Evaluation of strategies for improvement of grass silage quality for smallholder dairy farmers in Tanzania

PhD THESIS

Silage production strategy has never been widely adopted by majority of dairy farmers in Tanzania. On the other hand, most silage making technologies are expensive and not technically viable under smallholder production. Studies which have been done in developing cheap, practical and technically feasible silage making technologies concentrated on a single dimensional approach of intervention, i.e. pre-ensiling treatments, additives, others on ensiling amount and storage positions. Others dealt with single grass leaving other potential locally available grasses uninvestigated. Studies which specify the most appropriate combination of different techniques to achieve high quality fodder grass silage under smallholder farmers have rarely been conducted. To address these issues, six experiments were conducted in Magadu farm of Sokoine University of Agriculture, Morogoro, Tanzania to evaluate the effectiveness of various technologies used in fodder grass silage production under smallholder dairy farmers. The first study dealt with determination of the effect of grass species, wilting and ensiled amount in shopping plastic bags on silage quality. Elephant grass (Pennisetum purpureum) and guatemala grass (Tripsacum laxum) were established and harvested when they were 1.5 and 1 m tall respectively. One portion of grass was wilted for 24 hours and the other portion was unwilted before ensiling. Each portion was chopped into 4 cm particle length before ensiling. The chopped materials were ensiled either in portions of 5 or 10 kg in plastic bag silos. Treatments were assigned to a completely randomized design in factorial arrangement (2 x 2 x 2) as two grasses (elephant and guatemala grasses), two pre-ensiling treatments (unwilted and wilted) and two ensiled amounts (5 and 10 kgs) with two replications. The silage was opened and sampled after 60 days, analyzed for dry matter (DM) losses, chemical composition, fermentation and sensoric qualities and in vitro DM digestibility (IVDMD). The results showed that elephant grass silage had higher (p< 0.05) sensoric scores, crude protein (CP), ash, lactic acid (LA), acetic acid (AA) and bulk density (BD) but lower (p< 0.05) DM, water soluble carbohydrate (WSC), IVDMD, pH, Ammonia Nitrogen (NH3N) and DM loss than guatemala grass silage. Wilted fodder grass showed higher (p< 0.05) sensoric scores, DM, CP, ash, WSC, pH, LA, AA and bulk density but lower Neutral Detergent Fiber (NDF), NH3N, butyric acid (BA) and DM loss than unwilted fodder grass. The 5 kg silage showed higher (p< 0.05) bulk density than 10 kg silage in plastic bag silo. There was no difference between the two ensiled amounts (5 and 10 kg) in terms of appearance, smell and texture scores, DM, CP, WSC, ash, NDF, IVDMD, pH, NH3N, LA, AA, BA and DM losses. The second experiment was conducted to investigate the effect of grass species and different levels of maize bran on silage quality. Elephant grass, guatemala grass and rhodes grass were established and harvested when they were at the age of 120, 63 and 56 days, respectively. Nutritive value declines quickly as plant matures thus were harvested at respective recommended stages of growth for each grass. The harvested grasses were chopped into 4 cm length and subdivided into four portions each of which was treated with different level of maize bran (0, 5, 10 and 15%). The chopped materials were ensiled in portions of 5 kg in plastic bag silo. Treatments were assigned to a completely randomized design in factorial arrangement (3 x 4) as three grasses (elephant, guatemala and rhodes grasses) and four maize bran levels (0, 5, 10 and 15%) with two replications. The silage was opened after 60 days, sampled and analyzed for chemical composition, fermentation, sensoric qualities, IVDMD and keeping quality stability. Elephant grass produced higher (p< 0.05) quality silages than those produced by guatemala and rhodes grasses as indicated by higher (p< 0.05) sensoric qualities, CP, LA and stability but lower (p< 0.05) pH and NH3N. Maize bran at 10% level produced higher (p< 0.05) quality silages than maize bran at 0, 5 and 15% levels as indicated by higher (p< 0.05) sensoric scores, CP, WSC, LA, AA and stability but lower (p< 0.05) NDF, pH and NH3N. The interaction between grass species and different maize bran levels showed that, elephant grass silage with maize bran at 10% level produced best silage as indicated by highest sensoric scores, CP, LA and stability but lowest pH. The third experiment focused on assessing the effect of grass species and different levels of molasses on silage quality. The grasses were harvested when the re-growth was 1.5 and 1 m tall for elephant and guatemala grasses respectively and at flowering stage of growth for rhodes grasses. The harvested grasses were chopped into 4 cm length and subdivided into three portions each of which was treated with different levels of molasses (0, 3 and 5%), packed in 5 kgs plastic bag silo and stored in thatched barn. Treatments were assigned to a completely randomized design in factorial arrangement (3x3) as three grasses (elephant, guatemala and rhodes grasses) and three levels of molasses (0, 3 and 5%) with two replications. The silage was opened and sampled after 60 days, analyzed for sensoric qualities, chemical composition, IVDMD and fermentation. Elephant grass produced higher quality silage and preserved better than guatemala and rhodes grasses as indicated by higher (p< 0.05) sensoric qualities, CP, ash, LA, AA and stability but lower DM, pH, NH3N and BA. Silages produced from molasses at 5% level had higher (p< 0.05) quality and preserved better than silages mixed with molasses at 0 and 3% levels as indicated by higher (p< 0.05) DM, CP, WSC, ash, IVDMD, LA, AA and stability but lower (p< 0.05) NDF, pH and NH3N. Elephant grass at 5% level of molasses showed highest CP and WSC but lowest pH and NH3N. The fourth experiment aimed at determination of the effect of wilting, chopping length and maize bran additive levels on fodder grass silage quality. Elephant grass was harvested when the re-growth was 1.5 m tall. The harvested grass was divided into two portions. One portion of grass was wilted for 24 hours prior to ensiling while the other portion was ensiled unwilted. Before ensiling each portion was chopped into either 2 or 4 cm length pieces using a machete. Within each chop length (2 or 4 cm) the material was subdivided into four portions each of which was treated with one of the four levels (0, 5, 10 and 15%) of maize bran. Treatments were assigned to a randomized factorial design (4 x 2 x 2) as four maize bran level (0, 5, 10 and 15%), two pre-ensiling treatments (wilted and unwilted) and two chopped treatments (2 or 4 cm) with two replications. The silage was opened and sampled after 60 days analyzed for; sensoric qualities, chemical composition, in vitro DM digestibility, fermentation characteristics and stability. Wilted grass produced silages with higher (p< 0.05) appearance, smell and texture scores, DM, CP, WSC, LA AA, pH and stability but lower (p< 0.05) IVDMD, NDF, NH3N and butyric acid than unwilted grass silages. Silages produced from 2 cm chop showed higher (p< 0.05) sensoric scores, DM, CP, WSC and IVDMD, LA, AA and stability but lower (p< 0.05) NDF, pH, NH3N and butyric acid than those from 4 cm chop. Silages treated with maize bran at 10 and 15% levels had higher (p< 0.05) DM, WSC and ash but lower (p< 0.05) NDF than the other levels (0 and 5%). Silages treated with maize bran at 10% level had higher (p< 0.05) sensoric scores, CP, IVDMD, lactic acid, acetic acid and stability but lower (p< 0.05) NDF, pH, NH3N and butyric acids than the other levels (0, 5 and 15%). Wilting with 2 cm chop and maize bran at 10% produced silage with highest IVDMD, CP and acetic acid. Experiment 5 dealt with the investigation of the effect of wilting, chopping lengths and different levels of molasses on grass silage. Elephant grass (Pennisetum purpureum) was harvested when the re-growth was 1.5 m tall. The harvested grass was divided into two portions. One portion of grass was wilted for 24 hours prior to ensiling while the other portion was ensiled unwilted. Before ensiling each portion was chopped into either 2 or 4 cm length pieces using a machete. Within each chop size (i.e. 2 or 4 cm) the material was subdivided into three portions each of which was treated with one of the three levels (0, 3 and 5%) of molasses as additive. Treatments were assigned to a randomized factorial design (3 x 2 x 2) as molasses additive levels (0, 3 and 5%), two pre-ensiling treatments (wilted and unwilted) and two chopping treatments (2 or 4 cm) with two replications. The silage was opened and sampled after 60 days, analyzed for sensoric qualities; chemical composition IVDMD, fermentation characteristics and stability. Wilted grass produced silages with higher (p< 0.05) appearance, smell and texture scores, higher DM, CP, WSC, ash, LA, AA, pH and stability but lower NDF, IVDMD, NH3N and BA than unwilted grass silages. Silages produced from 2 cm chop showed higher (p< 0.05) sensoric scores, DM, CP, WSC, IVDMD, LA, AA and stability but lower NDF, pH, NH3N and BA than those from 4 cm chop. Silages treated with molasses at 5% level had higher (p< 0.05) sensoric scores, DM, CP, WSC, ash, IVDMD, LA, AA and stability but lower (p< 0.05) NDF, pH, NH3N and BA than silages treated with molasses at 0 and 3% levels. Wilting with 2 cm chop and molasses at 5% produced silage with highest sensoric scores, DM, WSC, LA but lowest NDF, pH and NH3N. Experiment 6 dealt with determination of the effect of grass species, ensiled amount in shopping plastic bags and storage positions on grass silage quality. The grasses were harvested when the re-growth was 1.5 m and 1 m tall for elephant and guatemala grasses and at flowering stage of growth for rhodes grasses. The harvested grasses were chopped in 2 cm particle length before ensiling. The chopped materials were ensiled in portions of 5, 10 or 12 kg in plastic bag silos. Then the plastic bags were stored either in thatched barn or in trench. Treatments were assigned to a completely randomized design in factorial arrangement (3 x 3 x 2) as three grasses (elephant, guatemala and rhodes grasses), three ensiled amount of grasses in shopping plastic bag silos (5, 10 and 12 kgs) and two storage positions (thatched barn and trench) with two replications. The silage was opened and sampled after 60 days, analyzed for chemical composition, fermentation, sensoric qualities, IVDMD and stability. Elephant grass produced higher (p< 0.05) quality silage and preserved better than those of guatemala and rhodes grasses as indicated by higher (p< 0.05) sensoric scores, CP, ash, LA, AA and stability but lower (p< 0.05) pH, NH3N and butyric acid. The 5 and 10 kg ensiled amounts showed higher (p< 0.05) appearance, smell and texture scores, DM, CP, WSC, ash, IVDMD, LA, AA and stability but lower (p< 0.05) NDF, pH, NH3N and BA than 12 kg ensiled amounts. There was no difference (p< 0.05) between the two ensiled amounts (5 and 10 kg) in terms of appearance, smell and texture scores, DM, CP, WSC, ash, NDF, IVDMD, pH, NH3N, LA, AA and BA. There were no differences (p< 0.05) between silages stored in thatched barn and trench in terms of sensoric scores, DM, CP, WSC, ash, NDF and IVDMD, pH, NH3N, LA, AA, BA and stability. It was therefore concluded that, wilted elephant grass chopped 2 cm, treated with either maize bran at 10% or molasses at 5% levels, ensiled either at 5 or 10 kg in shopping plastic bag silo, stored in either thatched barn or in trench was the most optimal combination of technologies to achieve high quality grass silage.