ANIMAL FEED SCIENCE AND TECHNOLOGY

E LS EV I ER Animal Feed Science and Technology 47 (1994) 125-139

Nutritional and antinutritional characteristics of selected Vicia genotypes

V.A. Aletor *'1, A.V. Goodchild, A.M Abd E1 Moneim International Center for Agricultural Research in the Dry Areas (ICARDA), PO Box 5466,

Aleppo, Syria

(Received 30 December 1992; accepted 26 October 1993)

Abstract

Mature seeds of Vicia narbonensis (91 lines), Vicia sativa (23 lines), Vicia ervilia ( 16 lines) and Vicia palaestina ( 16 lines) were evaluated for crude protein (CP), protein pre- cipitable tannin (PPT), vanillin-HC1 catechin equivalent (CE) and trypsin inhibitor ac- tivity (TIA). Other chemical and in vitro characteristics measured included ash, organic matter (OM), neutral detergent fibre (NDF), acid detergent fibre (ADF), in vitro or- ganic digestibility ( IVOMD) and in vitro dry matter (DM) digestibility (IVDMD).

Crude protein content ranged from 264.38 + 6.34 g kg- 1 DM in V. ervilia to 341.87 + 8.55 g kg- 1 DM in V. palaestina with all species showing little intraspecies variability. The DM, OM and ash contents of all the lines were similar. However, the ADF and NDF values for V. narbonensis were significantly (P< 0.05) higher than those of the other species. Mean IVDMD and IVOMD values of the species also differed significantly (P< 0.05 ). The higher IVDMD (90.95%) and IVOMD (88.30%) values of V. ervilia paralleled the relatively lower ADF and NDF values.

PPT was mainly restricted to V. narbonensis while none was detected in any of the lines of either V. sativa or V. ervilia. Tannin expressed as CE was no more than 8.27 g kg- t in any of the species and seven of the 91 lines of V. narbonensis had no detectable CE levels. There was considerable intraspecies variability in CE as indicated by the high coefficients of variation (CV), which ranged from 17.48% in V. palaestina to 68.99% in V. sativa. In comparison with most conventional legumes, TIA values were generally low in all the spe- cies, ranging from 0.36 g kg -~ DM in selection IFLVP 2524 to 5.00 gkg -1 DM in IFLVS 2560. Values for TIA in V. palaestina were particularly low (0.51 + 0.13 g kg- ~ DM) and a CV of 25.86% suggests that selection for lower inhibitor levels is possible.

Seed weight was significantly (P<0.05) correlated with CE in both V. sativa and V. palaestina, while CP was weakly negatively correlated ( r= - 0.48 ) with TIA among the V.

*Corresponding author. tPresent address: Institut fiir Tieremahrung, University of Bonn, Endenicher Allee 15, Bonn 1, Germany.

0377-8401/94/$07.00 © 1994 Elsevier Science B.V. All rights reserved SSD10377-8401 (93) 00568-G

126 V.A. Aletor et al. / Animal Feed Science and Technology 4 7 (1994) 125-139

sativa genotypes. There was no significant relationship between seed weight and TIA in any of the species investigated.

1. Introduction

The principal agricultural activity in the dry areas of West Asia and North Af- rica (WANA) is small ruminant (sheep and goat) production with rangeland grazing providing most of the animal feed (Cocks et al., 1986). During periods of high livestock feed prices, feed legumes such as Vicia species, which may be fed to small ruminants in the form of straw and grain, are increasingly attractive to farmers in WANA (Abd E1 Moneim et al., 1988). In view of the overall objec- tives of The International Center for Agricultural Research in the Dry Areas (ICARDA) of developing sustainable farming systems in the dry areas, there is a clear need for feed legume crops. One of the most attractive legume species for grain and straw production is Narbon vetch, Vicia narbonensis. It is a good source of protein with seeds containing up to 28% protein and yielding about 365 kg protein ha- ~ while the straw contains about 9% protein (Abd E1 Moneim, 1992 ).

ICARDA currently maintains breeding lines of Vicia species which include V. narbonensis L., Vicia sativa L., Vicia ervilia and Vicia palaestina R. In this study, we present compositional data on selected lines of these species with respect to their crude protein (CP), trypsin inhibitor activity (TIA), catechin equivalents (CE) and protein precipitable tannin (PPT) content. These parameters are of interest because both trypsin inhibitors and tannins are implicated in poor pro- tein digestibility (Liener, 1969, 1989 ), while tannins generally have been asso- ciated with poor digestibility and palatability (Menhansho et al., 1987; Hager- man et al., 1992). We hope that this information, along with in vitro characteristics, will arouse greater interest in the exploitation of their potential in human and/or animal feeding.

2. Materials and methods

The Vicia genotypes evaluated were obtained from the breeding programme at ICARDA (Tel Haydia, Syria; 36°40'N, 37°20'E, 390 m elevation). The current breeding procedure uses germplasm evaluation and seed multiplication in nurs- ery rows, and evaluates selected genotypes in microplots and promising geno- types in advanced yield trials. The samples analysed were mature grains compris- ing 91 lines of V. narbonensis (Narbon vetch), 23 lines of V. sativa (common vetch), 16 lines of V. ervilia (bitter vetch) and 16 lines of V. palaestina. About 50 g each of the dried seeds were ground before use using a UD cyclone sample mill (UD Corporation, Fort Collins, CO, USA) fitted with a 1 mm mesh sieve.

The dry matter (DM), CP and ash were determined as described by the Min-

V.A. Aletor et al. / Animal Feed Science and Technology 47 (1994) 125-139 127

istry of Agriculture, Fisheries and Food (1981), while neutral detergent fibre (NDF) and acid detergent fibre (ADF) were determined by the method of Goer- ing and Van Soest (1970). Some random selections were also evaluated for in vitro organic matter digestibility (IVOMD) and in vitro dry matter digestibility (IVDMD) using the technique of Tilley and Terry (1963). The weight of 100 seeds was estimated using a mechanical seed counter (Numigral, Franken, Neth- erlands) before weighing.

2. I. Assay for antinutritional factors

2.1.1. Protein precipitable tannin (PPT) and catechin equivalent (CE) All samples for tannin determination were milled just before analysis to pre-

vent the oxidation of phenolic compounds. A 200 mg measure of the flour was defatted in a screw-top test-tube with 5 ml of ethyl ether for 15 min using a Gal-

Table 1 Crude protein (CP), protein precipitable tannin (PPT) , tannin measured as vanillin-HC1 catechin equivalent (CE) and trypsin inhibitor activity (TIA) of V. narbonensis seeds selected in microplots

Selection Weight per 100 DM CP PPT CE TIA no. seeds (g) (%) (gkg -~ DM) (gkg - l DM) (gkg - l DM) (gkg - 1 D M )

IFLVN2376 19.371 90.79 301.02 ND 1.38 4.13 IFLVN 2377 21.168 90.65 294.54 4.37 1.65 3.49 IFLVN2378 20.868 90.65 306.56 4.62 1.65 3.34 IFLVN 2379 17.928 90.72 285.16 4.62 2.48 3.85 IFLVN 2381 19.192 90.67 303.85 4.19 1.38 3.04 IFLVN 2382 17.664 90.59 303.34 4.54 1.93 3.71 IFLVN2384 20.096 90.72 276.01 4.44 1.38 3.20 IFLVN 2385 20.686 90.64 277.69 4.19 1.38 3.34 IFLVN2386 16.968 90.77 274.43 4.19 1.38 3.04 IFLVN2389 18.85 90.60 276,93 4.28 1.38 3.41 IFLVN 2394 15.838 90.67 285.43 4.37 1.93 3.71 IFLVN2395 21.24 90.81 285.87 4.61 2.75 4,57 IFLVN2396 16.399 90.70 288.09 4.44 2.48 4.13 IFLVN 2397 19.208 90.60 288.30 4.54 2.48 4,58 IFLVN 2398 13.507 90.53 276.15 4.45 2.76 3,13 IFLVN2460 14.296 90.69 278.20 ND 2.76 3,20 IFLVN 2463 18.691 90.62 290.22 4.11 2.76 3.49 IFLVN 2482 16.371 90.70 288.31 ND 1.93 3.34 IFLVN2597 14.658 90.51 284.06 4.38 2.49 3.35 IFLVN2598 20.866 90.72 295.19 4.53 2.48 4.35 IFLVN 2599 13.129 90.78 283.98 4.36 3.30 2.83 IFLVN 2600 17.8 90.65 301.27 4.19 1.93 3.34 IFLVN 2601 18.607 90.65 325.32 ND 1.38 3.34 IFLVN 2602 15.329 90.71 292.25 4.28 1.93 3.70

Mean 17.864 90.67 290.09 4.38 2.06 3.57 SD 2.428 0.07 11.95 0.16 0.58 0.47 CV (%) 13.593 0.08 4.12 3.59 28.17 13.21

SD, standard deviation; CV, coefficient of variation; ND, none detected.

128 I~.A. Aletor et al. / Animal Feed Science and Technology 47 (1994) 125-139

Table 2 Crude protein (CP), protein precipitable tannin (PPT), tannin measured as vanillin-HCl catechin equivalent (CE) and trypsin inhibitor activity (TIA) of I~. narbonensis seeds multiplied in the nursery

Selection no. Weight per 100 DM CP PPT CE TIA seeds (g) (g) (gkg -1DM) (gkg - I D M ) (gkg -~DM) (gkg -~ DM)

IFLVN 111 11.226 90.82 297.40 4.36 4.40 3.77 IFLVN 112 12.411 90.75 322.09 4.44 2.48 3.12 IFLVN 113 11.417 90.94 295.69 4.35 3.30 3.47 IFLVN 114 12.651 90.71 301.40 4.19 2.76 3.20 IFLVN 123 15.252 90.68 296.87 4.53 2.48 3.77 IFLVN 124 12.045 90.82 273.40 5.14 4.68 2.90 IFLVN 125 12.807 90.72 307.76 4.71 2.76 3.20 IFLVN 126 11.901 91.12 315.30 4.77 2.47 3.25 IFLVN 127 13.186 91.27 258.90 4.94 3.83 2.88 IFLVN 129 11.666 91.16 275.78 5.04 4.39 3.47 IFLVN 130 12.404 91.21 261.92 5.46 4.66 3.32 IFLVN 543 12.648 91.19 263.08 5.55 4.93 3.18 IFLVN 569 14.168 90.68 309.88 4.53 1.93 3.48 IFLVN 580 15.746 90.71 295.78 4.53 2.76 3.91 IFLVN 584 9.713 91.19 258.91 4.68 1.92 2.30 IFLVN 803 9.684 90.91 286.88 5.22 3.85 3.33 IFLVN 1142 8.344 91.06 253.02 5.73 3.29 4.48 IFLVN 1144 13.379 90.99 275.85 4.96 3.57 3.33 IFLVN 1147 7.295 91.08 282.72 4.35 3.84 3.90 IFLVN 1148 6.518 91.04 307.12 5.04 3.57 3.39 IFLVN 1150 7.682 91.04 281.96 5.04 2.20 4.48 IFLVN 1152 6.807 91.13 288.05 4.60 4.39 3.47 IFLVN2272 10.675 90.87 292.73 5.23 2.75 3.70 IFLVN2524 12.723 90.83 298.58 4.70 2.75 4.05 IFLVN2525 12.774 90.95 291.15 4.18 2.20 3.11 IFLVN2528 14.031 90.56 282.13 4.11 1.93 4.36 IFLVN 2529 11.58 90.77 278.07 4.10 2.75 3.12 IFLVN2530 12.389 91.36 304.18 4.50 1.92 3.10 IFLVN2531 14.282 90.92 290.59 4.78 2.75 3.04 IFLVN2532 12.441 90.61 292.68 4.19 1.93 2.90 IFLVN2533 12.952 90.57 257.26 5.16 3.04 2.54 IFLVN2556 12.371 90.79 280.65 5.45 2.20 2.83 IFLVN 2624 5.612 89.28 311.38 ND 1.96 4.28 IFLVN 2628 6.917 91.13 298.36 5.12 1.92 4.40 IFLV-N 2644 9.198 90.93 292.09 4.18 1.37 4.05 IFLVN 2649 6.88 90.80 279.41 5.23 4.41 3.91 IFLVN 2663 8.623 91.06 247.42 4.51 0.82 2.24 IFLVN 2696 9.508 91.15 259.13 4.17 1.92 3.10 IFLVN 2697 8.58 91.12 258.56 4.08 2.47 3.18 IFLVN 2702 8.488 90.78 272.75 ND 1.38 3.26 IFLVN 2703 8.55 90.85 279.91 5.05 2.20 3.26 IFLVN2706 12.43 90.76 313.02 ND 0.83 3.21

Mean 10.951 90.89 285.47 4.74 2.81 3.41 SD 2.579 0.32 18.72 0.45 1.05 0.54 CV (%) 23.547 0.35 6.56 9.51 37.48 15.84

SD, standard deviation; CV, coefficient of variation; ND, none detected.

V.A. Aletor et al. / Animal Feed Science and Technology 47 (1994) 125-139 129

Table 3 Crude protein (CP) , protein precipitable tannin (PPT) , tannin measured as vanillin-HCl catechin equivalent (CE) and trypsin inhibitor activity (TIA) of advanced lines of V. narbonensis seeds

Selection Weight per 100 DM CP PPT CE TIA no. seeds (g) (%) (gkg - 1 D M ) (gkg - I DM) (gkg - 1 D M ) (gkg -~ DM)

IFLVN 2561 10.596 90.66 296.82 4.10 ND 3.77 IFLVN 2380 17.425 90.69 274.56 ND ND 2.68 IFLVN 2383 19.821 90.75 295.54 4.19 ND 2.68 IFLVN2387 18.833 90.83 270.18 ND 0.83 2.47 IFLVN 2388 16.035 90.89 306.85 ND ND 2.89 IFLVN2390 16.917 90.04 294.20 ND 0.83 3.00 IFLVN 2391 17.686 90.87 305.38 ND ND 2.83 IFLVN2392 16.533 90.91 277.97 ND ND 2.32 IFLVN2393 20.139 90.88 305.68 ND ND 2.61 IFLVN 2461 21.405 91.09 272.59 ND 1.37 3.53 IFLVN 2462 19.914 91.15 277.56 ND 1.92 3.69 IFLVN2464 22.247 90.87 281.06 4.09 1.38 3.92 IFLVN 2465 18.357 90.89 284.52 4.09 1.38 4.13 IFLVN 2466 19.954 90.97 309.77 ND 0.55 4.34 IFLVN 2467 21.63 90.78 311.30 ND ND 3.04 IFLVN2468 21.386 90.67 303.85 ND 0.55 2.33 IFLVN 2469 21.001 90.81 313.51 ND 0.55 3.48 IFLVN2470 22.638 90.72 293.21 ND 0.55 3.22 IFLVN 2471 24.412 90.71 294.34 ND 1.10 2.01 IFLVN2473 13.735 91.00 287.80 ND 0.55 3.21 IFLVN 2474 19.609 90.79 304.66 ND 0.55 4.87 IFLVN2475 23.086 90.85 294.88 ND 0.83 3.82 IFLVN 2476 19.144 90.87 297.35 4.18 1.65 3.39 IFLVN 2477 21.354 90.83 292.63 ND 1.38 4.00 IFLVN 2478 19.576 91.13 287.50 4.17 2.19 4.51

Mean 19.337 90.83 293.35 4.14 1.07 3.31 SD 2.977 0.21 12.42 0.04 0.51 0.73 CV (%) 15.393 0.23 4.23 1.01 48.14 22.11

SD, standard deviation; CV, coefficient of variation; ND, none detected.

lenkamp Lab Quake followed by centrifugation. The extract was discarded and the air-dry residue was thereafter extracted with 60% methanol; the PPT content of the extract was determined using standard bovine serum albumin (Lot 41H0520; Sigma, St. Louis, MO, USA) as outlined by Hagerman and Butler ( 1978 ). CE, which detects simple flavonoids as well as condensed tannins, was determined by the vanillin-HC1 method of Burns (1971 ) as modified by Price and Butler (1977) and Price et al. (1978). Commercial ( _+ )-catechin (Lot C- 0774; Sigma) was used as reference standard.

2.1.2. Trypsin inhibitor activity (TIA) The TIA content of the grains was measured as the extent to which an extract

of the flour inhibited the action of bovine trypsin (EC 3.4.21.1 ) on the substrate benzoyl-DL-arginine-p-nitroanilide hydrochloride (Lot 110H0329; Sigma) as

130 V.A. Aletor et al. / Animal Feed Science and Technology 47 (1994) 125-139

Table 4 Crude protein (CP), protein precipitable tannin (PPT), tannin measured as vanillin-HC1 catechin equivalent (CE) and trypsin inhibitor activity (TIA) in seeds from advanced lines of V. sativa

Selection Weightper 100 DM CP PPT CE TIA no. seeds (g) (%) (gkg - l DM) (gkg -1DM) (g kg - l DM) (gkg -~ DM)

IFLVS 545 4.995 90.66 336.53 ND 2.76 3.57 IFLVS 1134 4.794 90.63 306.96 ND 2.21 4.36 IFLVS 1135 5.534 90.72 302.69 ND 1.38 3.82 IFLVS 1136 4.397 90.79 343.21 ND 4.41 4.17 IFLVS 1437 4.908 91.12 344.27 ND 3.29 3.98 IFLVS 1448 2.08 90.74 307.25 ND 5.51 4.44 IFLVS2019 5.822 90.72 315.81 ND 2.20 4.96 IFLVS 2037 2.539 90.54 294.35 ND 6.07 4.88 IFLVS 2044 4.227 90.69 355.94 ND 1.93 4.61 IFLVS 2057 3.802 90.61 344.88 ND 4.97 4.49 IFLVS 2062 6.729 90.60 298.45 ND 1.66 4.88 IFLVS 2065 2.103 90.70 315.88 ND 8.27 4.61 IFLVS 2083 4.898 90.67 318.85 ND 1.93 4.00 IFLVS 2086 5.37 90.73 323.49 ND 1.93 4.09 IFLVS2096 4.711 90.81 328.38 ND 1.93 2.95 IFLVS 2106 4.518 90.88 306.34 ND 3.30 3.74 IFLVS 2108 5.313 90.55 322.92 ND 0.55 4.88 IFLVS 2109 4.86 90.96 301.36 ND 1.92 4.08 IFLVS 2556 5.509 90.75 312.95 ND 0.83 4.53 IFLVS 2557 2.464 89.93 306.68 ND 5.00 4.65 IFLVS2539 5.716 89.74 292.29 ND 0.28 4.05 IFLVS 2603 5.522 89.95 284.71 ND 1.39 4.01 IFLVS 2560 4.805 90.07 335.18 ND 1.39 5.00

Mean 4.592 90.59 317.36 ND 2.83 4.29 SD 1.209 0.33 18.69 ND 1.95 0.50 CV (%) 26.336 0.37 5.89 ND 68.99 11.58

SD, standard deviation; CV, coefficient of variation; ND, none detected.

modified by Smith et al. (1980). The flour was first defatted at room tempera- ture with 50 ml portions of petroleum ether (b.p. 40-60°C) in covered flasks and then mechanically shaken until fat-free. The extract was discarded and the residue air-dried. Extracts were thereafter made from 1 g of the flour with 10 mM NaOH for 3 h at room temperature using an orbital shaker (Gallenkamp, UK). The pH of the resulting slurry was adjusted to 9.5 with 1 M NaOH or 1 M HCI. After extraction, the suspension was shaken and diluted with distilled water such that 1 ml of the diluted extract produced trypsin inhibition of between 40% and 60% at 37 °C. The respective dilutions were noted and, in the present assay, di- lutions of 5 or 6 were suitable for all the species except F. palaestina extracts which needed no dilution. The TIA was subsequently determined in terms of mg pure trypsin g - l sample using the equation TIA = 2.632DA/S, where D is dilution factor, A is change in absorbance at 410 nm resulting from trypsin inhibition per cm 3 of diluted sample extract, S is weight of sample (Smith et al., 1980 ).

V.A. Aletor et al. / Animal Feed Science and Technology 47 (1994) 125-139 131

Table 5 Crude protein (CP), protein precipitable tannin (PPT)m tannin measured as vanillin-HC1 catechin equivalent (CE), and trypsin inhibitor activity (TIA) in seeds from advanced lines of V. ervilia

Selection Weight per 100 DM CP PPT CE TIA no. seeds (g) (%) (gkg -~ DM) (gkg - ~ D M ) (gkg -~ DM) (gkg -~ DM)

IFLVE2508 5.032 90.35 262.65 ND 1.66 2.71 IFLVE 2509 4.619 90.27 259.11 ND 0.83 3.41 IFLVE2510 4.384 90.23 261.44 ND 1.66 3.50 IFLVE 2511 4.204 90.50 270.28 ND 1.66 3.40 IFLVE 2512 4.883 90.40 254.65 ND 1.66 3.50 IFLVE 2513 4.891 90.40 259.07 ND 2.21 3.58 IFLVE2514 4.621 90.17 257.85 ND 2.22 3.59 IFLVE 2515 4.772 90.30 272.54 ND 2.21 3.77 IFLVE 2516 5.042 90.15 255.68 ND 2.22 3.59 IFLVE2517 4.734 90.19 270.54 ND 1.66 2.89 IFLVE2518 4.549 90.23 264.32 ND 1.66 4.29 IFLVE2519 4.381 90.47 267.82 ND 3.32 3.23 IFLVE2520 4.068 90.14 268.69 ND 2.77 3.77 IFLVE 2521 5.083 90.24 267.40 ND 1.39 3.06 IFLVE2522 4.878 90.26 260.14 ND 2.22 3.41 IFLVE 2563 3.851 90.48 277.96 ND 2.76 3.58

Mean 4.625 90.30 264.38 ND 2.01 3.46 SD 0.353 0.12 6.43 ND 0.59 0.36 CV (%) 7.634 0.13 2.43 ND 29.42 10.36

SD, standard deviation; CV, coefficient of variation; ND, none detected.

2.2. Data analyses

All values (means of duplicate determinations) were analysed statistically and coefficients of variation (CV) within the lines were calculated (Steel and Torrie, 1960). Mean values on the proximate and in vitro parameters in the randomly selected lines were compared using MSTATC software while the correlations be- tween the weight of 100 seeds, CP and the anti-quality factors were computed using SPSS.

3. Results

3.1. Dry matter (DM), seed weight and crude protein (CP) content

The weight of 100 seeds and the DM and CP contents of all the species are shown in Tables 1-6. The DM values of all the lines analysed were similar, and ranged from 90.26 _+ 0.16% in V. palaestina to 90.89 __ 0.32% in V. narbonensis. The weight of 100 seeds was least in V. palaestina (3.77_+0.28 g; Table 6) and highest in V. narbonensis (19.33 _+ 2.97 g; Table 3 ). The average weights of the seeds of V. sativa and V. ervilia were similar, and generally much lower than those

132 V.A. Aletor et al. / Animal Feed Science and Technology 47 (1994) 125-139

Table 6 Crude protein (CP), protein precipitable tannin (PPT), tannin measured as vanillin-HCl catechin equivalent (CE) and trypsin inhibitor activity (TIA) of seeds from advanced lines of V. palaestina

Selection Weight per 100 DM CP PPT CE TIA no. seeds (g) (%) (gkg - l DM) (gkg -~DM) (gkg -~ DM) (gkg -1DM)

IFLVP2523 3.586 90.45 348.70 ND 3.59 0.54 IFLVP2524 3.500 90.42 349.37 ND 3.32 0.36 IFLVP 2525 3.543 90.31 355.66 ND 3.60 0.47 IFLVP2526 3.312 90.29 329.60 ND 1.94 0.37 IFLVP2527 3.609 90.35 332.04 ND 2.21 0.51 IFLVP 2528 3.730 90,11 340.36 ND 3.61 0.68 IFLVP2529 4.050 90.43 334.29 ND 3.32 0.46 IFLVP 2530 4.150 90,39 350.04 ND 3.32 0.48 IFLVP2531 4.033 90,37 352.66 ND 3.32 0.45 IFLVP2532 4.170 90,45 345.94 ND 4.15 0.43 IFLYP2533 3.836 90.18 334.77 ND 3.33 0.51 IFLVP2534 4.310 90.24 346.63 ND 3.32 0.38 IFLVP2535 3.703 90.11 332.93 ND 2.50 0.47 IFLVP 2536 3.468 90.01 327.74 4.40 2.78 0.84 IFLVP2537 3.658 90.08 345.58 4.31 3.33 0.75 IFLVP2538 3.703 89.92 343.53 4.40 2.78 0.47

Mean 3.773 90.26 341.87 4.37 3.15 0.51 SD 0.280 0.16 8.55 0.04 0.55 0.13 CV (%) 7.427 0.18 2.50 1.02 17.48 25.86

SD, standard deviation; CV, coefficient of variation; ND, none detected.

of V. narbonensis. Intraspecies variabilities in seed weights were generally higher in V. narbonensis than in the other species as shown by their respective CVs. The species with the lowest and highest mean CP values were V. ervilia (264.38 _+ 6.43 g kg -~ DM; Table 3) and V. palaestina (341.87_+8.55 g kg -1 DM; Table 6), respectively. The CP content of V. sativa (Table 4) was similar to that of V. pa- laestina. The CV, an indicator of intraspecies variability in CP, was similar and small in all lines.

3.2. Protein precipitable tannin (PPT) and vanillin-HCl catechin equivalent (CE)

The PPT levels in K narbonensis lines (Tables I-3 ) were similar to those in K palaestina (Table 6). No PPT was detected in any line of K sativa or K ervilia. Most of the lines from the advance collection of K narbonensis and K palaestina had no detectable PPT levels. Flavonoid content (including condensed tannins), measured as CE, was no more than 8.27 g kg- ~ DM in all the Vicia species, and seven of the 91 K narbonensis lines had no detectable CE. All the lines from the other species contained CE with mean values ranging from 1.07_+0.51 g kg -~ DM in K narbonensis to 3.15 g kg-~ DM in K palaestina. There were large in- traspecies variations in CE content as show by the high CVs which ranged from 17.48% in V. palaestina to 68.99% in V. sativa.

V.A. A letor et al. / Animal Feed Science and Technology 47 (1994) 125-139 133

Table 7 Proximate and in vitro characteristics of Vicia seeds

DM CP OM Ash ADF NDF IVDMD IVOMD (gkg -1) (gkg -1) (gkg - l ) (gkg - t ) (gkg -~) (gkg - l ) (%) (%)

V. narbonensis (microplo 0 IFLVN 2379 907.20 285.16 968.00 3 5 . 2 1 1 3 7 . 5 4 1 8 7 . 0 6 88.98 85.79 IFLVN 2389 906.00 276.93 964.60 39.06 1 2 5 . 7 9 198.31 89.35 85.57 IFLVN 2395 908.10 285.87 967.50 35.80 97.83 1 9 8 . 2 0 88.54 85.27 IFLVN 2601 906.50 325.32 963.40 40.37 1 1 2 . 2 8 209.55 89.29 85.46 Mean 906.95a 293.32a 965.88a 37.61a 118.36b 198.28b 89.04a 85.52a SD 0.79 18.81 1.93 2.17 14.85 7.95 0.32 0.19

V. narbonensis (nursery) IFLVB 124 908.20 273.40 973.00 29.70 1 4 9 . 1 7 235.97 86.08 83.55 IFLVN 543 901.90 263.08 964.10 39.47 1 6 5 . 1 3 234.06 87.43 83.43 IFLVN 2702 907.80 272.75 965.60 37.86 1 7 6 . 0 0 296.75 85.92 82.48 IFLVN 2706 907.60 313.02 963.80 39.86 156.74 304.00 86.23 82.44 Mean 906.38a 280.56b 966.63a 36.72a 161.76a 267.70a 86.42b 82.98b SD 2.59 19.18 3.74 4.12 9.97 32.79 0.60 0.52

V. narbonensis (advance) IFLVN 2380 906.90 274.56 965.20 3 8 . 3 5 1 3 7 . 5 4 1 9 4 . 0 7 88.76 85.22 IFLVN 2462 911.50 277.56 967.80 35.34 1 0 7 . 2 3 1 4 9 . 6 0 89.22 86.48 IFLVN 2478 906.60 287.50 967.40 35.88 1 1 3 . 9 1 1 9 1 . 3 9 89.76 86.06 IFLVN2561 906.58a 296.82 970.60a 33.09 1 2 8 . 7 2 249.83 88.31 85.17 Mean 3.61 284.11a 1.70 35.67a 121.85b 196.22bc 89.01a 85.73a SD 8.76 1.87 11.94 35.62 0.54 0.56

V. sativa (multilocation) IFLVS 1448 907.40 307.25 967.60 35.75 95.00 1 3 6 . 6 0 83.53 79.94 IFLVS 2037 905.40 294.35 967.10 36.37 95.63 135.77 83.85 80.78 IFLVS 2108 905.50 322.92 966.80 36.68 75.68 1 4 7 . 8 3 86.69 83.72 IFLVS 2559 897.40 292.29 973.90 28.81 69.88 1 5 6 . 3 3 87.66 85.66 Mean 903.93a 304.20a 968.85a 34.40a 84.05c 144.13bc 85.43b 82.53abd SD 3.85 12.23 2.93 3.25 11.45 8.50 1.78 2.29

V. ervilia (microploO IFLVE 2509 902.70 259 .11 970.10 32.87 59.25 1 4 8 . 0 7 90.22 87.29 IFLVE 2519 914.70 267.82 972.60 29.97 55.14 1 5 6 . 4 3 91.92 89.31 IFLVE 2520 901.40 268.69 970.20 32.69 58.57 1 4 0 . 2 4 91.23 88.53 IFLVE 2521 902.40 267.40 971.70 31.08 54.14 1 1 3 . 2 2 90.44 88.05 Mean 905.30a 265.76d 971.15b 31.65a 56.78d 139.49c 90.95c 88.30c SD 5.45 3.86 1.05 1.20 2.18 16.21 0.67 0.73

V. palaestina (microplo 0 IFLVP 2523 904.50 348.70 965.90 37.35 89.80 1 8 1 . 0 3 81.48 78.09 IFLVP 2526 902.90 329.60 970.00 32.97 1 0 1 . 1 0 174 .51 82.00 79.74 IFLVP 2527 903.50 332.04 969.60 33.39 91.38 125.32 83.05 80.66 IFLVP 2532 914.50 345.94 967.90 35.00 85.16 120.71 81.40 78.40 Mean 906.35a 339.07c 968.35a 34.68a 91.86c 150.39c 81.98d 79.22d SD 4.74 8.35 1.62 1.72 5.80 27.52 0.66 1.04

Means followed by dissimilar letters in the same vertical column differ significantly (P< 0.05 ). SD, standard deviation.

134 V.A. Aletor et al. / Animal Feed Science and Technology 47 (1994) 125-139

Table 8

Corre la t ion coefficients ( r ) be tween weight o f seeds, c rude pro te in and an t inu t r i t iona l factors

P P T CE TIA

V. narbonensis Weight o f 100 seeds

n 91 91 91 r - 0 . 4 5 - 0 . 1 2 - 0 . 0 2 P * NS NS

Crude prote in

n 91 91 91 r - 0 . 2 1 0.01 0.17 P * NS NS

V. sativa Weight o f 100 seeds

n 23 23 23 r - - 0 . 6 5 - 0 . 3 1 P - *** NS

Crude prote in

n 23 23 23 r - - 0 . 0 4 - 0 . 4 8 P - NS *

V. ervilia Weight o f 100 seeds

n 16 16 16 r - - 0 . 4 1 - 0 . 3 4 P - NS NS

Crude pro te in

n 16 16 16 r - 0.32 - 0 . 0 5 P - NS NS

V. palaestina Weight o f 100 seeds

n 16 16 16 r - 0 . 2 0.55 - 0 . 2 7 P NS * NS

Crude pro te in

n 16 16 16 r - 0 . 1 6 0.65 - 0 . 2 8 P NS * NS

* P < 0.05; **P< 0.01; ***P< 0.001. n, n u m b e r o f observat ions ; NS, not significant .

3.3. Trypsin inhibitor activity (TIA)

All species contained low levels of TIA with mean values ranging from 0.51 + 0.13 g kg- 1 DM in V. palaestina (Table 6) to 4.29 + 0.50 g kg- 1 DM in V. sativa (Table 4 ). TIA values were highest in IFLVS 2560 (5.00 g kg- ~ DM) and

V.A. Aletor et al. / A nimal Feed Science and Technology 47 (1994) 125-139 135

least in IFLVP 2524 (0.36 g kg-~ DM). All the V. palaestina lines had low TIA, with values ranging from 0.36 g kg -~ DM in IFLVP 2524 to 0.84 g kg -~ DM in IFLVP 2536 with a CV of 25.86%. Mean TIA values in all lines of V. narbonensis, V. sativa and V. ervilia were generally similar, but substantially higher than those of V. palaestina (Table 6).

3.4. Proximate and in vitro studies

Some chemical and in vitro data on randomly chosen lines from each of the species are summarized in Table 7. The OM and ash contents of the samples were generally similar. However, the ADF and NDF values of V. narbonensis lines were significantly (P< 0.05) higher than those of the other species. The mean IDMD and IOMD values of the species also differed significantly (P< 0.05 ). The higher IDMD (90.95%) and IOMD (88.30%) values of F. ervilia appeared to parallel the lower ADF ( 56.78 g kg- ~ DM) and NDF ( 139.49 g kg- ~ DM) values.

3.5. Correlation analysis

Table 8 summarizes the correlation coefficients (r) between seed weight, CP and the different antinutrients. Seed weight was significantly and negatively cor- related with PPT only among the V. narbonensis lines (P< 0.05; r= - 0.45 ). Seed weight was also correlated significantly (P< 0.05 ) with CE in both V. sativa and V. palaestina. There were no significant correlations between seed weight and TIA in any of the species. Crude protein was positively correlated ( r= 0.05 ) with CE only in V. palestina lines. Similarly, there was a significant correlation (P< 0.05 ); r= -0 .48 ) between CP and TIA only among the V. sativa genotypes.

4. Discussion

It is well documented (Hove and King, 1978; Farrell and Vohra, 1983; Liener, 1989 ) that the nutritional importance of a given food/feedstuffin a diet depends not only on the nutrient composition of the raw feedstuff, but also on the amount that is consumed as well as the presence of antinutritional constituents. With regard to CP, the present data suggest that Vicia species could be important con- tributors of plant protein supplements to foods and feeds. For example, the mean CP contents of V. palaestina (341.87+8.55 g kg -~ DM) and V. sativa (317.36 + 18.68 g kg- 1 DM) compare favourably with those reported for the more conventional soya bean and lupins (Hove and King, 1978) and clearly surpass those reported for Pisum, Vicia faba, Ca janus ca jan and Phaseolus species by Hove and King (1978), Ologhobo (1980) and Aletor and Aladetimi (1989). Similarly, the CP contents of virtually all the V. narbonensis and V. ervilia lines were much higher than those commonly reported for lentil, Pisum and Vigna species. Given the desirability of the agronomic characteristics and yield poten- tial of Vicia, it is conceivable that it could play a valuable role as a supplemental

136 V.A. Aletor et al. / Animal Feed Science and Technology 47 (1994) 125-139

protein source to cereal straw and other low-nitrogen farm byproducts currently fed to ruminants in several parts of WANA. However, appropriate feeding re- gimes and the necessary processing technique (s), if any, remain to be established to ensure its optimal utilization. In this context, the data on the relative seed sizes could be of value, especially if cooking is envisaged as a mode of processing. This is because earlier studies with chick peas, lentils and faba beans (Williams et al., 1985; Eskine et al., 1985; Singh et al., 1988 ) have shown significant positive cor- relation between seed size and cooking time. Indeed, Singh et al. (1988) have suggested that selection for seed weight (an adequate predictor of cooking time) could eliminate labour-intensive tests for cooking time, and thus will enable leg- ume quality analysts to concentrate on the more critical aspects of legume quality such as protein digestibility, S-amino acid content and antinutrient constituents.

Although the immense contribution of grain legumes to dietary protein supply is well acknowledged, their ability to synthesize myriad antiphysiological factors remains a major drawback to their direct use as food by man and livestock. The present study shows that Vicia species contain tannins and trypsin inhibitors in varying levels. Unlike 1I. narbonensis, in which most of the lines analysed con- tained PPT and CE, it was of interest that no PPT was detected in any of the lines of 1I. sativa and 1I. ervilia, while only three of the 16 lines of II. palaestina con- tained detectable levels. The fact that many simple flavonoids (including care- chin) do not precipitate proteins (Hagerman and Butler, 1978 ) tends to suggest that, unlike V. narbonensis, most of the 1I. sativa, 1I. ervilia and I1. palaestina lines may contain predominantly low molecular weight phenolics. These, unlike the condensed tannins, may not be deleterious to animals unless consumed in large amounts. However, they generally contribute to poor palatability of rations (Menhansho et al., 1987 ).

Although direct comparison of analytical results on antinutritional factors is often made difficult by differences in analytical techniques, using similar analyt- ical methods Ologhobo (1980), Smith et al. (1980), Aletor and Ojo (1989) and Aletor et al. (unpublished data, 1992) reported much higher levels of TIA for Glycine max, Phaseolus, Vigna and Lathyrus species. Trypsin inhibitor levels in 1I. palaestina were particularly low, and were no more than residual levels found in cooked or roasted soya bean. Most legumes contain trypsin inhibitors whose inactivation by heat treatment improves the nutritional value. Consequently, trypsin inhibitor levels are an important quality index, especially for leguminous protein sources.

Data on the other chemical and in vitro parameters showed a significant spe- cies difference in ADF, NDF, IVDMD and IVOMD. Vicia ervilia generally had the highest IVDMD and IVOMD values, which seemed to parallel the lower CE and NDF values. In general, the ADF and in vitro digestibility data for the spe- cies were similar to those reported for the more conventional feed legumes in- cluding V. faba (Garfido et al., 1989). The in vitro data corroborate our earlier suggestion that Vicia seeds could be of great value as protein supplement to the vast amount of cereal straw and grains currently fed to animals in regions where they are primarily cultivated. However, wholesale incorporation into animal diets

V.A. Aletor et al. / Animal Feed Science and Technology 47 (1994) 125-139 137

must be avoided, and the potential toxicity of feeding seeds from species with high fl-cyano-L-alanine levels, especially to monogastric animals, recognized. Al- though sample size in the present study did not permit reliable correlations to be made between tannins and in vitro digestibilities, Buckley et al. ( 1983 ) and Gar- rido et al. ( 1989 ) reported significant negative correlations between dietary tan- nins in V. faba and other plant materials and in vitro digestibilities. The inhibi- tion of protein and carbohydrate degradation in vitro by tannins is believed to arise via a combination of several modes of action, including complexation with proteins, polysaccharides, bacterial cell membranes, or with enzymes involved in protein or carbohydrate degradation (Konishi et al., 1987; Muller et al., 1989 ).

The correlations between seed size, CP and the various antinutrients were gen- erally poor, suggesting that a selection programme based on seed size and protein may be difficult. With regard to tannins, a selection programme based on match- ing seed and flower colours is envisaged in our germplasm collection since seed and flower colours are known to be strongly correlated with tannin levels (Ca- brera and Martin, 1986; Garrido et al., 1989).

5. Conclusion

It is evident that the protein content and in vitro characteristics of these species of Vicia compare quite favourably with those of the more conventional and higher- priced grain legumes. Additionally, their content of antiproteinases, typified by trypsin inhibitors and tannins, appear to be much lower than found in the more conventional legumes. Clearly, much more information is needed on their con- tent of other antiphysiological agents, especially fl-cyanoalanine, and on simple treatments to ensure their safe use as dietary components for monogastrics and ruminants under feed-lot. Identification and development of cultivars with lower levels of these antinutritional factors, especially offl-cyano-L-alanine, should be a major focus of current breeding programmes. The need for more evaluatory tests on Vicia has become compelling in the wake of reported cases of purposeful adulteration, with Vicia, of some of the more conventional and higher-priced grain legumes.

Acknowledgements

We greatly appreciate the technical assistance of Adil E1 Awad, Ibrahim Said, Mohammed Heilani and other members of staff of the Forage and Legume Qual- ity Laboratories. Our thanks also go to Farouk Shamo for the statistical analyses.

References

Abd El Moneim, A.M., Cocks, P.S. and Sweedan, Y., 1988. Yield stability of selected forage vetches ( Vicia spp. ) under rainfed conditions in West Asia. J. Agric. Sci., 111: 295-301.

138 V.A. Aletor et al. / Animal Feed Science and Technology 47 (1994) 125-139

Abd El Moneim, A., 1992. Narbon vetch ( Vicia narbonensis): a potential feed legume crop for the dry areas in West Asia. J. Agron. Crop Sci., in press.

Aletor, V.A. and Aladetimi, O.O., 1989. Compositional evaluation of some cowpeas varieties and some under-utilized edible legumes in Nigeria. Nahrung, 33: 999-1007.

Aletor, V.A. and Ojo, O.I., 1989. Changes in differently processed soya bean (Glycine max) and lima bean (Phaseolus lunatus) with particular reference to their proximate, mineral and inherent anti- nutritional constituents. Nahrung, 33: 1009-1016.

Burns, R.E., 1971. Methods for evaluating tannin content in grain sorghum. Agron. J., 63:511-512. Buckley, K.E., Devlin, T.J. and Marquardt, R.R., 1983. Factors affecting in vitro rumen digestion of

faba bean cultivar. Can. J. Anim. Sci., 63: 89-96. Cabrera, A. and Martin, A., 1988. Variation in tannin content of Viciafaba L. J. Agric. Sci., 196: 337-

382. Cocks, P.S., Thomson, E.F., Sonnmel, K. and Abd E1 Moneim, E.M., 1986. Degradation and rehabil-

itation of agricultural lands in North Syria. ICARDA- 110 Ar, En Aleppo. International Center for Agricultural Research in the Dry Areas, Aleppo, Syria.

Eskine, W., Williams, P.C. and Nakkoul, H., 1985. Genetic and environmental variation in the seed size, protein yield and cooking quality of lentils. Field Crops Res., 12: 153-161.

Farrel, D.J. and Vohra, P., 1983. In: Recent Advances in Animal Nutrition in Australia. University of New England Publishing Unit, Amidale.

Garrido, A., Cabrera, A., Gomez, A. and Guerrero, J.E., 1989. Relationship between tannin content and in vitro nutritive value in seeds of 24 strains of V.faba L. In: J. Huisman, T.F.B. Van der Poel and I.E. Liener (Editors), Recent Advances of Research in Anti-nutritional Factors in Legume Seeds. Pudoc, Wageningen, pp. 156-159.

Goering, H.K. and Van Soest, P.J., 1970. Forage Fibre Analysis. Agric. Handbook No. 379. Agric. Research Service, USDA, Washington, DC.

Hagerman, A.E. and Buffer, L.G., 1978. Protein precipitation method for the quantitative determi- nation of tannins. J. Agric. Food Chem., 26: 809-812.

Hagerman, A.E., Robbins, C.T., Weerasuriya, Y., Williams, T.C. and McAuthur, C., 1992. Tannin chemistry in relation to protein digestibility. J. Range Manage., 45: 57-62.

Hove, E.L. and King, S., 1978. Composition, protein quality and toxins of seeds of grain legumes, Glycine max, Lupinus spp., Phaseolus spp., Pisum sativum and Vicia faba. NZ J. Agric. Res., 2 l: 457-462.

Konishi, K., Adachi, K., Kita, K. and Horikochi, I., 1987. Inhibitory effect of tannic acid on the respiratory chain of Photobacterium phosphoreum. Chem. Pharm. Bull., 35:1169-1175.

Liener, I.E., 1969. In: I.E. Liener (Editor), Toxic Constituents of Plant Foodstuffs. Academic Press, New York.

Liener, I.E., 1989. Anti-nutritional factors research: state-of-the-art. In: J. Huisman, T.F.B. Van der Poel and I.E. Liener (Editors), Recent Advances of Research in Anti-nutritional Factors in Leg- ume Seeds. Pudoc, Wageningen, pp. 6-13.

Menhansho, H., Butler, L.G. and Carlson, D.M., 1987. Dietary tannin and salivary proline-rich pro- teins: interaction, induction and defence mechanism. Annu. Rev. Nutr., 7: 423-430.

Ministry of Agriculture, Fisheries and Food, 1981. The analysis of agricultural materials. Tech. Bull. RB 427. HMSO, London.

Muller, H.M., Leimuller, E. and Rittner, U., 1989. Effect of tanniferous plant materials on protein and carbohydrate degradation in rumen fluid in vitro. In: J. Huisman, T.F.B. Van der Poel and I.E. Liener (Editors), Recent Advances of Research in Anti-nutritional Factors in Legume Seeds. Pudoc, Wageningen, pp. 156-159.

Ologhobo, A.D., 1980. Biochemical and nutritional studies ofcowpeas and lima bean with particular reference to some inherent anti-nutritional factors. Ph.D Thesis, University of Ibadan, Nigeria.

Price, M.L. and Butler, L.G., 1977. Rapid visual estimation and spectrophotometric determination of tannin content of sorghum. J. Agric. Food Chem., 25: 1268-1273.

Price, M.L., Scoyov, S.V. and Butler, L.G., 1978. A critical evaluation of the vanillin reaction as an assay for tannins in sorghum grains. J. Agric. Food Chem., 26: 1214-1219.

Singh, K.B., Erskine, W., Robertson, L.D., Nakkoul, H. and Williams, P.C., 1988. Influence of pre- treatment on cooking quality parameters in dry food legumes. J. Sci. Food Agric., 44:135-142.

V.A. Aletor et al. / Animal Feed Science and Technology 47 (1994) 125-139 139

Smith, C., Megen, V.M., Twaalfhoven, L. and Hitchcock, C., 1980. The determination of trypsin inhibitor levels in foodstuffs. J. Sci. Food Agric., 3 l: 341-350.

Steel, R.G. and Torrie, J.H., 1960. Principles and Procedures of Statistics. McGraw-Hill, New York. Tilley, J.M.A. and Terry, R.A., 1963. A two-stage technique for the in vitro digestion of forage crops.

J. Br. Grassl. Soc., 18:104-111. Williams, P.C., Singh, K.B. and Nakkoul, H., 1985. Relation of some aspects of Kabuli-type chickpeas

to cooking time. J. Sci. Food Agric., 34: 492-496.