Nutritive and Chemical Composition and In Vitro Digestibility of Cladodes of the Opuntia Species

Abstract

Sixteen cultivars (three resistant species and thirteen susceptible cultivars to cochineal insects) were included in the experiment to evaluate their nutritional and chemical compositions and to determine the in vitro digestibility of cladodes of the Opuntia species. Cultivars showed highly significant (p < 0.001) differences in their chemical composition. The cultivars’ content of ash, crude protein (CP), dry matter (DM), organic matter (OM), in vitro dry matter digestibility (IVDMD), in vitro organic matter digestibility (IVOMD), acid detergent fiber (ADF), neutral detergent fiber (NDF) and Cell (cellulose) varied, with ranges of 8.18–22.75%, 3.19–10.40%, 86.40–91.88%, 77.45–91.82%, 68.25–87.3%, 55.81–78.95%, 14.88–26.15%, 29.66–67.05% and 10.99–19.92%, respectively. The ash content showed a significant highly negative correlation with OM (r = −1.00; p < 0.001), IVDMD (r = −0.603; p < 0.001) and DOMD (r = −0.904; p < 0.001), respectively. The Garao, Aloqa and O. robusta var. X11 cultivars showed the highest similarities, as they are found close to the first branch of the dendrogram. In total, 90% of the variation among the nutritional traits was attributed to the first four PCs, with 55.97% to the first two principal components (PCs). The nutritional composition of the resistant species was greater than or equal to the susceptible ones.

1. Introduction

Prickly pear cacti (Opuntia species) comprise species with a wide climatic tolerance. They are able to proliferate in rainfall regimes of 250 to 1200 mm per annum, in very hot summers of over 40 °C and in cold winters with temperatures frequently falling below 0 °C for brief durations [1].

Opuntia ficus indica is commonly used as a fresh edible food ingredient and in products like jams, alcoholic beverages, natural liquid sweeteners and animal feed. The cladodes contain bioactive compounds such as fiber, minerals, flavonoids, phenolics and other nutrients. They also contain high amounts of fiber, including pectin, mucilage, lignin, cellulose and hemicelluloses, and these substances generally support the metabolism of lipids and sugars [2].

Cactus pears could play a stabilizing role in agriculture, as they can prevent stock losses during droughts, save natural grazing from over-grazing, increase farm income and alleviate poverty in rural areas. Although they have been considered poor in terms of crude fiber and crude protein, O. species are considered to be high in in vitro digestibility and in water content. They are often the only source of green forage in the dry season in arid and semi-arid climatic areas. Opuntia cladodes are regarded as a good, inexpensive source of energy, which may reduce the use of concentrated feeds and expensive fodder crops in dry areas [3]. They are a drought resistant fodder crop and have high biomass yield, palatability and soil adaptability [4]. Opuntia species are used as supplementary feed to sheep, dairy goats [5] and dairy cows [6].

In Tigray, northern Ethiopia, prickly pear covers an estimated area of 360,000 ha, which is about 6.74% of the total land of the region. It has a significant role in terms of economic and cultural value, for it is commonly used as a food and feed during food and feed shortage periods, bee forage and a source of cash income and short time employment, and it provides soil and water conservation services.

Cactus pear fruits are most commonly consumed by humans, while the fruit peels and the cladodes are fed to livestock. Young cladodes of some Opuntia species, including those of O. ficus indica and O. stricta, are consumed as nopal elsewhere. However, in Ethiopia, the consumption of cladodes as human food is rare, and the cladodes are commonly considered as livestock feed.

The cactus pear populations in northern Ethiopia display diverse morphological and other phenotypic variations. Some are revered for the sweet fruits they produce while others are mostly used for livestock feed. In the dry areas, particularly in the southern and eastern zones of Tigray, livestock feed is in short supply and the cladodes of the cactus pear have been used to fill the gap in feed shortage. However, the nutritional content of the cactus pear species used as livestock feed in Tigray has rarely been studied despite the more than a century old tradition of using the plant as livestock feed. Information on the entire nutritional and chemical composition of the cactus pear populations in Tigray is limited.

Nutritional value analyses are used to predict the performance of animals fed on a particular feed ration. Besides, new cactus pear cultivars have been introduced to the country with the objective of identifying cochineal-resistant lines. These lines are expected to be used as human food and livestock feed. These newly introduced cactus species have shown superior agronomic performance, including higher resistance to cochineal insects, but their nutritive value has not been adequately described. Hence, this experiment was conducted to determine the nutritional and chemical composition, as well as the in vitro digestibility, of cladodes of the Opuntia species.

2. Materials and Methods

2.1. Cladode Harvest and Preparation

Approximately one-year-old cladodes were manually harvested from randomly selected cactus pear plants from within the cactus pear collections in the Mekelle University main campus and other parts of the Tigray region, as described in

Table 1. Physical characteristics of Opuntia species.

Cultivars	Species	Spine	Location	Altitude (m)	Latitude (N)	Longitude (E)
Singin	O. ficus-indica	Spiny	Adigrat	2467	158,325	375,482
Halimbo	O. ficus-indica	Spiny	Adigrat	2471	158,324	375,482
Gabayle	O. ficus-indica	Spiny	Irob	2054	161,335	375,619
Silhuna	O. ficus-indica	Spineless	Irob	2034	161,366	375,618
Netcho	O. ficus-indica	Spiny	Irob	2040	161,365	375,618
Garao	O. ficus-indica	Spiny	Irob	2054	161,335	375,619
Aloqa	O. ficus-indica	Spiny	Aloqa/Adigrat	3004	157,509	375,442
Wukro-1	O. ficus-indica	Spiny, late maturing	Wukro	2023	152,580	375,632
Keyah/red	O. ficus-indica	Spiny	Wukro	2015	152,577	375,632
Lemats	O. ficus-indica	Spineless	Wukro	2013	152,576	375,631
Chewchawa	O. ficus-indica	Spiny	Mekhonee	1650	141,296	375,806
Tinkish	O. ficus-indica	Spiny	Mekhonee	1599	141,303	375,813
Wedwada	O. ficus-indica	Spiny	Mekhonee	1667	139,541	375,921
O. robusta var. X-11	O. robusta	Spineless
O. stricta var. wild	O. stricta	Spineless	-	-	-	-
O. stricta var. mexicana	O. stricta	Spineless	-	-	-	-

. Sixteen cultivars were included in the experiment (three cochineal resistant species and thirteen cochineal susceptible cultivars). Some of the cultivars are described in Figure 1. Samples were prepared by washing fresh cactus cladodes and removing the thorns manually. Cladodes were cut into pieces using a sharp knife.

Treatments were arranged in a complete randomized design (CRD) and were replicated three times. The samples were dried in a forced draft oven at 100 °C to constant mass; their moisture was determined, and the samples were milled using a laboratory mill with a one-millimeter sieve. The dried plant materials were stored in tightly sealed plastic bottl

2.2. Chemical Proximate Analyses

Dry matter (DM) content: The weight of the fresh as well as dried samples were recorded and the dry matter content was determined (AOAC, 2000). The following equation was used to determine the dry matter content (g/kg fresh material):

$$dry matter (gDM/kg wet weight) = \frac{weight \mathit{before} drying - weight \mathit{after} drying}{weight before drying} \times 100$$

Organic matter (OM) content: An amount of 2 g of milled cladode material was dried overnight at 100 °C and cooled in a desiccator, and the weight of the silica bowl and dried sample was recorded. Samples were incinerated at 550 °C for three hours. The weight of the ash contained in the silica bowls was recorded after cooling in a desiccator (fisher brand ^TM stainless steel desiccator with stainless shelves; Billings, MT, USA) for 20 min (AOAC, 2000). Organic matter (OM) was calculated by subtracting the percentage of ash from 100. The following equation was used to determine the OM content: OM (g/kg DM) = 100 − %ASH

2.3. Crude Protein Content and Fiber Contents

Crude protein (CP) content: Approximately 0.2 g of DM material was weighed into a foil cup and inserted into the Leco® Nitrogen analyzer (Leco® Corporation, St. Joseph, MI, USA, 2001). The total nitrogen (N) content was determined via combustion in oxygen, and a factor of 6.25 was used to convert N content into CP content. All CP determinations were performed with two replicates (Dumas, 1831).

Fiber content: Samples were analyzed for contents of different fiber fractions, i.e., neutral detergent fiber (NDF), acid detergent fiber (ADF) and acid detergent lignin (ADL). The acid detergent fiber (ADF), neutral detergent fiber (NDF) and acid detergent lignin (ADL) were determined using the method of [7]. Cellulose and hemicellulose were determined by differences from NDF, ADF and lignin using the methodology described by [7]. The formula used to predict total digestible nutrients (TDN) was 82.38–(0.7515 × ADF) as described by Bath and Mabele (1989).

2.4. In Vitro DM and OM Digestibility

In vitro digestibility of DM (IVDMD) analysis of opuntia species samples was carried out according to Tilley and Terry (1963). Ruminal fluid was collected from two steers fitted with ruminal cannula. The opuntia samples were placed in crystal tubes and incubated with a mix of McDougall saliva (1948) and ruminal fluid (4:l) for 72 h. Sample leftovers were recovered with Whatman 541 (12.5 cm diameter) filter paper and dried (105 °C, 24 h). Three tubes (sub-samples) were used for each sample. The DM degradation was evaluated in two runs. In vitro organic matter digestibility (IVOMD) was determined through methods suggested by ANKOM (2008).

2.5. Statistical Analyses

One-way analysis of variance (ANOVA) was used to test the effect of Opuntia cladode cultivars on nutritive value using the General Linear Model (GLM) procedure of the SPSS Software (IBM SPSS statistics 20). Least square means were used to test the significance of differences between treatment means (p < 0.05). Dependent variables that were found to be significantly different (p < 0.05) were further subjected to a multiple comparison test using Tukey’s test. In addition, Pearson’s correlation coefficient (r) was used to decide whether there was a linear relationship between IVDMD, IVOMD, CP, ash, NDF, ADL, ADF, TDN, SC, HC and Cell in the SPSS software (IBM SPSS statistics 20). A chemometric approach composed of principal component analysis (PCA) implemented in the SPSS software was used to analyze the nutrition content. The standardized data of each sample were then subjected to PCA. This way, the data could be reduced to a set of new latent variables called PCs. The loadings represented the relative contribution of each variable in the PCs, and the score values showed the location of each sample within a PC [8]. The relationship between the variables and samples was shown in PCA graphs. A hierarchical cluster analysis was performed with standardized data (Z scores) using Euclidean distances following the complete linkage method using Genstat 14th Edition.

3. Results and Discussion

Chemical composition

The opuntia species tested in the current study showed variations in chemical composition (p < 0.001) within and among the species. Ref. [9] also found similar results, where great variability in chemical composition was recorded in the cladodes of Opuntia spp. they tested [10]. Nevertheless, the chemical composition of cladodes depends on a number of factors such as the chemical characteristics of the soil, geographical location, environmental conditions, age and species [11,12,13].

3.1. Proximate Composition

The dry matter (DM) content showed a highly significant (p < 0.001) variation among species and cultivars (

Table 2. Dry matter (DM), ash, crude protein (CP), neutral detergent fiber (NDF), acid detergent fiber (ADF), acid detergent lignin (ADL), IVDMD = in vitro dry matter digestibility and IVOMD = in vitro organic matter digestibility of different Opuntia species.

Cultivars	DM	Ash	OM	CP	ADF	ADL	NDF	IVDMD	IVOMD
Singin	91.00 ± 0.00 g	19.44 ± 0.02 j	80.56 ± 0.02 e	6.25 ± 0.00 c	25.95 ± 0.05 j	6.24 ± 0.24 g	43.53 ± 0.01 ef	77.28 ± 0.09 f	62.90 ± 0.14 f
Halimbo	89.23 ± 0.01 cd	13.69 ± 0.01 c	86.16 ± 0.16 l	10.11 ± 0.11 ij	26.15 ± 0.15 j	6.21 ± 0.21 g	43.00 ± 0.20 e	83.15 ± 0.15 j	71.19 ± 0.11 k
Gabayle	89.08 ± 0.00 cd	14.50 ± 0.01 d	85.51 ± 0.01 k	8.94 ± 0.01 f	22.57 ± 0.09 i	5.65 ± 0.25 f	43.98 ± 0.02 fg	87.30 ± 0.20 l	74.15 ± 0.15 l
Slhuna	88.30 ± 0.10 b	21.81 ± 0.19 m	78.38 ± 0.01 b	9.38 ± 0.01 fg	17.63 ± 0.23 c	4.35 ± 0.02 d	60.90 ± 0.11 k	71.49 ± 0.01 b	55.81 ± 0.11 a
Netcho	89.35 ± 0.05 cd	8.18 ± 0.02 a	91.82 ± 0.02 n	8.01 ± 0.03 de	18.34 ± 0.06 d	4.48 ± 0.01 d	44.97 ± 0.01 h	85.85 ± 0.16 k	78.95 ± 0.05 m
Garao	87.09 ± 0.01 a	18.17 ± 0.02 g	81.83 ± 0.02 h	10.40 ± 0.40 j	18.74 ± 0.00 de	4.36 ± 0.01 d	39.29 ± 0.01 d	75.44 ± 0.06 e	62.53 ± 0.04 ef
Aloqa	89.35 ± 0.01 cd	18.68 ± 0.01 h	81.33 ± 0.01 g	9.94 ± 0.06 hi	16.33 ± 0.35 b	4.28 ± 0.01 cd	38.00 ± 0.10 c	82.90 ± 0.00 i	68.58 ± 0.02 ij
Wukro-1	90.70 ± 0.10 g	16.02 ± 0.00 e	83.98 ± 0.00 j	4.75 ± 0.07 b	19.67 ± 0.03 fg	5.12 ± 0.00 e	38.81 ± 0.02 d	78.38 ± 0.02 g	68.42 ± 0.01 hi
Keyah/red	91.88 ± 0.13 h	19.14 ± 0.02 i	80.86 ± 0.02 f	3.19 ± 0.11 a	19.29 ± 0.01 ef	5.44 ± 0.02 ef	49.05 ± 0.05 i	74.61 ± 0.01 d	62.43 ± 0.00 e
Lemats	91.02 ± 0.02 g	22.75 ± 0.25 n	77.45 ± 0.05 a	5.31 ± 0 E-7 b	19.32 ± 0.05 ef	5.41 ± 0.01 ef	44.59 ± 0.07 gh	79.27 ± 0.01 h	59.56 ± 0.04 d
Chewchawa	86.40 ± 0.40 a	14.47 ± 0.01 d	85.53 ± 0.01 k	9.57 ± 0.14 gh	21.85 ± 0.01 h	4.23 ± 0.01 cd	49.68 ± 0.02 j	82.83 ± 0.07 i	71.41 ± 0.02 k
Tinkish	88.94 ± 0.01 c	14.53 ± 0.01 d	85.47 ± 0.01 k	5.13 ± 0.01 b	20.29 ± 0.09 g	3.66 ± 0.02 ab	49.91 ± 0.19 j	75.52 ± 0.01 e	67.62 ± 0.02 g
Wedwada	87.85 ± 0.02 b	19.75 ± 0.02 k	80.25 ± 0.02 d	6.57 ± 0.05 c	21.53 ± 0.00 h	3.62 ± 0.01 ab	44.53 ± 0.03 gh	68.25 ± 0.01 a	57.67 ± 0.02 b
O. robusta var. X-11	89.97 ± 0.01 ef	21.52 ± 0.01 l	78.49 ± 0.01 c	8.38 ± 0.01 e	14.88 ± 0.01 a	3.23 ± 0.02 a	29.66 ± 0.03 a	74.07 ± 0.04 c	58.14 ± 0.01 c
O. stricta var. wild	90.50 ± 0.20 fg	16.87 ± 0.00 f	83.13 ± 0.00 i	7.75 ± 0.05 d	17.55 ± 0.05 c	3.88 ± 0.03 bc	67.05 ± 0.05 l	83.80 ± 0.20 j	68.90 ± 0.10 j
O. stricta var. mexicana	89.51 ± 0.00 de	12.77 ± 0.01 b	87.12 ± 0.12 m	6.03 ± 0.08 c	15.24 ± 0.06 a	4.25 ± 0.02 cd	37.21 ± 0.09 b	77.60 ± 0.00 f	68.05 ± 0.00 h
Mean	89.38 ± 0.26	17.02 ± 0.67	82.99 ± 0.67	7.48 ± 0.38	19.71 ± 0.57	4.65 ± 0.16	45.26 ± 1.57	78.61 ± 0.93	66.02 ± 1.13
p-value	p < 0.001	p < 0.001	p < 0.001	p < 0.001	p < 0.001	p < 0.001	p < 0.001	p < 0.001	p < 0.001

Data are means ± standard error. Values with the same letter in a column are not significantly different (p ≤ 0.001).

). The DM content of the cultivars tested ranged from 86.40 to 91.88% (Table 2). The DM content of the cultivars tested in the current study was relatively higher compared to earlier other findings [14,15]. Similarly, [5,16,17,18] reported 91.71–95.45% of DM contents in cladodes of cactus pear, which was comparable to the DM content of rice straw (91.88–92.21%) and almond trees (97.75–98%). Cultivar Keyih of the O. ficus-indica species had the highest DM (91.88%) content, which was significantly higher than the rest, O. stricta and O. robusta included. Cultivars Lemats, Singin and Wukro-1 had DM content statistically similar to that of O. stricta wild but significantly higher than the rest of the cultivars, including O. robusta and O. stricta var mexicana (Table 2). The DM contents of O. robusta and O. stricta wild were significantly higher than nine and eleven of the O. ficus indica cultivars, respectively. To the contrary, Ref. [19] reported DM for O. ficus-indica that was significantly (p < 0.05) higher than that of the O. Robusta they studied. The DM content of the O. robusta and O. stricta species was above the average value (Table 2).

The test cultivars showed significantly higher (p ≤ 0.001) variation in ash content (Table 2). The ash content ranged from 8.18 to 22.75% of the DM with an average of 17.02% (Table 2), whereas [20] reported an ash content of 15% in Brazilian species O. monacantha, which is relatively lower than in the current study. An ash content of 18.9% was found in cladodes of cactus pear by [21]. Moreover, the ash content reported by [22] was in the range of 13.7–29.3%. Ref. [4], on the other hand, reported comparable ash content of 16.7% in three varieties belonging to different cacti species. The variations in ash content may be due to geographical location, soil, climatic conditions, maturation and minerals [23,24].

In the present study, the ash content was the highest in cultivar Lemats followed by var. Silhuna, both spineless O. ficus indica types, followed by O. robusta, and the least in Netcho, a spiny O. ficus indica, followed by spineless O. stricta var mexicana (Table 2). Though [25,26] reported higher values of ash for the spineless forms compared to spiny types, the results in the current study showed inconsistent results, in which the top three were of the spineless type and the remaining of both forms. To the contrary, [2,27,28] reported the ash values of spiny cladodes higher than those of spineless cladodes.

The (organic matter) OM content in the current study ranged from 77.45 to 91.82% with an average of 82.99% and showed a highly significant (p ≤ 0.001) variation among varieties and species (Table 2). Significantly lower OM content was recorded in three spineless cultivars belonging to the O. ficus-indica type along with the O. robusta cultivar, while the majority of spiny O. ficus-indica had significantly higher OM content than the spineless O. ficuls indica cultivars (Table 2). Similarly, higher OM content of spineless cactus pears was reported by [28,29,30]. From among the 16 cultivars tested, 13 had OM content of ≥80% (Table 2). Comparable OM contents in cladodes belonging to different Opuntia varieties, ranging from 83.29% to 91.7%, were reported by several authors [31,32,33,34,35,36]. On the other hand, relatively lower OM content (68.3% to 79.85%) was reported by several other authors [37,38,39,40]. Whereas [26] reported that OM content was greater in O. robusta than in O. ficusindica, in the current study the O. robsuta cultivar had significantly lower OM content than the O. stricta cultivars as well as the majority of the O. ficus indica cultivars. Organic matter (OM) content was lower in spineless cultivars, including in O. robusta (Table 2). Nevertheless, variations in OM content could be attributed to species, varieties and the environment upon which a particular variety is grown. Generally, cochineal-resistant cultivars included in the current study (O. robusta and the O. stricta types) were not inferior in OM content to the rest of the cultivars tested.

3.2. Crude Protein Content (CP)

The crude protein (CP) content showed a highly significant difference (p ≤ 0.001) between variants (Table 2). Similarly, [30] also indicated a highly significant difference (p ≤ 0.001) among silages of opuntia. The crude protein content of different Opuntia cladodes varied between 3.19 and 10.40%, with an average of 7.48% (Table 2). This result was lower than earlier reported values ranging from 6.23 to 15.6% [2,30,34,39,40,41]. However, the average CP content reported in the current study (7.48%) is within the optimum range of 6.0–8.0 % DM for the growth of ruminal micro-organisms, and most of the cultivars tested (3/4) had CP values within or above the optimum range. The overall average crude protein content is close to the amount (7.7%) required for maintaining 40 kg goats [28]. Furthermore, the average crude protein content of the different Opuntia cladodes analyzed (Table 2) was higher than those reported by various authors, which were in the range of 3.80% to 6.3% [6,18,20,22,28,33,37,42,43]. Feed resources that contain less than seven percent CP may not adequately support microbial activity in the rumen, thus limiting the quantity of protein for the host animal [31]. The cultivar with the significantly lowest CP content was Keyah, followed by Wukro-1, Lemats and Tinkish (Table 2). In order to meet the maintenance and production requirements of livestock, feed sources with low CP contents (such as cladodes with low CP content) need to be balanced with supplementary protein, such as non-protein nitrogen [32,33].

There were cases where the CP content of O. ficus indica was found to be greater than O. stricta [34] and O. robusta [30]. On the other hand, the CP content of O. ficus indica was reported to be lower than that of O. robusta [9,44,45]. Considering the individual performance of the tested cultivars for CP content in the current study, the majority that had significantly higher CP content belong to the O. ficus indica types. However, it is difficult to generalize, as the representation of the different species in terms of numbers was not similar and the O. robusta and O. stricta types were under-represented. The values of crude protein showed significant variation among species and among cultivars (Table 2). The CP level in cladodes varies according to species and age [5,35]. In this study, however, the test cladodes collected from the mother plants were of the same age, and thus age differences may not be considered as a factor, though differences in genotype, edaphic factors and geographical location may have significant contributions.

3.3. Fiber Contents

The neutral detergent fiber (NDF) content showed a highly significant variation (p < 0.001) among the cultivars (Table 2). The NDF content ranged from 29.66 to 67.05% with an overall average of 45.26% (Table 2), which is comparable to the NDF content in O. ficus-indica cladodes of 46.6% [37]. On the other hand, the NDF contents recorded in the current study were higher than those reported by [21,38,39,40,46,47], the values of which were 25.1%, 25%, 22.8%, 22.8%, 39.7% and 21.9%, respectively. The highest NDF content of 67% belonged to the O. stricta wild variant, followed by Siluhna (61%), Chewchawa and Tinkish (50%) and Keyih (49%), while the least was recorded in O. robusta (29.66%), followed by O. stricta var mexicana (37%). The rest of the cultivars, all belonging to the O. ficus indica type, had NDF contents of 38% to 45%.

One of the nutritional aspects of cactus pears that must be met is their low NDF content [42]. According to [48], in order to avoid metabolic disorders, the NDF level recommended in diets for ruminants is 28%. Forages with high fiber contents have poor DMI due to their rumen fill effect, as well as low digestibility [41]. In addition, the higher NDF in forage indicates a higher lignin content, which in turn reduces digestibility and leads to less forage consumption [39,49]. On the other hand, low NDF (13.67–20.88%) content values from opuntia cladodes may indicate that they could not be regarded as a source of roughage [50]. Though the majority of the cultivars tested in the current study had high NDF content, which may make them relatively suitable as sources of roughage (Table 2), the O. robusta cultivar had the closest value (29.65%) to the recommended value of 28%, followed by that of the O. stricta var mexicana (37%). Both cultivars are resistant to the cochineal insect pest, and in particular, the latter is commonly used as livestock feed elsewhere (Dubeux, personal communication). On the other hand, the O. stricta wild, with the highest NDF content of 67%, may not be suitable as a ruminant feed per se. More or less similar NDF contents of 32.9% was obtained from O. robusta cladodes by [35].

Acid detergent fiber (ADF) content showed highly significant (p < 0.001) differences between species and among cultivars (Table 2). This result was in agreement with the findings of [39,43,51], who reported the significantly (p < 0.05) and highly significant (p < 0.001) influence of cultivars on the value of ADF content. The ADF values in the current study ranged from 14.88 to 26.15%, with an average value of 19.71% (Table 2). Refs. [14,22] reported an ADF content range of 14–25%. Significantly higher ADF contents were recorded in the O. ficus-indica cultivars Halimbo and Singin (Table 2). However, the least were recorded in O. robusta (14.88%) and O. stricta var mexicana (15.24%), followed by Aloqa (16.33%), Siluhna (17.63%) and O. stricta wild (17.55%); the latter two cultivars were not significantly different in ADF values from each other. The characteristic low-ADF concentrations can be associated with high non-structural carbohydrate concentrations, making the cactus pear a valuable source of fermentable carbohydrates for ruminants [52]. Thus, the cladodes of both O. robusta and O. stricta var mexicana would make a more suitable diet than the rest of the cultivars as they contain lower levels of ADF.

The acid detergent lignin (ADL) content ranged from 3.23 to 6.24% with an average of 4.65% (p < 0.001) (Table 2). Significantly higher ADL content was recorded in Singin (6.24%) and Halimbo (6.21%) than the rest of the cultivars, while significantly lower ADL values were obtained from O. robusta (3.23%), Wedwado (3.62%) and Tinkish (3.66%) (Table 2). In animal nutrition, the lignin plays role in depressing fiber fraction utilization in the ruminant digestive tract, and it affects animal digestibility [53]. Ref. [54] also increases as ADL increases, the content of indigestible neutral detergent fiber in grasses. So, O. robusta was the best in its nutritional value because of the low ADL content. In general, the ADL content of the cultivars under determination was considered a low amount.

Hemicellulose (HC) content of the cultivars showed highly significant (p < 0.001) variation (

Table 3. Total digestible nutrients (TDN), hemicellulose (HC), soluble carbohydrate (SC) and cellulose (Cell) of Opuntia species.

Cultivars	HC (%)	Cell (%)	SC (%)	TDN (%)
Singin	17.63 ± 0.01 b	19.45 ± 0.03 h	30.78 ± 0.01 e	62.92 ± 0.00 a
Halimbo	16.66 ± 0.16 b	19.92 ± 0.04 h	33.13 ± 0.18 g	62.59 ± 0.03 a
Gabayle	21.41 ± 0.07 de	16.92 ± 0.16 f	32.59 ± 0.02 fg	65.42 ± 0.07 b
Slhuna	43.41 ± 0.48 k	13.28 ± 0.21 c	7.96 ± 0.26 a	69.13 ± 0.17 h
Netcho	26.63 ± 0.05 h	13.86 ± 0.05 cd	38.84 ± 0.00 h	68.60 ± 0.05 g
Garao	20.55 ± 0.01 d	14.38 ± 0.01 de	32.04 ± 0.29 f	68.30 ± 0.00 fg
Aloqa	21.67 ± 0.25 e	12.05 ± 0.36 b	33.39 ± 0.17 g	70.11 ± 0.26 i
Wukro-1	19.14 ± 0.01 c	14.55 ± 0.03 e	40.42 ± 0.09 i	67.60 ± 0.02 de
Keyah/red	29.76 ± 0.04 j	13.85 ± 0.01 cd	28.62 ± 0.14 d	67.88 ± 0.01 ef
Lemats	25.27 ± 0.02 g	13.91 ± 0.04 cde	27.55 ± 0.12 c	67.86 ± 0.04 ef
Chewchawa	27.83 ± 0.03 i	17.63 ± 0.02 g	26.29 ± 0.11 b	65.96 ± 0.01 c
Tinkish	29.62 ± 0.28 j	16.63 ± 0.11 f	30.43 ± 0.21 e	67.13 ± 0.07 d
Wedwada	23.00 ± 0.03 f	17.91 ± 0.01 g	29.16 ± 0.04 d	66.20 ± 0.00 c
O. stricta var. wild	49.54 ± 0.06 l	13.68 ± 0.03 c	8.29 ± 0.06 a	69.19 ± 0.04 h
O. robusta var. X-11	14.78 ± 0.04 a	11.65 ± 0.03 ab	40.45 ± 0.04 i	71.20 ± 0.01 j
O. stricta var. mexicana	21.97 ± 0.15 e	10.99 ± 0.08 a	44.00 ± 0.01 j	70.93 ± 0.05 j
Mean	25.55 ± 1.63	15.04 ± 0.47	30.25 ± 1.74	67.56 ± 0.43
p-value	p < 0.001	p < 0.001	p < 0.001	p < 0.001

Data are means ± standard error. Values with the same letter in a column are not significantly different (p ≤ 0.001).

). The hemicellulose content ranged from 14.78 to 49.54% with an average of 25.55% (Table 3). The hemicellulose proportion of the NDF was 49.83–71.28% (Table 3). Ref. [22] reported that hemicellulose represents a high proportion of NDF. Lower hemicellulose contents were recorded in O. robusta, Singin and Halimbo (Table 3).

Cellulose is the other insoluble unavailable polysaccharide quantified in Opuntia species. The cellulose content of Opuntia cladodes varies from 10.99 to 19.92% (p < 0.001) (Table 3). Rodrigues et al. (2016) also reported a wide variability in the cellulose content of different O. ficus-indica populations. The average content of cellulose in the current study was 15.04% (Table 3). A similar result of 14.5% was reported by Calabrò et al. (2018). However, Refs. [37,41] reported higher cellulose content: 21.3% and 33.2%, respectively. The lowest cellulose content was found in O. stricta var. mexicana, followed by O. robusta (Table 3).

Total digestible nutrients (TDN) of Opuntia in the current study varied significantly (p ≤ 0.001) and were in the range of 62.59–71.20% (Table 3). The average TDN content in our study was 67.56% (Table 3). This is comparable to the reported 69.3% TDN content in cactus pear silage [45,55], 61% in cactus cladodes and 70.87% in O. ficusindica [56]. Ref. [57] indicated that the TDN content in cactus is higher than in any of the other roughages listed. This indicates that cactus can increase the digestible nutrients in the feed.

Soluble Carbohydrate (SC): One of the main characteristics of prickly pears is their relatively high soluble carbohydrate content, rendering them more fermentable by lactic acid bacteria [36,58]. Because of its high carbohydrate quality, spineless cactus can be used an emergency feed or as part of a complete diet, provided that the diet contains an adequate amount of degradable protein [39]. The high concentration of soluble carbohydrates in cactus facilitates the incorporation of nitrogen into microbial protein, which is the main source of metabolizable protein for the host animal, and is also an important feature concerning cattle nutrition, consistent with its common use as forage [57]. The SC content of the cultivars in the current study varied significantly (p ≤ 0.001); it ranged from 7.96 to 44.00% and had a mean value of 30.25% (Table 3). The high ruminal digestibility values may be attributed to the soluble carbohydrates present in the cactus, which are related to the nitrogen-free extract and low NDF content [39]. Significantly lower SC values were recorded in the O. stricta wild type and Slhuna, and higher values were found in O. stricta var. mexicana and O. robusta. This gives a good indication that the cultivars O. stricta var. wild and Slhuna are less fit for feed, while the latter two cultivars are more suitable to serve as source of diet for ruminants.

3.4. In Vitro-Dry Matter Digestibility (IVDMD)

The in vitro dry matter digestibility (IVDMD) content of O. species in this study showed highly significant (p < 0.001) differences among the cultivars and ranged from 68.25 to 87.3%, with an average of 78.61% (Table 2). IVDMD values in cactus pear cladodes ranging from 48.5 to 86.6% were similarly recorded by several authors) [19,22,39,59,60,61,62]. The high digestibility of all nutrients in treated groups [63] can be taken from the point that cacti contain minerals, which may be associated with improvement in the digestibility of all nutrients [26,36,64]. The IVDMD content in O. robusta was among the least performing O. ficus indica cultivars, while O. stricta var. mexicana was amongst the best performing cultivars (Table 2). However, Ref. [19] indicated that the IVDMD of O. robusta was generally higher than that of O. ficus-indica, and the O. robusta cladodes had more desirable nutritive traits than O. ficus-indica. Nevertheless, O. robusta in the current study is represented by a single cultivar and it is difficult to make such conclusions. However, both the O. robusta and O. stricta var. mexicana cultivars had IVDMD values high enough to make them worthy of using them as livestock feed. The O. stricta wild also showed a high IVDMD value, despite the limitations it had in terms of the overall digestibility and nutritive parameters (Table 2 and Table 3).

In the present study, IVDMD was highest in O. ficusindica var. Gabyle, followed by var. Netcho. Both cultivars had a high amount of OM and low amount of ash (Table 2). Most of the cultivars with high amounts of ash content showed low amounts of IVDMD (Table 2). This indicates that the amount of IVDMD increases with increases in OM content, and as the ash percentage decreases, the IVDMD increases (Table 2). As ADF and NDF lacked significant correlation with IVDMD, low IVDMD may result from the high amount of ash, and not from the ADF and NDF values (Table 2). Similarly, Ref. [19] also reported that low IVDMD values may be partly related to high ash content rather than ADF level. Ref. [19] stated that as the ash percentage increased, the IVDMD decreased. In addition, the comparison between nutritional composition and in vitro digestibility indicated that DM, NDF, ADF and ash were negatively correlated with the total digestibility [35]. A positive correlation was observed between CP and total digestibility of cladodes [35]. Ref. [61] also indicted that the IVDDM was greater (p < 0.05) for corn silage than cactus pear silage, probably due to its greater CP concentration. ADF is an important indicator of fodder digestibility and is negatively correlated with digestibility [41]. The reduction of ADF could lead to an increase in digestibility because ADF content in the feed has a negative correlation with digestibility [41]. The high degradability of the prickly pear cactus is due to the amount of nonstructural carbohydrates (on average 68%) and low lignin content [22]. Ref. [19] reported that the digestibility of most chemical constituents significantly increased in line with the low fiber content of Opuntia. On the other hand, Ref. [46] indicated that the reduced digestibility of Opuntia cladodes is a result of high fiber and low protein content, which negatively influence microbial proliferation in the rumen and thus digestibility. Digestibility of DM was positively influenced by the addition of cactus pear silage in a diet. These results might be associated with the low ADF and ADL concentrations [26,45]. High ruminal DM degradability of cacti is likely due to their high nonstructural carbohydrate and low lignin contents, which may explain the high degradation rate of the slowly degradable DM fraction [39]. Cactus cladodes have high digestibility, similar to good-quality forage [46].

Spineless cacti are favored for feeding to animals because spineless cacti can be easily consumed by animals as they lack spines [65]. Spineless cacti also have a high acceptability for ruminants and large quantities can be voluntarily consumed [66]. The IVDMD in 90-day-old spineless cactus cladodes was reported to be 66.7% [35]. This, however, is lower than the IVDMD of spineless cultivars/species in the current study.

In Vitro Organic Matter Digestibility (IVOMD)

Digestible organic matter is important for ruminal microbial protein synthesis as an energy source [55,63]. In the current study, the values of IVOMD showed a highly significant (p < 0.001) variation among the species and among cultivars. The values of IVOMD ranged from 55.81 to 78.95%, with an average of 66.02% (Table 2). Similarly, Refs. [17,40] also reported Organic matter digestibility (OMD) as 66.6% in O. ficusindica and 59.41–60.54% in rice straw, respectively. Clones also showed IVOMD content of 69.5–82% in Opuntia forage [67]. However, the IVOMD values in the current study were lower than those reported by [55] of 73.37–83.2% and [65] of 79.7%. Nevertheless, the IVOMD values are higher than those reported for rice straw, as indicated above. The highest IVOMD was recorded in O. ficusindica var. Netcho, a variety with the highest OM (Table 2). Therefore, the IVOMD may be influenced by the high availability of OM.

3.5. Pearson’s Correlation Coefficient between Parameters

The correlations between the tested cultivars for their chemical composition are presented here below in

Table 4. Pearson’s correlation coefficients (r) among nutritional characteristics of 16 cultivars of Opuntia species.

	DM	Ash	OM	CP	ADF	ADL	NDF	IVDMD	IVOMD	HC	Cell	SC	TDN
DM	1	0.201	−0.203	−0.662 **	−0.056	0.408 *	−0.026	0.064	−0.078	−0.005	−0.213	0.093	0.057
Ash		1	−1.000 **	−0.110	−0.133	−0.056	0.023	−0.603 **	−0.904 **	0.072	−0.146	−0.381 *	0.136
OM			1	0.109	0.130	0.056	−0.015	0.606 **	0.905 **	−0.063	0.142	0.374 *	−0.133
CP				1	0.031	−0.119	−0.020	0.330	0.194	−0.032	0.081	−0.162	−0.036
ADF					1	0.687 **	0.096	0.167	0.185	−0.263	0.968 **	−0.043	−1.000 **
ADL						1	−0.045	0.347	0.204	−0.288	0.484 **	0.089	−0.687 **
NDF							1	0.062	0.005	0.935 **	0.134	−0.906 **	−0.095
IVDMD								1	0.864 **	−0.001	0.086	0.104	−0.170
IVOMD									1	−0.063	0.157	0.301	−0.187
HC										1	−.216	−0.863 **	0.265
Cell											1	−0.083	−0.968 **
SC												1	0.042
TDN													1

Asterisks indicate significance at * p < 0.05 and at ** p < 0.001.

. There was a significant positive correlation (r = 0.408; p < 0.05) between DM and ADL, whereas there was a highly significant negative correlation (r = −0.66; p < 0.01) between DM and CP. Ref. [67] also indicated a highly negative correlation between CP and DM content. The ash content showed a significant negative correlation with OM (r = −1.00; p < 0.001), IVDMD (r = −0.603; p < 0.001), IVOMD (r = −0.904; p < 0.001) and SC (r = −0.38; p < 0.05), respectively. As ash (r = −0.87, p < 0.002), NDF (r = −0.44, p < 0.02) and ADF (p <0.001) percentage increased, the IVDMD decreased [19,35]. In the current study, no correlation was observed between CP and digestibility (Table 4). However, a positive correlation (p <0.05 to p <0.001) was reported between the CP and total digestibility of cladodes by [35]. OM had significant (r = 0.606; p < 0.001), (r = 0.905; p < 0.001), (r = 0.37; p < 0.05) correlation with IVDMD, IVOMD and SC, respectively (Table 4). So, this indicates that as the OM content increases, the nutritional value of the cultivars increases because of the reduction in ash content (Table 4). NDF showed a highly significant positive correlation (r = 0.935; p < 0.001) with HC. ADF also showed a significant positive correlation (r = 0.687; p < 0.001) with ADL and cellulose (r = 0.968; p < 0.001) and a highly significant negative correlation (r = −1.00; p < 0.001) with TDN. In addition, ADL showed a significant positive correlation with cellulose (r = 0.484; p < 0.001) and a significant negative correlation with TDN (r = −0.69; p < 0.001). IVDMD also showed a significant positive correlation (r = 0.864; p < 0.001) with IVOMD. In addition, HC and Cell also showed highly significant negative correlations with SC (r = −86; p < 0.001) and TDN (r = −0.97; p < 0.001), respectively (Table 4). Similarly, Ref. [54] also reported that there were high correlations of (p < 0.05) ADF with cellulose, ADL with ADF and NDF with HC. In [68], the authors also indicated that there was a positive correlation between ash, NDF and ADF.

3.6. Multivariate Analyses

3.6.1. Hierarchical Cluster Analysis

The hierarchical cluster analysis highlighted an overall pattern of genetic diversity and the relationship between cultivars. According to the dendrogram obtained by the UPGMA hierarchical grouping method, the highest variability was observed between Singin, Qeyah, Lemats, Wukro-1, Tinkish, Wedwada and other cultivars (Figure 2). Ref. [69] also stated that the highest similarity was found between the genotypes which are located at the first branch of the dendrogram. In this regard, cultivars like Garao, Aloqa and O. robusta var. X11 showed the highest similarity, as they are found close to the first branch of the dendrogram. The genotypes also showed a similarity of 65–98% and clustered into two main groups and five branches: GI (Singin, Qeyah, Lemats, Wukro-1, Tinkish, Wedwada, Halimbo, Gabayle and Chewchawa) and GII (Slhuna, O. stricta var. wild, Netcho, O. stricta var. mexicana, Garao, Aloqa, O. robusta var. X11) (Figure 2). In the current study, clustering was done based on the nutritional characteristics of the cultivars. The fact that the cochineal-resistant species/cultivars (O. stricta wild, O. stricta var mexicana and O. robusta) were assigned into different branches in the dendrogram shows that they do not have much to share in terms of their nutritive values; rather, they were closely related to the susceptible types despite being of different species. Ref. [52] also indicated that resistance factors did not influence the nutritional value of genotypes, and cactus genotypes which are resistant to the carmine cochineal showed nutritional characteristics similar to or better than traditionally used susceptible cactus genotypes. From the perspective of geographical origin (Table 1), the dendrogram indicated that the clustering was not influenced by the geographical area they were collected from. In [70], the authors also indicated that the distribution of populations in the dendrogram showed that geographical origin was unrelated to the clustering pattern.

3.6.2. Principal Component Analysis (PCA)

PCA was employed to reduce the data dimensionally and to represent the experimental data in a simpler way; Bartlett’s test of sphericity was statistically significant (p ≤ 0.001) (

Table 5. KMO and Bartlett’s test to check the effectiveness of implementing principal component analysis on the data of chemical composition.

		Chemical Composition
KMO Measure of Sampling Adequacy		0.568
Approx. Chi-Square		1339.446
Bartlett’s Test of Sphericity	df	78
	Sig.	<0.001

). Based on the eigenvalues greater than 1, the results showed that 90% of the variation among the nutritional traits was attributed to the first four principle components (

Table 6. Eigenvalues of the covariance matrix using sixteen Opuntia cultivars and thirteen variables.

	Eigenvalues	% of Variance	Cumulative %
Principal Component 1	4.237	32.589	32.589
Principal Component 2	3.039	23.379	55.968
Principal Component 3	2.717	20.901	76.869
Principal Component 4	1.741	13.390	90.259
Principal Component 5	0.917	7.051	97.310
Principal Component 6	0.223	1.713	99.023
Principal Component 7	0.114	0.880	99.903
Principal Component 8	0.012	0.090	99.993
Principal Component 9	0.000	0.003	99.996
Principal Component 10	0.000	0.003	99.999
Principal Component 11	0.000	0.001	100.000
Principal Component 12	0.000	0.000	100.000
Principal Component 13	0.000	0.000	100.000

), and each of the remaining components contributed only seven percent or less. This indicates that the four components could provide a good summary of the dataset. However, [71,72] stated that the first three principal components contributed to 80% and 94% of the variations, respectively.

The traits most heavily positively loading to PCA1 were DOMD, OM, IVDMD, ADL and SC, whereas ash and TDN associated negatively. PCA2 was highly positively associated with ADF, Cell and ash content and negatively loaded to TDN, SC and OM in large amounts. PCA3 explained 20.9% of the variation and was positively associated with NDF, HC and CP and negatively associated with SC, DM and ash, respectively (

Table 7. Eigenvectors of the principal components of the sixteen Opuntia cultivars and thirteen variables.

	Principal Component	Principal Component	Principal Component	Principal Component
	1	2	3	4
DOMD	0.792	−0.462	0.327	0.173
Ash	−0.737	0.522	−0.289	−0.081
OM	0.735	−0.523	0.296	0.084
IVDMD	0.663	−0.309	0.344	0.221
ADL	0.565	0.490	−0.230	0.381
ADF	0.693	0.708	−0.072	−0.075
TDN	−0.696	−0.705	0.071	0.077
Cell	0.645	0.681	−0.003	−0.224
NDF	−0.160	0.395	0.872	0.215
HC	−0.404	0.131	0.870	0.235
SC	0.387	−0.529	−0.745	0.003
DM	−0.118	0.111	−0.297	0.892
CP	0.184	−0.128	0.312	−0.738

). The fourth PCA contributed 13.39% of variation and was positively associated with DM and TDN, in addition to other small contributors, and negatively associated with CP and Cell (Table 7). In [67], it was indicated that NDF showed among the highest factor loading, explaining 17% of the variability through the third principal component. The most important characteristics are those with weighting coefficients (eigenvectors) of greater magnitude, in absolute value, in the first principal components; these would be the most responsive characteristics in the selection process among cactus pear cultivars. Two principal components out of the thirteen components contribute no value to the total variance (Table 6). Such characteristics contributed little to the divergence among the genotypes and could be neglected [73].

The two most variation-contributing principal components were able to describe 55.97% of the total variability (Table 6 and Figure 3). Similarly, [8,74,75] also indicated that principal components (PC) 1 and 2 explained 53%, 54% and 50.5% of total variance of the data, respectively. However, ref. [76] reported that the two principal components explained 69.6% of the total variance in the case of the whole dataset. The first PC accounted for 32.59% and the second PC accounted for 23.38% of the total variance (Table 6 and Figure 3). Ref. [8] also indicated that the first PC (PC1) represented 39.71% of the total variance, whereas PC2 accounted for 13.41% of the total variance. In addition, [77] also indicated that the first PC (PC1) represented 39.3% of the total variance and PC2 accounted for 22% of the total variance.

4. Conclusions and Recommendations

Chemical composition showed variations among cultivars and species. Results showed that the cultivars had high DM, moderate ash, high OM and CP content and, in many cases, dry matter digestibility higher than 80%. This high dry matter digestibility of cladodes of prickly pear is useful for easy utilization and would eventually tell us its nutritional importance as an animal feed. Cultivars with higher ash contents were lower in their IVDMD content. As ADF and NDF lacked significance to the correlation of IVDMD, the low IVDMD may be from the high amount of ash and not from the ADF or NDF value. This indicates that the IVDMD was determined more significantly by the ash content of cultivars. Cultivars with higher OM contents were also higher in their IVOMD content. This indicates that the IVOMD was influenced by the availability of the organic matter content of cultivars. Based on the eigenvalues greater than 1, the results showed that 90% of the variation among the nutritional traits was attributed to the first four principal components (PCs), and each of the remaining components contributed only seven percent or less. This indicates that the four components could provide a good summary of the dataset. From among the cochineal-resistant cultivars, the O. robusta and O. stricta var Mexicana possess qualities that makes them acceptable for livestock feed. So, in vivo studies should be performed in the future to determine animal responses to these cactus species in order to better assess their fodder potential.