The fodder grass resources for ruminants: A indigenous treasure of local communities of Thal desert Punjab, Pakistan

Indigenous people have been using local grasses for rearing their animals for centuries. The present study is the first record of traditional knowledge of grasses and livestock feeding system from the Thal desert in Pakistan. A snowball sampling method was used to identify key participants. Information was collected from the respondents from six districts of Thal Desert through semi-structural questionnaire and site visits. The data was analyzed through Smith’s salience index and Composite Salience using ANTHROPAC package in R software. On the whole 61 grasses were recorded from the study area: most of them belong to the Poaceae family (52 species). Based on palatability grasses were categorized into three major groups i.e. (A) High priority, (B) Medium priority and (C) Low priority. Species in Group A, abundantly present in the study area represent a source of highly palatable forage for all ruminants. 232 (141M +91W) local participants were interviewed. Participants were grouped into three major age categories: 20–35 (48 participants), 36–50 (116 participants) and 51–67 years old (68 participants). ANTHROPAC frequency analysis confirmed the Smith’s salience index and Composite Salience; Cynodon dactylon was the favorite species (6.46 SI, 0.6460 CS) followed by Cymbopogon jwarancusa (5.133 SI, 0.5133 CS) and Sorghum sp. was the third most salient species (5.121 SI, 0.5121 CS). Grasses were mostly available during the months of August and October and had also ethnoveterinary importance. This document about the traditional feeding of livestock in Thal Desert can underline the importance of conserving a traditional knowledge, which was poorly documented before.


Background
In rural areas of Pakistan, agro-pastoral activities play a crucial role in the development of the local economy, accounting for more than half of the total agricultural income and 10.6% of the national GDP [1]. These activities are particularly important in the economy of the country's desert regions where land cultivation is difficult and livestock husbandry is the main and often unique survival strategy and income source for the local communities. Moreover, milk and meat production may counteract the impact of climatic unpredictability on fluctuations in food availability, especially in areas facing frequent crop shortages. According to data reported by Farooq et al. [2], in Pakistan 8.1% of buffaloes, 13.5% of cattle, 15.3% of sheep and 14.4% of goats are raised in desert districts. However, husbandry in these areas is often an uncertain and low-paid activity; shortage of fodder as a result of severe climatic conditions, high rate of diseases, limited availability of veterinary services and poor access to animal vaccination are important constraints limiting the local livestock productivity [2]. The sustainable production of livestock under harsh climatic conditions needs efficient strategies for improving fodder utilization and management [3]. From this perspective, traditional knowledge can be an important source of information on local wild forage resources and on their nutritive properties. Several studies have shown that smallholder farmers in many parts of the world have a deep practical knowledge about the importance and quality of plants used to feed animals. Ethnobotanical investigations on fodder plants have been carried out in Africa [4][5][6], Brazil [7], India [8,9] and China [10][11][12]. Many studies throughout the world highlight the diverse and abundant use of grasses and sedges as fodder; grasses and sedges are generally reported to be palatable and highly productive resources and to have high forage potential especially in arid and semiarid areas [12,13].
Previous studies have shown that Thal is rich in grasses and sedges [14]; most of the grasses are used by local population as fodder [10,13,15]. However, no detailed study carried out to analyze utilization and selection strategies of these plants by shepherds and farmers living in this zone. Extensive areas in the Thal have been overgrazed and are now strongly threatened by desertification [16,17]. Understanding the relative importance and preference of different species is crucial for a sustainable management of the local forage resources and can help animal husbandry technicians to optimize the selection of useful fodder species and to improve the livestock system efficiency. Moreover, recording this knowledge would be a much faster and cheaper method for learning about palatability and nutritive value of these plants.
The major aims of this study were: (1) To document traditional knowledge about the use of grasses and sedges as fodder in Thal and to assess similarities and differences with the studies previously conducted in the same [15] and in neighboring areas [11,12]. (2) To evaluate the impact of socioeconomic factors on the local ethnobotanical knowledge. (3) To rank, by order of preference, the different species used in the animal diet. (4) To quantify the influence of seasonal variation on the availability of these plants as animal feed.

Description of the study area
The Thal desert is located between 31˚10' N and 71˚30' E in the Punjab province, Pakistan (Fig 1). It is a subtropical sandy desert lying between the Indus River flood plains in the west and Jhelum and Chenab River flood plains in the east. About 50% of the Thal is under arid to hyper-arid climatic conditions (mean annual rainfall less than 200 mm) and the remaining half is characterized by semiarid climatic conditions (annual mean rainfall between 200 and 500 mm). Most of rainfall occurs between June and August. Average temperatures range between 3-8˚C in winter and 32-40˚C in summer. Wind erosion is a serious problem leading to the loss of topsoil and organic matter and damage to crop plants.
In Thal desert livestock is considered as a secure source of income for small farmers and the landless poor. According to Husain [18] the average herd size is 17 standard animal units. Livestock herds consist of animals of different age and sex; on average each farm has 22.8 goats, 16.7 sheep, 7 cattle, 2.51 buffaloes, 0.88 camels, 0.21 donkeys and 0.05 mules. Detailed information on grazing and stall feeding practiced in the area is given in Faraz et al. [19].

Ethnobotanical survey
The ethics committee/IRB of Pir Mehr Ali Shah Arid Agriculture University, Rawalpindi approved this study. Formal ethical consent was also obtained from all participants before the research started. Data were collected for two consecutive years (from March 2016 to March

PLOS ONE
The fodder grass resources for ruminants 2018), twice a year from each of the Thal desert six districts. Participants were selected by snowball-sampling technique [20] among village leaders, shepherds and both farm and domestic livestock caretakers. Interviews were carried out complying with the ethics guidelines commonly followed in ethnobotanical studies [21,22]. Information was gathered by using different approaches i.e. group discussions with participants, individual semi-structured questionnaires and participant observation (Fig 2) [23,24]. The questionnaires were drafted in the local language (Seriki and Punjabi) and included the following major questions: (i) Which grasses/sedges are used as fodder? (ii) Which grasses/sedges are preferred as feed for cattle, sheep, camels, buffaloes, and goats? (iii) What is the palatability of the different used plants? (iv) Which plant part do animals consume? (v) What are the feeding habits of different animals? (vi) Which livestock feeding system does local people adopt: free grazing or cut and carry? (vii) Do the listed fodder plants have any ethnoveterinary use? (viii) What are their other indigenous uses?
In the second stage of the field research we used direct observation of livestock grazing habits to evaluate the palatability of different plants, animal preferences and the growth stages of plants at the time of grazing.

Collection and identification of plants
Plant collection was performed with the help of local participants during the field survey. Identification of the gathered species was carried out by the herbarium specialist Dr. Mushtaq Ahmed from Quaid-i-Azam University, Islamabad and by the taxonomist Dr. Humaira Shaheen (Fig 3). Botanical nomenclature of species and families complies with online Flora of Pakistan (http://www.efloras.org/flora_page.aspx?flora_id=5) [24] and the herbarium specimens were kept in the Botany Department of Pir Mehr Ali Shah University of Arid Agriculture.

Data analysis
The most common method to measure relative abundance was visual assessment and observation of ethnobotanically important grasses in the study area [12]. Total study area was almost 20,000 square kilometers. We randomly divided each district into 45-50 plots and plot size was (10X10m = 100m 2 ). Results were constructed by percentage of relative abundance through the following formula:

RA ¼
Total percentage cover of species over all plots Number of plots estimated � 100 Based on the abundance value, grasses were categorized into the following groups i.e. abundant, common, frequent, occasional and rare (Table 1).
Relative frequency of citation (RFC) was calculated to sort listed plants by priority order, using the following formula [12, [24][25][26]: Where fc is the number of participants that mentioned the fodder use of the species and "n" is the total number of participants included in the study.
Pairwise comparison (PWC) was also used to determine the priority order of the listed species [12,27]. Ten participants (5 key participants and 5 randomly selected) were chosen for the PWC. The participants were asked, one at a time, to select their preferred fodder plants from all possible pairs of species. Each species got a score of 1 if the participants selected it. Adding the scores and ranking them to obtained the final score.
Smith's salience index and Composite Salience [28] were used to evaluate species saliency by weighing the average of the inverse rank of a species across multiple free-lists where each list was weighed by the number of species in the list. ANTHROPAC [28] was used to generate Smith's salience indexes.
Pairwise ranking or comparison was used to evaluate the degree of preference or levels of importance. The values for use reports across the selected species were summed up and ranked. Ten participants (six key and four randomly taken participants) in the study area ranked grasses according to their use e.g. 1st, 2nd, 3rd, 4th and 5 th respectively. Ranking can be used for evaluating the degree of preference or level of importance of selected plants [28][29][30].
Respondent Consensus Factor (Fic): The Respondent consensus factor was derived in order to seek the importance of species used as fodder, Forage, Mixed feed and veterinary uses [31].
Where Nur is the number of use-reports in each disease category; Nt is number of species used.

Socioeconomic factors
In total, 232 local participants were interviewed ( Table 2); 141 were men and 91 were women. A smaller number of female participants were expected and this can be partially explained with the local cultural restrictions preventing women from working outside their homes or farms. Participants were grouped into three major age categories: 20-35 (48 participants), 36-50 (116 participants) and 51-67 years old (68 participants). With regard to the profession, 34% (36 women and 44 men) were shepherds, 26% (27 women and 33 men) were farmed livestock caretakers and 40% (28 women and 64 men) domestic livestock caretakers. Thirty-six (16%) of the interviewed people were illiterate, 24 (10%) never completed their primary education, 120 (52%) completed 5 years of primary school and 52 (22%) participants had middle education level (Fig 4) [24].

Use of fodder species
The participants reported the use of 61 plant species that were distributed into 40 genera and 3 botanical families. The most represented genus was Cyperus with 5 species, followed by Cenchrus and Eragrostis with 4 species each. Most species belonged to Poaceae family (51 species; 84% of the reported plants) while 8 species (13%) were categorized into Cyperaceae family.
Typhaceae were represented by only one species: Typha elephantina. Fifty-five species (92% of the reported species) were classified as native and 5 (8%) as exotic. The following exotic species were reported by participants: Chloris gayana, Imperata cylindrica, Paspalum dilatatum, Sorghum bicolor and Vetiveria zizanioides. These results seem to reflect composition and distribution patterns of the local flora. In a floristic checklist of Thal desert, Shaheen et al. [14] observed that Poaceae was the main family with 52 species. Of the 52 Poaceae naturally occurring in the area, 48 (94%) were reported to be used as fodder in our study; 5 were not cited by participants and 4 (Brachiaria reptans, Eragrostis atrovirens, E. cilianensis, Themeda anathera) were reported in our study but not in the floristic inventory. All the eight Cyperaceae cited were included in the study conducted by Shaheen et al. [14].
Our comparative analysis revealed that 15 species are used as fodder in all the considered studies. We found a mean similarity (Jaccard index) rather high (36.4 ± 6.9) with values ranging from 30.8 (this study vs [11]) to 50.0 ([12] vs [11]). These studies were all conducted in zones lying in the proximity of the study area that share not only similar ecological factors but also the same socioeconomic and cultural history. Nevertheless, our study listed 20 grasses not previously reported in the fodder category for this area. These results provide an important new contribution to the knowledge on wild fodder plants in Pakistan. At the same time, they also show the importance of collecting new ethnobotanical information even in already studied areas.

Socioeconomic factors
Participants mentioned 8.27 ± 4.49 taxa (range 1-18). Gender (H = 0.373; P > 0.05) and education (H = 5.29; P> 0.05) had no influence on the knowledge of fodder plants. Gender influence on traditional knowledge is controversial [32] and many studies have showed that the statistical strength of this relation depends on the local cultural context and on the categories of use that the researchers focus on. A lack of differentiation between men and women, as observed in this study, could mean that there is not a clear division of labor in the area. A similar finding was observed by Aumeeruddy et al. [32] in Northern Pakistan, where women have a detailed knowledge on characteristics and properties of the different fodder species, suggesting that they fully share with men the responsibility of livestock rearing and forage collection. Khan and Khan [33] observed that most of the women of Cholistan desert have an important

PLOS ONE
The fodder grass resources for ruminants role in managing livestock, spending almost 8 to 13 hours a day in this activity. Differently Nunes et al. [7] and Bruschi et al. [6] showed that men prevail in the knowledge about fodder plants. The greater male knowledge found in these two studies may be explained by different gender-based experiences and skills: men spend much of their time moving with their herds while women are more frequently involved in managing food and family care. The age of participants resulted to be statistically significant (H = 9.97; P < 0.05). As also shown in many other ethnobotanical studies [34][35][36] elderly people seem to retain more traditional knowledge on the use of plants. For young people (25-35 years old), the average number of known fodder plants was 6.65 ± 4.12 while for middle-aged (36-50) and elderly participants (> 50) there was an average number of 8.25 ± 4.13 and 9.42 ± 4.74, respectively. Occupation also strongly affected the number of fodder species reported by participants (H = 14.58; P < 0.01). Domestic livestock caretakers mentioned a higher number of plants (9.50 ± 4.43) followed by farmed

PLOS ONE
The fodder grass resources for ruminants

Pairwise ranking of wild palatable plants
Cymbopogon jwarancusa subsp. jwarancusa with 1 st rank was the most preferred among all selected grass species, followed by Cynodon dactylon, Cenchrus ciliaris, Typha elephantina and Cyperus alopecuroides that had 2 nd 3 rd , 4 th and 5 th rank respectively. Pycreus flavidus received the lowest score, therefore resulting as the less preferred species (Table 3). The most highly ranked species (Cymbopogon jwarancusa subsp. Jwarancusa, Cynodon dactylon, Cenchrus ciliaris, Typha elephantina and Cyperus alopecuroides) are also the most dominant in the area (Shaheen, unpublished data). This finding seems to support the "appearance hypothesis" stating that the most abundant species are better known and mostly used [37]. Plants commonly

PLOS ONE
The fodder grass resources for ruminants growing in the area allow local people to have more experience of their properties and consequently have a greater probability of being introduced into the local culture.

Correlation used for pairwise comparison
On the basis of RFC value, pairwise comparison was used to correlate fodder grasses and the knowledge of the respondent. Ten out of 232 respondents were chosen on the basis of their profession (ethnoveterinary practitioner) but were potential respondents due to sufficient indigenous knowledge. Based on RFC values knowledge of respondent R1 showed a strong correlation with R4, as R2 (0.56; p<0.001) with R1 and R7 (0.55;p<0.001), R2 had a strong correlation with R3 and R8 (0.48, 0.58; p<0.001) but R2 had the strongest correlation with R9 (0.71; p<0.001). All correlation and the distribution of RFC values are shown in Fig 5. The positive correlation between respondents suggests that respondents report similar information about the plant; for example, R2 and R9 both were ethnoveterinary practitioners more than 50 years old, so they had similar knowledge.

Availability and prioritizing fodder grasses on the basis of RFC and PWC
RCF values ranged from 1 to 0.51 with a mean value of 0.71. Twenty-five species had RFC values higher than average while the remaining 35 species had RFC value lower than average ( Fig  6, Table 4). Cymbopogon jwarancusa and Cynodon dactylon showed the highest value (1.00) while Imperata cylindrical (0.52) and Vetiveria zizanioides (0.51) had the lowest. Fic in Table 5

PLOS ONE
The fodder grass resources for ruminants conformed that Cymbopogon jwarancusa, Cynodon dactylon and Cenchrus ciliaris have highly useful as fodder. Based on these RFC values fodder species were classified into three categories of priority: species with higher priority (group A), species with medium priority (group B) and species with low priority (group C). Twenty-eight (45.9%) species were highly preferred by the participants, followed by twenty-three (37.7%) species that had medium priority while ten (16.3%) grass species were the least preferred (Fig 7). Values ranged between 1-0.69 for group A, between 0.69-0.54 for group B and between 0.54-0.51 for group C. Similar results were shown by Harun et al.
[12] in their study. These results were confirmed by cluster analysis based on RFC in which the reported species were classified into three major groups compliant with the results of priority ranking analysis. Similar results were found when we performed cluster analysis using PWC data. Cymbopogon jwarancusa was the preferred species in both approaches ( Table 6).

PLOS ONE
The species included in Group A (high priority) are ecologically dominant and largely available in the area. Moreover, taxa included in this group have a good palatability and are also available during the dry season when other grazing resources are exhausted.

Palatability of grasses and the method of feeding
Preferred palatable species are often leafy, with less stem, a low leaf table and leaves of low tensile strength [37,38]. Palatability analysis showed that 77% of the reported species are grazed in the study area (Table 7). In particular; grasses included in the group A of the priority ranking were consumed by all ruminants locally raised. Goats are the only animals to feed on every type of grass growing in Thal desert although palatability results show a preference for 58% of the reported species. 40% of the species represented the favorite fodder for sheep and 26% the favorite fodder for buffaloes. Camels are very selective animals and use only few specific grasses as fodder (Fig 8). Different parts showed to have different edibility: for example 42% of grass species were consumed as whole plant (e.g. Cynodon dactylon, Eragrostis minor, Cenchrus ciliaris, Cenchrus pennisetiformis, etc.) while 38% and 19% of them were consumed as aerial parts and as leaves, respectively. The reason why so many grasses are grazed as a whole is probably related to their small size and tender herbaceous texture (e.g. Cynodon dactylon, Lasiurus sindicus, Phalaris minor, Cyperus rotundus, Eragrostis minor etc. similar results shown in other literature [12,13]. Due to the sandy nature of soils occurring in the study area these plants have shallow root systems and can easily be pulled out from the soil. Species growing in the form of dense patches are hard to be consumed as a whole and animals feed only on the aerial parts. Beliefs on livestock feeding habits are common in the area: for example, some local shepherds reported that putting the herd out to pasture in open fields improves their health and milk production. According to them freely grazing animals are able to select the best grasses, avoiding the toxic or less nutritious ones. They justify this belief by comparing milk production of freely grazing animals with forage-fed cattle and also by saying that during dry season, when free grazing is not possible, there is a considerable reduction in animal health and milk production [38,39].

Role of the fodder species on milk production
Ten out of the 80 interviewed shepherds (based on the respondent knowledge) were randomly sampled to examine in detail the role of fodder species on the milk production. We focused our attention on the shepherds because, during the interviews, they showed a deeper knowledge about the plant species influencing quantity and quality of milk. According to them, Cynodon dactylon was the best species for milk production (6.46 SI, 0.6460 CS) followed by https://doi.org/10.1371/journal.pone.0224061.g008 Table 8. Results of ANTHROPAC analysis of overall salience index of milk for producing species. Cymbopogon jwarancusa (5.133 SI, 0.5133 CS). Cymbopogon jwarancusa was also reported to give a peculiar aroma, increasing the milk's value. Sorghum sp. was the third most salient species (5.121 SI, 0.5121 CS) ( Table 8). This findings were confirmed when we extended our analysis to all the participants. According to the results of the ANTHROPAC frequency analysis, ranking the plants in the order of their citation frequency (Fig 9), Cynodon dactylon had 73.21% frequency of milk production, following by Cymbopogon jwarancusa (70.54%) and Sorghum sp. (67.86%).

Relative abundance and seasonal availability
Relative abundance analysis showed that most of the cited species (55%) were abundantly present in the study area and most of them belonged to priority Group A (Fig 10). 13.39% of the species were available in August and in October while 12.54% were available in July. In Pakistan, July, August and October are months characterized by monsoon rains fostering the grass biomass development (Fig 11).

People use livestock for improving their economic life
Livestock production makes the main contribution to agriculture value-added services in the study area. Ten local participants were asked to rank animals from one to five on the basis of their economic value. Milk production is the major income source for people living in the Thal desert; for milk production, to cows and buffaloes are raised more frequently than camels or goats (Fig 12). Goats, sheep, buffaloes and cows are also raised for meat production. During religious celebrations (such as pilgrimages and Eid ul Azha) shepherds and farmers take livestock to the local market for sale and this is another major income source as also shown in [40]. Skins from sheep, buffaloes, cows and camels are also sold for making leather goods; teeth and bones are used for making different objects (e.g. buttons, jewelry and decoration pieces) (Fig 12). Dung of buffaloes and cows is dried and used as fuel or, fresh, as a natural

PLOS ONE
The fodder grass resources for ruminants fertilizer to improve the soil fertility. Ox, buffaloes and sometimes camels are used for ploughing. Camels are commonly used for transportation in desert areas.

Indigenous uses and ethno-veterinary uses of grasses
Eighteen of the 61 reported species were locally used in ethno-veterinary practice. Cymbopogon jwarancusa was the most cited veterinary grass (48) and was reported to heal infertility and skin diseases in ruminants (Table 9). Other species (Cenchrus spp., Arundo donax, Desmostachya bipinnata, Dichanthium annulatum, Digitaria ciliaris, Eleusine indica, Eragrostis spp., Saccharum spontaneum) were frequently reported as remedies to treat urinary and digestive diseases in livestock. Similar results are shown in different studies [12, 41,42]. Urinary and digestive diseases were the most frequently reported disorders; this finding is probably due to the sandy nature of the soil, causing the accumulation of sand-laden feed material in the digestive apparatus and urinary tract of livestock. Fic analysis showed in Table 5

Conclusion
The present study provides an inventory of plant species, plant parts and diversity in palatability and feeding behavior. The data analysis highlighted the possible motives behind the greater acceptability ratio of high priority fodder grasses: i.e. diversity in their palatability for major ruminant species, availability in the study area, and versatility of feeding methods. This study is not only significant for the conservation of ethnobotanical knowledge but may also help in facilitating sustainable feeding for ruminants. Subsequently, the information may play a major role in improving the livelihood of smallholder farmers. A blend of traditional and scientific knowledge is required to produce a worthwhile criterion for selecting fodder grasses. If some of the grasses show promising nutritional and pharmacological value, then necessary steps to conserve the area and the species should be taken. This will help to boost up the economy of the country.