Abstract

The paper aims to understand how the public perceives river water quality and related risks and behaviour. Using the stratified semi-purposive sampling process, the study explores the perception of people residing along the river Yamuna in India. The method applied involved a structured questionnaire survey of 2706 respondents and four focused group discussions with people residing within two kilometres of the river bank. Non-parametric tests such as Kruskal Wallis, Mann Whitney U-test and One-Sample Wilcoxon Signed-Rank Test were used to analyse the data. The findings suggest that the majority of the respondents formulate their perceptions using non-scientific methods like sensorial and heuristics. Perception on sources of pollution is shaped by personal experiences and people do not perceive diffused sources of pollution that affect river water quality. Respondents attributed the pollution in the river to anthropogenic activities and their risk perception was found to be linked to their direct dependence on the river for their daily needs. The paper suggests behavioural change strategies to focus on social, governance, and technological drivers.

1. Introduction

The need for public participation in water management has been emphasized on several occasions since the International Conference on Water and the Environment (ICWE) held in Dublin in 1992 (ICWE, 1992; Tippett et al., 2005; Pahl-wostl et al., 2008) and has been increasingly forming a part of government policies on water resources in the last two decades (Yufei et al., 2007; Ministry of Water Resources (MoWR), 2012). Despite widespread recognition, there are several challenges associated with securing public participation as it is linked to a wide variety of pre-conceived thoughts, beliefs and values associated with people (Mostert, 2003; Korff et al., 2012). Thus, to identify the type, scope and scale of engagement and corresponding methods for positive behavioural response from people, it is important to gain insight into public perception (Tran, 2006; Clay et al., 2007; Dobbie & Green, 2013). Perception is man's predominant form of cognitive contact with his or her environment, and is of seminal importance when understanding his or her behaviour (Efron, 1969). Therefore, a detailed understanding of how individuals perceive and behave, becomes imperative for policymakers and researchers, especially the ones seeking solutions that require behavioural change to manage natural resources (Anderson et al., 2007; Larson et al., 2009; Hu & Morton, 2011). Literature suggests that public perception of the quality of water varies with education (Anderson et al., 2007), age, income level (Larson & Lach, 2008), gender (Larson et al., 2010), social and psychological belief (Tran, 2006) and geographic proximity of a person to the water source (Hu & Morton, 2011). Research on analysis of public perception for public participation in water quality restoration programmes has been conducted in many successful river management projects, such as the river Thames in the United Kingdom (Tunstall, 2000), river Murray in Australia (Larson et al., 2013) and river Rhine in Europe (Buijs, 2009) and in developed countries such as Spain, Australia, United States, Canada, United Kingdom, Japan, etc. (Tunstall, 2000; Yamashita, 2002; Turgeon et al., 2004; Lepesteur et al., 2008; Roca & Villares, 2008; Le Lay et al., 2013). However, the same is not the case with respect to developing countries. Furthermore, it is interesting to note that most published studies reported in the body of literature have been prior to planning an intervention.

The study presented in this paper is different from the earlier studies, as it has been carried out at a river site where there have been multiple attempts during the past two decades to revive the river. Water quality restoration measures predominantly include infrastructure interventions to reduce pollution load into the river, an approach popular in all river action plans of India. The Yamuna Action Plan (YAP) was launched in 1993 with a soft loan of approximately 217 million dollars from Japan Bank for International Co-operation (JBIC) to take up pollution abatement works in the river Yamuna by building diversion sewer lines, sewage pumping stations and sewage treatment plants (STPs). The YAP I was extended to YAP II in 2004 with an additional budget of 142 million dollars and to YAP III in 2016 with allocation of 127 million dollars and with similar objectives (Tokyo Engineering Consultants (TEC), 2009; Ministry of Environment and Forests (MoEF), 2011; The Indian Express, 2016).

The study aims to bring out evidence regarding public perception of river water quality, factors influencing perception, perception about sources of pollution, significance of the river, risk perception due to change in water quality and reasons behind water-related behaviour of the people. Further, the study suggests interventions that would help in creating desirable behavioural changes among the people and provides policy relevant information to design effective public participation strategies for river quality restoration projects. A mixed method approach comprising of both quantitative and qualitative techniques was used to achieve the objectives of the study. The oral questionnaire survey was chosen to capture the response of a diverse set of respondents and to ensure that every question was answered. Focus Group Discussions (FGDs) helped in gaining deeper insights into the findings from the survey as well as validating the findings through triangulation of the data. The next section explains the materials and methods used for the study which include site description, data collection methodology, analysis tools and sampling. This section is followed by results and discussion, policy implications, and the conclusion.

Materials and methods

Site description

Three stretches of the river Yamuna have been considered in this study (Figure 1). These include a 22 km critically polluted stretch of the river along Delhi (Sharma et al., 2015), a 183 km relatively cleaner stretch upstream of Delhi and a 155 km stretch downstream of Delhi (Rani et al., 2013). The survey was conducted in the districts of Yamunanagar, Muzaffarnagar, Panipat, Baghpat and Sonipat upstream of Delhi, in Delhi, and at Faridabad and Mathura districts downstream of Delhi. The ranking for water quality used in the study is aligned with the designated best use classification of surface water quality by the Central Pollution Control Board (CPCB) as shown in Table 1. These categories for the purpose of the study have been ranked as ‘1’ for very poor and ‘5’ for very good.

Table 1.

Designated best use classification of surface water. Source:CPCB (2007).

Designated best use Class of water Water quality ranking (used in the study) 
Drinking water source without conventional treatment but after disinfection 
Outdoor bathing 
Drinking water source after conventional treatment and disinfection 
Propagation of wild life and fisheries 
Irrigation, industrial cooling, controlled waste disposal 
Designated best use Class of water Water quality ranking (used in the study) 
Drinking water source without conventional treatment but after disinfection 
Outdoor bathing 
Drinking water source after conventional treatment and disinfection 
Propagation of wild life and fisheries 
Irrigation, industrial cooling, controlled waste disposal 
Fig. 1.

Schematic diagram of the study area.

Fig. 1.

Schematic diagram of the study area.

Since 1993, Government of India (GoI) has been making efforts to clean the river through its mega project, Yamuna Action Plan (YAP) which mainly focuses on water treatment based infrastructure projects (TEC, 2009). According to the Ministry of Environment and Forests (MoEF), the Government invested over 350 million dollars on YAP I and II between 1993 and 2011; 40% of which was used within Stretch 2 for building infrastructure and creating public awareness to restore the water quality of the river Yamuna (MoEF, 2011). Despite these investments and interventions, the water quality of the Yamuna in Stretch 2 continues to be of Class E (Sharma et al., 2015). The water quality in Stretch 1 is reported to be of Class B and in Stretch 3 it is of Class C (Khan & Singh, 2013; Rani et al., 2013; CPCB, 2016).

Data collection and analysis

Questionnaire design

The first section of the questionnaire focused on the demographic status and location aspects of the respondent. The information was recorded by assigning codes for variables:

  • gender – male, 1; female, 2

  • income category – low income (LI), 1; medium income (MI), 2; high income (HI), 3

  • level of education – illiterate, 1; semi-literate, 2; literate, 3

  • location – Stretch 1, 1; Stretch 2, 2; Stretch 3, 3

The income status of an individual has been classified differently by various agencies such as National Council of Applied Economic Research (NCAER) and McKinsey Global Institute (MGI) (Beinhocker et al., 2007). Drawing from these reports, for the purpose of this study, we have considered ‘low income’ as those having monthly household earning of less than INR 15,000, ‘medium income’ for monthly household earning between INR 15,000–60,000 and ‘high income’ for monthly household earning of INR 60,000 and above (1 US Dollar ∼ 64 INR). To ascertain that the income being mentioned by the person was reported correctly, questions on family size, nature of employment of the earning members and type of residence were asked in order to arrive at a judgemental income.

The Census of India has defined literacy as the ability to read and write with understanding of any language, and minimum educational qualification is not necessary to be considered literate (National Commission to Review the Working of the Constitution (NCRWC), 2001). The Census further classifies literates into ‘below primary’, ‘primary’, ‘middle’, ‘matric/secondary’, ‘higher secondary’, ‘diploma’ and ‘graduates and above’ levels (Census, 2001). The classification given by the Census was analysed according to the number of years of education under each category and for the purpose of this study, a person with 5 years or less schooling is considered as illiterate, 6 to 12 years is considered as semi-literate and greater than 12 years is considered as literate.

The second section assesses the perception of river water quality which was measured on a scale of 1–5. The following parameters were considered to ascertain the basis of perceiving water quality (de França Doria, 2010):

  • Sensorial: decision based on taste, colour and smell

  • Contextual: relating to the presence of refuse in the water, on the bank, sewer line and drain outlets entering the river and the presence of aquatic life

  • Scientific: relating to knowledge of presence of harmful chemicals based on evidence

  • Heuristics: opinion based on experience, intuitive judgement, educated guess, etc.

  • Culture and belief

The perception about the various sources of pollution was determined based on the respondent's response to the choices presented; namely, ‘industrial’, ‘domestic’, ‘agricultural activities’, and ‘customs and rituals’. Furthermore, they were also given the option to cite any source beyond the choices presented. The third section focused on understanding the perception of risk among the public with respect to the changes in water quality. Unlike risk assessment by scientists, risk perception is a personal understanding of risk built over time due to experience, knowledge and memory. It is highly subjective and relates to an individual's judgement about the risks they believe they are exposed to (Slovic, 1994; Rohrmann, 2008). These were first assessed by identifying the significance of the river among the people. To capture this, participants were asked to choose from the following categories: ‘emotional/cultural’, ‘occupational’, ‘environmental’, ‘daily needs’, ‘recreational’, ‘no relation’ and ‘no idea’. Risk perception was further measured with a single item scale by asking the respondents: ‘Do you perceive any risk to quality of your life from the changing water quality of the river Yamuna?’ The respondents could reply with either yes (1) or no (0). Those who answered in the affirmative were asked to select the nature of the risk. The respondents were given five choices, ‘health’, ‘economic’, ‘environmental’, ‘socio-cultural’ and ‘any other’. In all the multiple-choice questions, the respondents were given the flexibility of choosing multiple options as the response to each option was to be given as yes (1) or no (0).

The fourth section of the questionnaire was designed to understand the behaviour of people related to water. In 1975, Fishbein and Ajzen coined the ‘theory of reasoned action (TRA)’ for evaluating the factors responsible for a specific behaviour. TRA suggests that a person's behaviour is determined by his intention to perform the behaviour. The behavioural intent, in turn, is a function of his attitude and subjective norm. Whereas attitude is a personal trait, subjective norm is the influence of society on the perception of the person to perform a specific behaviour (Ajzen, 1991). The reasons behind water related behaviour were thus computed using eight questions and the respondents were asked to record their answers using a five-point Likert scale, with 1 representing strongly disagree, and 5 representing strongly agree. The questions sought responses on public ownership of their action for degrading water quality, the reasons for their action – habit, convenience, societal influence, lack of option, ignorance, absence of penalty and whether awareness would bring the required change in behaviour.

A pilot run of the questionnaire was performed on 25 volunteers selected from each stretch to pre-test and fine tune the questions. Furthermore, Cronbach's alpha reliability test was performed on Likert-scale questions for measuring internal consistency of scores. The alpha coefficient for the eight items was found to be 0.777 suggesting that the items had high internal consistency (Gliem & Gliem, 2003). To ensure consistency in the response, the data was collected in two phases around the same time period during the year May to June 2014 and April to May 2015. Survey forms were filled in by the researchers during face-to-face interactions with the respondents. This ensured that all questions were answered with no missing values. The interviews were conducted in Hindi which was the native language of the respondents.

Sampling

Keeping in mind the broad spectrum of respondents, semi-purposive sampling was used for the study (Crona et al., 2009; Sovacool et al., 2012). People residing within two kilometres of the river bank and having the opportunity to view the river more frequently than others, were surveyed for a perception analysis. This was based on the commonly known fact that dependency on the river for a livelihood reduces as one moves away from the river due to diversification in the availability of sources of water. However, to capture the diversity of views, each segment was further divided into sections comprising residential areas, educational institutions, market place and other public places like parks, religious places, etc. A sampling grid was designed to sample participants representing key independent variables such as gender, education and income. The questionnaire was administered to 2706 individuals in three stretches (see Figure 1); Stretch 1 (n = 834), Stretch 2 (n = 1,168), and Stretch 3 (n = 704). The demographic attributes of the surveyed population sample are shown in Table 2.

Table 2.

Demographic attributes of surveyed population (n = 2,706).

  Illiterate
 
Semi-literate
 
Literate
 
 
  LI MI HI LI MI HI LI MI HI Total 
Stretch 1 Male 117 49 10 20 75 40 10 233 557 
 Female 50 49 10 10 14 24 10 105 277 
Stretch 2 Male 123 55 10 84 92 49 25 44 89 571 
 Female 95 50 15 89 95 20 40 30 163 597 
Stretch 3 Male 40 14 15 74 25 39 64 40 320 
 Female 70 54 55 39 96 44 19 384 
 Total 495 271 58 273 389 161 215 195 649 2,706 
  Illiterate
 
Semi-literate
 
Literate
 
 
  LI MI HI LI MI HI LI MI HI Total 
Stretch 1 Male 117 49 10 20 75 40 10 233 557 
 Female 50 49 10 10 14 24 10 105 277 
Stretch 2 Male 123 55 10 84 92 49 25 44 89 571 
 Female 95 50 15 89 95 20 40 30 163 597 
Stretch 3 Male 40 14 15 74 25 39 64 40 320 
 Female 70 54 55 39 96 44 19 384 
 Total 495 271 58 273 389 161 215 195 649 2,706 

Four FGDs were conducted with homogeneous groups to gain detailed insights into the perception, extent of knowledge and behavioural patterns of the public towards water quality of the river. Of the four groups, one group had only female participants. The FGD began with a general question regarding the water quality of the river Yamuna. Participants were asked to express their views on the river and its changing water quality. This was followed by a few probing questions on the reasons behind their views on the water quality of the river. The participants were also asked to share their opinions on the significance of the river and its impact on their well-being due to the changing water quality. All FGDs were audio recorded.

Data analysis

Numerically coded data was used for carrying out statistical analysis. Descriptive statistics were calculated across all variables to summarize the data and to gain understanding of its distribution and spread. Since the data did not follow a normal distribution, non-parametric tests, Kruskal Wallis and Mann Whitney U-tests (Larson & Lach, 2008; Steinwender et al., 2008) were used to assess the difference of means with respect to the above referred variables.

The One-Sample Wilcoxon Signed-Rank Test was conducted to gauge the public perception of water quality of the river Yamuna. This was done with the aim of determining whether the median of the sample was equal to the hypothesized value of 3 (neither good nor bad). The reporting of all statistically significant differences was based on the 0.05 level of significance (p-value). Pairwise comparison of the means was carried out through the Bonferroni post hoc test on variables that consisted of more than two independent samples. The means were later compared with scientifically available data on water quality of the river Yamuna to understand the similarities and differences in the perception and reality. The responses from the FGDs were also considered after content analysis to support the quantitative analysis.

The questionnaire data measuring water-related behaviour using the Likert scale was initially subjected to exploratory analysis. This included plotting of a scatter graph amongst all the independent variables and the computation of a correlation matrix, in order to explore possibilities of subjecting advanced multivariate methods to extract more insights from the data. It was noted that the correlation between most of the independent variables was statistically significant. This suggested exploring possibilities of using multivariate data analysis techniques like principal component analysis (PCA) and factor analysis (FA) to extract the grouping variables, if any, in the form of major factors which could help in bringing the required behavioural change in people.

Results and discussion

Public perception of water quality

A summary of the means of public perception of water quality has been provided in Table 3. It is evident from the values of the weighted mean scores in Stretches 1, 2 and 3 that the public perceives water quality as ‘medium’ in Stretch 1 but ‘bad’ in Stretches 2 and 3. The Wilcoxon Signed-Rank Test on one sample revealed that at a significance level of 5%, the median value of public perception of water quality was equal to the hypothesized value of 3 in Stretch 1 and lower in Stretches 2 and 3. The finding contradicts the scientific evidence which reveals that water quality is ‘good’ (Class B) in Stretch 1 and ‘medium’ (Class C) in Stretch 3 (Rani et al., 2013). Underestimating the water quality in Stretch 1 indicates that people have higher expectations where few visible polluting sources exist. The FGDs revealed that people were unaware of YAP and their perception is being influenced by the multiple visible sources of pollution on Stretches 2 and 3. It is important to note here, that scientific information is rarely accessible to the common man who does not have the requisite skills to interpret the data if provided to them.

Table 3.

Mean values of public perception of water quality on a scale of 1 to 5 (1 representing very bad and 5 representing very good).

  Illiterate
 
Semi-literate
 
Literate
 
 
  LI MI HI LI MI HI LI MI HI Mean 
Stretch 1 Male 3.48 4.18 3.50 4.25 3.00 2.88 5.00 1.00 2.46 3.04 
 Female 2.70 4.18 4.50 5.00 3.57 3.58 5.00 3.00 2.57 3.24 
Pooled data   3.63     3.34     2.59   3.11   
Stretch 2 Male 1.41 1.36 2.00 1.77 1.44 1.61 1.20 1.34 1.54 1.50 
 Female 1.52 2.10 2.00 1.79 1.79 1.50 1.75 2.50 1.69 1.78 
Pooled data   1.59     1.68     1.66   1.64   
Stretch 3 Male 1.38 1.57 1.44 1.00 1.53 1.60 1.00 1.08 1.13 1.28 
 Female 1.64 1.37 1.00 1.36 1.26 1.30 1.41 1.32 1.53 1.42 
Pooled data   1.48     1.41     1.24   1.36   
  Illiterate
 
Semi-literate
 
Literate
 
 
  LI MI HI LI MI HI LI MI HI Mean 
Stretch 1 Male 3.48 4.18 3.50 4.25 3.00 2.88 5.00 1.00 2.46 3.04 
 Female 2.70 4.18 4.50 5.00 3.57 3.58 5.00 3.00 2.57 3.24 
Pooled data   3.63     3.34     2.59   3.11   
Stretch 2 Male 1.41 1.36 2.00 1.77 1.44 1.61 1.20 1.34 1.54 1.50 
 Female 1.52 2.10 2.00 1.79 1.79 1.50 1.75 2.50 1.69 1.78 
Pooled data   1.59     1.68     1.66   1.64   
Stretch 3 Male 1.38 1.57 1.44 1.00 1.53 1.60 1.00 1.08 1.13 1.28 
 Female 1.64 1.37 1.00 1.36 1.26 1.30 1.41 1.32 1.53 1.42 
Pooled data   1.48     1.41     1.24   1.36   

A Mann Whitney U-test for gender groups revealed that the perceptions of water quality among men and women were statistically different in all stretches. Women in general were found to be slightly more positive about the river water quality as compared to men. A possible rationale behind this finding that emerged in the FGD was that women were not as educated in comparison to men and had limited exposure to sources of information like the newspaper. A Kruskal Wallis test revealed a significant difference based on income only in Stretch 1, wherein the lower income group perceived water quality to be better than the higher income group. The reason that emerged for this was higher dependency of the lower income group on river water for their daily requirements in comparison to the higher income respondents, who had access to the piped water system and were not dependent on the river. Also, a substantial difference in the perception of water quality was seen based on education in Stretch 1 and Stretch 3. A pairwise comparison was conducted through a Bonferroni post-hoc test, to understand whether a difference existed between literate and illiterate, semi-literate and literate, or illiterate and semi-literate respondents. The result revealed a significant difference in the perception only among illiterate and literate respondents in Stretch 1 and 3. No difference in perception was observed between the illiterate and semi-literate, and the semi-literate and literate community. The literate people were more conservative in water quality perception compared to illiterate respondents.

Factors that influence public perception of water quality

The sensorial factor has emerged as the most important basis of public perception of water quality in all stretches and across all independent variables. Among the total sample surveyed, 93.8% considered parameters like taste, colour and odour as their basis of perception. On the contrary only a meagre 6.2% of the total respondents, mostly from the literate and semi-literate community from Stretch 2, based their perception on scientific parameters of water quality. The reasons for such findings can be attributed to the lack of access to scientific data and information and growing attention, investment and public debate on the revival of the river water quality in Stretch 2.

However, 46.9% of the respondents from Stretches 2 and 3, considered contextual indicators as another significant basis for developing water quality perception. The FGD revealed that the reason for this was visible drains carrying raw sewage into the river, the diminishing aquatic life, open defecation and the presence of trash on the banks.

Perception based on heuristic was notable only in Stretch 1 where water quality is good. The FGD findings indicate that this was primarily because people of this stretch have not experienced any negative consequence due to the river water quality in their knowledge and recall time. Significant difference was also observed between income groups and education groups in Stretch 1 while considering heuristic (p < 0.001) as the basis of perception. The illiterate lower income group primarily based their perception on sensorial factors and seldom considered heuristics as the basis of perception.

Culture and belief was not reported to be the prominent factor affecting perception except by a small percentage of male respondents in Stretch 3. The literature review also did not provide any succinct information on the cross linkage of culture, demographics and water quality perception. Several researchers have argued that traditions associated with culture and belief play a role in building perception about the environment (Larson et al., 2009; de França Doria, 2010). The findings of FGDs revealed that the respondents from an older age category were more critical about the deteriorating water quality as they had seen the river when it was cleaner. However, perception based on knowledge of scientific parameters did not resonate even in the FGDs. The participants said that unlike air pollution, which is broadcast both on television and on digital boards in the city, there were no mechanisms in place to share scientific information about the water quality. Refusal of animals to drink water from the river, stomach infection from consumption of vegetables grown on the river bank, and black coatings in the vessels in which river water was stored were other factors people cited as having influence on their perception.

Perception of sources of pollution

Domestic and industrial discharge was perceived to be the main cause of pollution irrespective of the socio-economic status and the place of residence of the respondents. Significant difference (p < 0.001) was observed in the perception about the sources of pollution among women of different education levels and income groups in Stretch 1. The reason for this, that emerged from the FGD, was that perception is highly dependent on personal experience. Since, interaction of the lower income women with the river water is for domestic use only; they are unable to perceive other sources of pollution. Furthermore, only 10% of the respondents of the literate group from Stretch 2 cited agriculture as a source of pollution; suggesting that the public is unable to perceive diffused sources of pollution, like agricultural activities, as the cause for deteriorating water quality of the river. Similarly, pollution due to customs and rituals was also perceived mainly by the literate respondents along all of the three stretches. The findings of the FGD reveal that education enables people to gain access to information and to visualize the problem from multiple perspectives. A small percentage (1.3%) of respondents could not identify specifically with the options provided in the questionnaire. They stated that the drains from Stretch 2 that bring untreated sewage directly to the river were responsible for polluting the river.

Significance of the river and risk perception

Significance of the river

The emotional and cultural significance of the river in the lives of the people was found to be the highest across all three stretches. Stretch 1 (88%) and Stretch 3 (79%) had a higher percentage of respondents who were emotionally and culturally connected with the river than Stretch 2 (52%). Significant differences in the responses towards cultural significance were observed between men and women in Stretch 2 (p = 0.039) and Stretch 3 (p = 0.046). Men were found to be more culturally connected than women. In addition, the environmental significance of the river was mentioned by 50% of the respondents in all of the three stretches. However, a significant difference was seen between literate, semi-literate and illiterate respondents in Stretches 2 (p < 0.001) and 3 (p = 0.011). Contrasting opinions were noted on the basis of income group in Stretch 2 (p < 0.001). Respondents from the higher income, literate category believed that the deteriorating water quality had an adverse impact on their quality of life, including health, and has had a great impact on the ecosystem of the river. Elevated levels of pollution in the middle stretch, in addition to the encroachment on flood plains and river catchment area, has led to depletion of the water resource, decrease in biodiversity, increase in air pollution and rise in extreme weather events like floods and drought.

Dependence on the river for their daily needs, like bathing and washing, was cited by 52% of the respondents. A contrast in responses was observed between income groups in Stretch 1 and among men and women in Stretch 2. The lower income group in Stretch 1 and illiterate and semi-literate women in Stretch 2 were more dependent on the river water for daily needs. However, only a meagre 13% of the overall respondents mentioned dependence on the river for their occupational needs. The findings of the FGD brought out that the deteriorating water quality has severely impacted the occupational significance of the river in the lives of people.

Risk perception

The percentage of the population perceiving risk due to changes in water quality of the river Yamuna was higher in Stretch 1 (93%) and Stretch 3 (88%), in comparison to Stretch 2 (73%), despite water quality being scientifically assessed as ‘very bad’ in Stretch 2 (Sharma et al., 2015). Minimum direct exposure to the river in Stretch 2 was one of the reasons highlighted in the FGDs. Respondents further stressed the health associated risks with the river. The participants stated that the river was a breeding ground for mosquitoes, resulting in several vector-borne diseases. The foul smell emanating from the river triggered respiratory problems among the people living near the river. In addition, the poor surface water quality has led to the contamination of the ground water and the consumption of untreated water caused gastrointestinal and associated diseases. They also reported that the contaminants in the water caused skin infection when used for bathing. The respondents also elaborated upon the economic risk associated with the deteriorating water quality. The perception of impact due to the economic risk was higher in Stretch 2 (36%), followed by Stretch 3 (31%). The impact as mentioned by the respondents was primarily on the household budget, as they had to either purchase water for consumption or install expensive treatment systems to purify the water prior to use. Other impacts were loss of livelihood options, low return on vegetables grown on Yamuna plain, stagnation in prices of property close to the river and the effect on the metallic parts of household equipment. Environmental risks due to pollution in the river Yamuna were also stated by 46% of the total respondents. They indicated that it was not only polluting the surrounding air, but has also led to reduced rainfall and increase in temperatures over the last decade. An impact on the ecosystem by way of loss in aquatic life, reduced crops, and threat to migratory birds was perceived by several respondents. A significant difference in perception of economic and environmental risks was observed between income groups (p = 0.031) and education groups (p < 0.001). While the literate respondents were more concerned about the environmental risk due to pollution in the river, the lower income respondents felt it was more an economic risk. Social and cultural risk was mainly cited by 36% of the respondents from Stretch 1. Differences based on gender (p < 0.001) were observed; with a greater number of men than women citing social and cultural impact. Low self-esteem due to inferior quality of the river water was yet another impact mentioned by the respondents.

Water-related behaviour of public

The mean scores of eight Likert-scale questions that were used to ascertain water-related behaviour of the public are presented in Table 4. The findings of the survey support studies that attribute anthropogenic activities as a leading cause for river pollution (Brewer & Stern, 2005). With an average score of 4.2 the respondents agreed that degradation in water quality can be attributed to human behaviour. It was believed that the reasons for disposing of waste in water were not ignorance but habit or convenience. Although, the absence of penal action emerged as a primary reason behind pollution of the river, a lack of options to dispose of waste was also considered an important reason behind indiscriminate disposal. The respondents also agreed that behavioural patterns and the cultural landscape of society have an influence on their actions. They stated that if disposing of waste on the banks or in water has a societal sanction then individuals do not feel embarrassed in adopting that behaviour.

Table 4.

Mean values of reasons behind water related behaviours and p-values between groups. Measured on a 5-point agreement scale with 1 representing strongly disagree and 5 representing strongly agree.

 Mean score
 
Independent variables Public action degrades water Habit Societal influence Convenience No alternate option Ignorance Absence of penalty Need for awareness 
Illiterate (I) 4.04 4.16 3.95 4.12 3.92 3.56 4.30 4.03 
Semi-literate (SL) 4.17 4.02 3.81 3.98 3.89 3.29 4.34 4.30 
Literate (L) 4.36 4.18 3.67 3.73 3.88 3.18 4.45 4.31 
p-value (α = 0.05) 0.001 0.161 0.029 0.001 0.942 0.009 0.300 0.004 
Sig. L-I 0.001   0.024 0.001  0.008   0.007 
  L-SL 0.072   0.587 0.067  1.000  1.000 
  SL-I 0.461  0.596 0.618  0.134  0.015 
Male 4.10 4.00 3.75 3.90 3.87 3.26 4.21 4.09 
Female 4.33 4.27 3.85 3.94 3.93 3.41 4.06 4.37 
p-value (α = 0.05) 0.001 0.001 0.231 0.655 0.492 0.140 0.001 0.001 
LI 4.13 4.07 3.85 3.92 3.91 3.47 4.42 4.18 
MI 4.21 4.14 3.76 3.91 3.86 3.23 4.32 4.20 
HI 4.45 4.29 3.80 4.00 4.05 3.35 4.45 4.45 
p-value (α = 0.05) 0.030 0.222 0.654 0.847 0.488 0.102 0.448 0.114 
Sig. LI-MI 0.766              
  LI- HI 0.025             
  MI-HI 0.127         
 Mean score
 
Independent variables Public action degrades water Habit Societal influence Convenience No alternate option Ignorance Absence of penalty Need for awareness 
Illiterate (I) 4.04 4.16 3.95 4.12 3.92 3.56 4.30 4.03 
Semi-literate (SL) 4.17 4.02 3.81 3.98 3.89 3.29 4.34 4.30 
Literate (L) 4.36 4.18 3.67 3.73 3.88 3.18 4.45 4.31 
p-value (α = 0.05) 0.001 0.161 0.029 0.001 0.942 0.009 0.300 0.004 
Sig. L-I 0.001   0.024 0.001  0.008   0.007 
  L-SL 0.072   0.587 0.067  1.000  1.000 
  SL-I 0.461  0.596 0.618  0.134  0.015 
Male 4.10 4.00 3.75 3.90 3.87 3.26 4.21 4.09 
Female 4.33 4.27 3.85 3.94 3.93 3.41 4.06 4.37 
p-value (α = 0.05) 0.001 0.001 0.231 0.655 0.492 0.140 0.001 0.001 
LI 4.13 4.07 3.85 3.92 3.91 3.47 4.42 4.18 
MI 4.21 4.14 3.76 3.91 3.86 3.23 4.32 4.20 
HI 4.45 4.29 3.80 4.00 4.05 3.35 4.45 4.45 
p-value (α = 0.05) 0.030 0.222 0.654 0.847 0.488 0.102 0.448 0.114 
Sig. LI-MI 0.766              
  LI- HI 0.025             
  MI-HI 0.127         

Further, three components with eigenvalues greater than 1 were extracted using PCA and FA to determine grouping of variables into important drivers which could help in bringing the required behavioural change in people. The factor loading of each variable is shown in Table 5. The first factor that has been extracted can be named as ‘social drivers’, as items like habit, societal influence, and convenience load highly on it. The second factor can be called ‘governance drivers’, as items like absence of penalty and awareness to bring in change in behaviour load heavily on it and the third factor can be named ‘technology drivers’, as items like lack of disposal options and ignorance load highly on it. These components clearly outline the immediate reasons behind river water pollution; namely, lack of available infrastructure for proper disposal of waste, weak governance and the behavioural and cultural conundrums in the society.

Table 5.

Extraction of factors for grouping of variables.

 Component Matrix
 
 Social Governance Technological 
Public action degrades river water 0.437 0.343 −0.346 
People dispose of waste in water and river bank due to personal habit 0.602 −0.201 −0.511 
People dispose of waste in water and river bank due to societal influence 0.660 −0.409 −0.133 
People dispose of waste in water and river bank due to convenience involved 0.673 −0.416  
People dispose of waste in water and river bank as no alternate option is available 0.479 −0.147 0.596 
People dispose of waste in water and river bank due to ignorance 0.519 0.140 0.510 
People dispose of waste in water and river bank due to absence of penalty 0.373 0.672  
Creating awareness can bring the required change in behaviour 0.496 0.540  
 Component Matrix
 
 Social Governance Technological 
Public action degrades river water 0.437 0.343 −0.346 
People dispose of waste in water and river bank due to personal habit 0.602 −0.201 −0.511 
People dispose of waste in water and river bank due to societal influence 0.660 −0.409 −0.133 
People dispose of waste in water and river bank due to convenience involved 0.673 −0.416  
People dispose of waste in water and river bank as no alternate option is available 0.479 −0.147 0.596 
People dispose of waste in water and river bank due to ignorance 0.519 0.140 0.510 
People dispose of waste in water and river bank due to absence of penalty 0.373 0.672  
Creating awareness can bring the required change in behaviour 0.496 0.540  

Policy implications

The research brings out that public perceptions regarding water quality are primarily based on sensorial parameters and contextual indicators, underlining the primacy of aesthetic quality. The river water quality classification methods such as Water Quality Index (WQI) and ambient water standards, promulgated by regulatory agencies, are predominantly a measure of dissolved impurities in water, and aesthetic parameters like floating solids, colour, odour, etc., though recognized as pollutants, are not a part of the water quality measurement standards. Further, the inability of people to perceive the water quality on the basis of scientific parameters often leads to misestimating of risks associated with water quality (Larson et al., 2009). Planners should ensure that the awareness programmes are designed in a way to help people develop a holistic understanding of the issues affecting water quality, and additional efforts should be undertaken to make actual water quality information available to them in comprehensible language. The study further suggests that the perception of sources of pollution is largely based on personal experience. People are unable to perceive non-point sources of pollution, like that from agricultural runoff, and they largely attribute pollution to the indiscriminate discharge from urban areas. The research also reveals that the perceptions of people and their preferences vary geographically, which highlights the necessity of undertaking surveys and FGDs, to acquire useful insights into public perception, knowledge, and preferences prior to planning interventions.

Research has shown that perceived benefits by the public from participation and a sense of ownership of the resource affects the willingness to participate (Jingling et al., 2010; Brandt & Svendsen, 2013; Marks et al., 2014). There is evidence from the study that while people attribute the cause for degradation of water quality to public action, they believe that the responsibility for cleaning the river lies with the government. The study also indicates that the adverse water-related behaviour by people is often due to habit and convenience rather than lack of infrastructure. While rivers provide social, cultural as well as economic benefits for the people, the study brings out that only a small percentage of the respondents are dependent on the river for occupational needs due to poor water quality. It is pertinent that policymakers should take these findings into consideration while designing river action plans and engage with influential actors such as community leaders, religious groups, and local elected representatives to catalyse apposite behavioural change among people. The respondents of FGDs were also of the opinion that water quality is not accorded the same level of importance as air quality. Success of use of mass media and digital display boards has been seen in the case of creating awareness about the air quality index in Indian cities (Times of India (ToI), 2015). Similar steps can be taken to inform communities about water quality levels. Furthermore, understanding local social norms and individual choices is also important for planning interventions and communication strategy. As brought out in the study, activities need to be undertaken at the community level in addition to the use of mass media platforms to mobilize public action. Finally, the findings reveal a more liberal assessment of water quality by women, and greater dependence of women, especially from rural areas, on the river for their daily needs, making them more vulnerable to infections associated with poor water quality. This finding demands adequate emphasis on gender mainstreaming in water management and integration of women in decision-making activities related to river management (United Nations Development Programme (UNDP), 2006).

Conclusion

This study is an attempt to report research that examines factors influencing public perception of water quality and public knowledge regarding sources of pollution, risk perception and water-related behaviour. The policymakers may subsequently use these findings to better address perceptions that underpin public behaviour, knowledge and associated action towards river water quality. The study brings out that technical interventions by way of building STPs and sewer diversion lines, as part of the YAP, do not seem to have had a positive influence on the perception of the public regarding water quality of the river, thereby signifying the need for a paradigm shift in strategy from a techno-centric approach to a people-centric approach, which takes into consideration the diverse viewpoints and cultural beliefs of all stakeholders. The authorities also need to simultaneously provide adequate infrastructure for waste disposal and put in place a robust mechanism for punitive measures against defaulters. In India, where people consider the river alive and sacred, a socio-emotional connection can also be leveraged to re-establish the community's sense of belongingness to the rivers and build grass-root government-public collaborations for enhanced effectiveness of intervention measures.

References

References
Ajzen
,
I.
, (
1991
).
The theory of planned behavior
.
Organizational Behaviour and Human Decision Processes
50
,
179
211
.
Anderson
,
B. A.
,
Romani
,
J. H.
,
Phillips
,
H.
,
Wentzel
,
M.
&
Tlabela
,
K.
, (
2007
).
Exploring environmental perceptions, behaviors and awareness: water and water pollution in South Africa
.
Population and Environment
28
(
3
),
133
161
.
Beinhocker
,
E. D.
,
Farrell
,
D.
&
Zainulbhai
,
S. A.
, (
2007
).
Tracking the Growth of India's Middle Class
.
Brandt
,
U. S.
&
Svendsen
,
G. T.
, (
2013
).
Is local participation always optimal for sustainable action? The costs of consensus-building in local agenda 21
.
Journal of Environmental Management
129C
,
266
273
.
Available at: http://www.ncbi.nlm.nih.gov/pubmed/23974351 (accessed September 17 2013)
.
Brewer
,
G. D.
&
Stern
,
P. C.
, (eds) (
2005
).
Decision Making for the Environment: Social and Behavioral Science Research Priorities
.
The National Academies Press
,
Washington, DC
.
Census
(
2001
).
Census of India: Literacy and Level of Education
. .
Clay
,
D. E.
,
Ren
,
C.
,
Reese
,
C.
,
Waskom
,
R.
,
Bauder
,
J.
,
Mesner
,
N.
,
Paige
,
G.
,
Reddy
,
K.
,
Neibauer
,
M.
&
Mahler
,
R.
, (
2007
).
Linking public attitudes with perceptions of factors impacting water quality and attending learning activities
.
Journal of Natural Resources and Life Sciences Education
36
,
36
44
.
CPCB
(
2007
).
Guidelines for Water Quality Monitoring
. In:
Central Pollution Control Board, Ministry of Environment and Forests
, pp.
1
35
.
CPCB
(
2016
).
Water Quality of River Yamuna in Delhi
,
New Delhi
.
Available at: http://cpcb.nic.in/water_quality_river_yamuna_08-09-2016.pdf (accessed January 1 2018)
.
Crona
,
B. I.
,
Ronnback
,
P.
,
Jiddawi
,
N.
,
Ochiewoe
,
J.
,
Maghimbi
,
S.
&
Bandeira
,
S.
, (
2009
).
Murky water: analyzing risk perception and stakeholder vulnerability related to sewage impacts in mangroves of East Africa
.
Global Environmental Change
19
(
2
),
227
239
. .
de França Doria
,
M.
, (
2010
).
Factors influencing public perception of drinking water quality
.
Water Policy
12
(
1
),
1
19
.
Available at: http://www.iwaponline.com/wp/01201/wp012010001.htm (accessed March 14 2013)
.
Dobbie
,
M.
&
Green
,
R.
, (
2013
).
Public perceptions of freshwater wetlands in Victoria, Australia
.
Landscape and Urban Planning
110
,
143
154
.
Available at: http://linkinghub.elsevier.com/retrieve/pii/S0169204612003155 (accessed September 17 2013)
.
Efron
,
R.
, (
1969
).
What is perception?
In:
Proceedings of the Boston Colloquium for the Philosophy of Science
,
Cohen
,
R. S.
&
Wartofsky
,
M.W.
(eds),
D. Reidel Publishing Company
,
The Netherlands
, pp.
137
173
. .
Gliem
,
J. A.
&
Gliem
,
R. R.
, (
2003
).
Calculating, interpreting, and reporting Cronbach's alpha reliability coefficient for Likert-type scales
. In:
Midwest Research to Practice Conference in Adult, Continuing, and Community Education, (1992)
, pp.
82
88
. .
Hu
,
Z.
&
Morton
,
L. W.
, (
2011
).
U.S. midwestern residents perceptions of water quality
.
Water
3
(
4
),
217
234
.
Available at: http://www.mdpi.com/2073-4441/3/1/217/ (accessed March 26 2013)
.
ICWE
(
1992
).
The Dublin Statement and Report on the International Conference on Water and the Environment
. .
Jingling
,
L.
,
Yun
,
L.
,
Liya
,
S.
,
Zhiguo
,
C.
&
Baoqiang
,
Z.
, (
2010
).
Public participation in water resources management of Haihe river basin, China: the analysis and evaluation of status quo
.
Procedia Environmental Sciences
2
,
1750
1758
. .
Khan
,
S.
&
Singh
,
S. K.
, (
2013
).
Assessment of the impacts of point load on river Yamuna at Delhi stretch, by DO-BOD modeling of river, using MATLAB programming
.
International Journal of Engineering and Innovative Technology
2
(
10
),
282
290
.
Korff
,
Y.
,
Daniell
,
K. A.
,
Moellenkamp
,
S.
,
Bots
,
P.
&
Bijlsma
,
R. M.
, (
2012
).
Implementing participatory water management: recent advances in theory, practice, and evaluation
.
Ecology and Society
17
(
1
). .
Larson
,
K. L.
,
White
,
D. D.
,
Gober
,
P.
,
Harlan
,
S.
&
Wutich
,
A.
, (
2009
).
Divergent perspectives on water resource sustainability in a public–policy–science context
.
Environmental Science & Policy
12
(
7
),
1012
1023
. .
Larson
,
K. L.
,
Ibes
,
D. C.
&
White
,
D. D.
, (
2010
).
Gendered perspectives about water risks and policy strategies: a tripartite conceptual approach
.
Environment and Behavior
43
(
3
),
415
438
.
Available at: http://eab.sagepub.com/content/43/3/415.short (accessed November 3 2014)
.
Le Lay
,
Y. -F.
,
Piégay
,
H.
&
Rivière-Honegger
,
A.
, (
2013
).
Perception of braided river landscapes: implications for public participation and sustainable management
.
Journal of Environmental Management
119
,
1
12
.
Available at: http://www.ncbi.nlm.nih.gov/pubmed/23434790 (accessed September 17 2013)
.
Lepesteur
,
M.
,
Wegner
,
A.
,
Moore
,
S. A.
&
McComb
,
A.
, (
2008
).
Importance of public information and perception for managing recreational activities in the Peel-Harvey estuary, Western Australia
.
Journal of Environmental Management
87
(
3
),
389
395
.
Available at: http://www.ncbi.nlm.nih.gov/pubmed/17412478 (accessed May 1 2013)
.
Marks
,
S. J.
,
Komives
,
K.
&
Davis
,
J.
, (
2014
).
Community participation and water supply sustainability: evidence from handpump projects in Rural Ghana
.
Journal of Planning Education and Research
34
(
3
),
276
286
.
Available at: http://journals.sagepub.com/doi/pdf/10.1177/0739456X14527620 (accessed December 26 2017)
.
MoEF
(
2011
).
Annual Report – Indian Ministry of Environment and Forests 2010–2011
. .
Mostert
,
E.
, (
2003
).
The challenge of public participation
.
Water Policy
5
(
2
),
179
197
.
MoWR
(
2012
).
National Water Policy
. .
NCRWC
(
2001
).
Literacy in the Context of Constitution of India
.
Available at: http://lawmin.nic.in/ncrwc/finalreport/v2b1-5.htm (accessed May 8 2017)
.
Pahl-Wostl
,
C.
,
Tàbara
,
D.
,
Bouwen
,
R.
,
Craps
,
M.
,
Dewulf
,
A.
,
Mostert
,
E.
,
Ridder
,
D.
&
Taillieu
,
T.
, (
2008
).
The importance of social learning and culture for sustainable water management
.
Ecological Economics
64
(
3
),
484
495
.
Rani
,
M.
,
Akolkar
,
P.
&
Bhamrah
,
H. S.
, (
2013
).
Water quality assessment of River Yamuna from origin to confluence to River Ganga, with respect to biological water quality and primary water quality criteria
.
Journal of Entomology and Zoology Studies
1
(
6
),
1
6
.
Roca
,
E.
&
Villares
,
M.
, (
2008
).
Public perceptions for evaluating beach quality in urban and semi-natural environments
.
Ocean & Coastal Management
51
(
4
),
314
329
. .
Rohrmann
,
B.
, (
2008
).
Risk Perception, Risk Attitude, Risk Communication, Risk Management: A Conceptual Appraisal
. In:
15th TIEMS Conference Booklet
,
February
. .
Slovic
,
P.
, (
1994
).
Perception of risk
.
Science
14
(
5
),
815
828
. .
Sovacool
,
B. K.
,
D'Agostino
,
A. L.
,
Rawlani
,
A.
&
Meenawat
,
H.
, (
2012
).
Improving climate change adaptation in least developed Asia
.
Environmental Science & Policy
21
,
112
125
.
Available at: http://linkinghub.elsevier.com/retrieve/pii/S1462901112000652 (accessed December 30 2014)
.
Steinwender
,
A.
,
Gundacker
,
C.
&
Wittmann
,
K. J.
, (
2008
).
Objective versus subjective assessments of environmental quality of standing and running waters in a large city
.
Landscape and Urban Planning
84
(
2
),
116
126
. .
TEC
(
2009
).
Yamuna Action Plan – A Bilateral Project of Government of India and Japan on River Conservation
. .
The Indian Express
(
2016
).
Clean-up part III: AAP govt, Centre join hands for Rs 825-cr Yamuna action plan | The Indian Express. The India Express. Available at: http://indianexpress.com/article/india/india-news-india/clean-up-part-iii-aap-govt-centre-join-hands-for-rs-825-cr-yamuna-action-plan-2789723/ (accessed December 17 2017)
.
Tippett
,
J.
,
Searle
,
B.
,
Pahl-Wostl
,
C.
&
Rees
,
Y.
, (
2005
).
Social learning in public participation in river basin management – early findings from HarmoniCOP european case studies
.
Environmental Science & Policy
8
(
3
),
287
299
. .
ToI
(
2015
).
Digital Display Boards to Give Spot Pollution Levels – Times of India
. .
Tunstall
,
S.
, (
2000
).
Public perceptions of the environmental changes to the Thames Estuary in London, UK
.
Journal of Coastal Research
16
(
2
),
269
277
.
Turgeon
,
S.
,
Rodriguez
,
M. J.
,
Theriault
,
M.
&
Levallois
,
P.
, (
2004
).
Perception of drinking water in the Quebec City region (Canada): the influence of water quality and consumer location in the distribution system
.
Journal of Environmental Management
70
(
4
),
363
373
.
Available at: http://www.ncbi.nlm.nih.gov/pubmed/15016444 (accessed August 13 2013)
.
Yufei
,
P.
,
Xueying
,
Z.
,
Min
,
L.
,
Rui
,
Z.
&
Lei
,
S.
, (
2007
).
Transparency and Public Participation in Water Resources Management in China (English)
.
World Bank
,
Washington, DC
. .