001	0.0001
Sugar, % DM	10.70^d	8.76^c	7.13^b	6.99^b	5.84^a	5.49^a	8.01^c	7.22^b	8.03^c	6.68^a	0.2432	0.0001	0.0004	0.0001
CF, % DM	3.86	3.70	3.82	3.81	3.78	3.83	3.98^b	3.65^a	4.00^b	3.57^a	0.06151	0.1112	0.0008	0.0001
Ash, % DM	1.70^c	1.62^bc	1.51^ab	1.46^a	1.42^a	1.41^a	1.58^c	1.48^a	1.52^b	1.49^ab	0.01411	0.0001	0.0001	0.0001
Zein, % DM	2.20^a	2.28^ab	2.15^a	2.49^bc	2.48^bc	2.50^c	2.24^b	2.51^c	2.70^c	1.96^a	0.04583	0.0001	0.0001	0.3020
Zein, % CP	23.77^a	24.73^bc	24.28^ab	28.04^c	28.33^c	28.30^c	24.98^b	27.63^c	28.08^c	24.28^a	0.4842	0.0001	0.0001	0.2900
Zein, % starch	3.28^b	3.24^ab	3.01^a	3.44^b	3.41^b	3.45^b	3.15^b	3.57^c	3.84^d	2.66^a	0.06629	0.0001	0.0001	0.1173

SAMP, sampling date; HYB, hybrid; T1, sampling done after 95 days from seeding; T2, sampling done after 4 days from T1; T3, sampling done after 8 days from T1; T4, sampling done after 13 days from T1; T5, sampling done after 18 days from T1; T6, sampling done after 27 days from T1; HYB×SAMP, interaction between SAMP and HYB; WP_DM, whole plant dry matter; GR_DM, grain dry matter; CP, crude protein; CF, crude fat.

^a-fMeans with different superscripts in the same row differ significantly within SAMP or HYB (P<0.05).

Table 2. Correlation between main parameters measured on corn whole plant (only dry matter) and grain collected at different sampling times from different hybrids.

Item	GR_DM	CP	Starch	Sugar	Zein, % DM	Zein, % CP	Zein, % starch	DMD_7h	STD_7h	VFA	GP_b	GP_c
WP_DM, %	0.83^*	-0.46^*	0.60^*	-0.72^*		0.39^*	-0.39^*	-0.41^*	-0.59^*	0.48^*	-0.67^*
GR_DM, %		-0.34^*	0.64^*	-0.89^*	0.33^**	0.61^*		-0.63^*	-0.64^*	-0.75^*	0.55^*	-0.87^*
CP, % DM			-0.67^*	0.44^*	0.65^*	0.26^**	0.75^*				-0.37^*
Starch, % DM				-0.72^*			-0.41^*	-0.24^**		-0.49^*	0.50^*	-0.55^*
Sugar, % DM						-0.45^*		0.62^*	0.61^*	0.72^*	-0.54^*	0.86^*
Zein, % DM						0.90^*	0.97^*	-0.60^*	-0.64^*	-0.25^**		-0.45^*
Zein, % CP							0.81^*	-0.65^*	-0.73^*	-0.46^*	0.34^**	-0.68^*
Zein, % starch								-0.49^*	-0.54^*			-0.29^**
DMD_7h, %									0.86^*	0.61^*		0.80^*
STD_7h, %										0.55^*	-0.35^**	0.75^*
Starch_Kd, /h										0.46^*	-0.29^**	0.76^*
VFA, mmol/L												0.77^*
GP_b, /h												-0.50^*

GR_DM, corn grain dry matter; CP, crude protein; DMD_7h, dry matter digestibility after 7 h in vitro fermentation; STD_7h, grain starch digestibility after 7 h in vitro fermentation; VFA, volatile fatty acids; GP_b, maximum potential gas production; GP_c, gas production rate; WP_DM, whole plant dry matter; starch_kd, starch digestion rate.

*P<0.01

**P<0.05. Only significant values are showed.

Table 3. Dry matter and starch digestibility, fermentation rate of starch and volatile fatty acids with their molar percentage.

	SAMP						HYB				SEM	P
	T1	T2	T3	T4	T5	T6	Gigantic	Klips	Y43	9575	SAMP	HYB	HYB×SAMP
DMD_7h, %	45.42	41.14	39.68^d	37.66^c	36.07^b	35.18^a	39.31^c	36.72^a	37.72^b	43.03^d	0.6172	0.0001	0.0001	0.0733
STD_7h, %	55.22	51.39	48.46^d	45.92^c	44.61^b	41.85^a	48.76^b	45.11^a	45.50^a	52.28^c	0.9591	0.0001	0.0009	0.1648
Starch_K^d, /h	0.0540	0.0483	0.0445	0.0413	0.0397	0.036	0.0452	0.0404	0.0408	0.0498
VFA, mmol/L	39.60	38.27	37.28^d	35.32^b	35.95^c	34.27^a	36.89^bc	35.94^a	37.23^c	36.52^ab	0.4991	0.0001	0.0764	0.0617
Acetate, % mol	64.27	64.48	65.05^b	65.20^bc	65.42^bc	65.53^d	64.82^a	65.17^b	64.60^a	65.37^b	0.1362	0.0001	0.0051	0.0994
Propionate, % mol	22.83	22.43	21.93^b	21.91^b	21.67^ab	21.45^a	22.40^b	21.93^a	22.19^ab	21.95^a	0.1138	0.0001	0.1181	0.0750
Butyrate, % mol	10.61	10.72	10.58	10.73	10.58	10.68	10.66^a	10.49^a	11.02^b	10.43^a	0.0923	0.6338	0.0001	0.7920
Isobutyrate, % mol	0.21	0.27	0.31	0.11	0.18	0.16	0.19	0.24	0.24	0.17	0.0699	0.1977	0.6762	0.3890
Valerate, % mol	1.01	1.03	1.03	1.03	1.01	1.03	1.01	1.05	1.05	0.99	0.0234	0.8447	0.1239	0.5754
Isovalerate, % mol	1.07^a	1.08^a	1.09^a	1.12^a	1.11^a	1.15^b	1.07^a	1.13^b	1.13^b	1.08^a	0.0205	0.0487	0.0045	0.6660

SAMP, sampling date; HYB, hybrid; T1, sampling done after 95 days from seeding; T2, sampling done after 4 days from T1; T3, sampling done after 8 days from T1; T4, sampling done after 13 days from T1; T5, sampling done after 18 days from T1; T6, sampling done after 27 days from T1; HYB×SAMP, interaction between SAMP and HYB; DMD_7h, grain dry matter digestibility after 7 h in vitro fermentation; STD_7h, grain starch digestibility after 7 h in vitro fermentation; VFA, volatile fatty acids.

^a-dMeans with different superscripts in the same row differ significantly within SAMP or HYB (P<0.05). Measurements were made after 7 h in vitro fermentation of corn grain collected at different sampling times from different corn hybrids

Table 4. pH and kinetic parameters of gas production calculated from a 72 h in vitro fermentation of corn grains collected at different sampling times from different hybrids.

	SAMP						HYB				SEM	P
	T1	T2	T3	T4	T5	T6	Gigantic	Klips	Y43	9575	SAMP	HYB	HYB×SAMP
pH	6.34	6.34	6.34	6.33	6.34	6.33	6.34^b	6.34^b	6.35^b	6.31^a	0.0075	0.9515	0.001	0.8501
GP_b, mL/g OM	380.92^a	381.83^a	387.35^b	386.33^b	392.33^c	387.13^b	385.84^a	385.81^a	384.98^a	387.29^b	2.0963	0.0001	0.4832	0.4054
GP_c, h^–1	0.0553^e	0.0530^d	0.0517^c	0.0503^b	0.0493^a	0.0491^a	0.0521^c	0.0502^a	0.0516^b	0.0519^b	0.0003	0.0001	0.001	0.0017

SAMP, sampling date; HYB, hybrid; T1, sampling done after 95 days from seeding; T2, sampling done after 4 days from T1; T3, sampling done after 8 days from T1; T4, sampling done after 13 days from T1; T5, sampling done after 18 days from T1; T6, sampling done after 27 days from T1; HYB×SAMP, interaction between SAMP and HYB; GP_b, maximum potential gas production; OM, organic matter; GP_c, gas production rate.

^a-eMeans with different superscripts in the same row differ significantly within SAMP or HYB (P<0.05).

Conclusions

Hybrids of the same FAO class of maturity and planted at the same time differed in WP_DM content when contemporary harvested at different time intervals from seeding. On the contrary, the rate at which the DM content increased with advancing maturity was similar among hybrids. These differences were not exactly reflected in GR_DM content behaviour, suggesting differences in drying rates for grain and vegetative parts of the corn plant.

The rate of ruminal starch digestion for corn grain, evaluated on the basis of two independent in vitro techniques, tended to slightly decline with advancing maturity together with a correlated increase of zein content. These results suggest that also in immature corn grain zein can play a role in controlling rumen starch availability for microbial degradation in the rumen. Hybrids usually utilised for corn silage production differ for some chemical and nutritional features, but differences are quantitatively small and likely of limited importance for practical ruminants feeding. These results need to be confirmed with multiple-year and multi-site trials, as environmental conditions can markedly influence the exploitation of different hybrids potential.

Acknowledgements

The authors gratefully acknowledge the financial support of KWS Italia, Agroservice. The authors are also thankful to the research facilities and other support provided by Azienda Agricola Vittorio Tadini, Podenzano (PC), Italy to carry out this study.