Understanding and Enhancing Food Conservation Behaviors and Operations

Abstract

This study explores the dynamics of food conservation behaviors and operations, shifting the lens from the prevalent narrative of food waste reduction to a marketing perspective that emphasizes consumer engagement in sustainable operations. Amidst the rapid urban transformation and economic progress of many countries, this research examines factors influencing individual behaviors toward responsible food operations. It aims to delineate the motivational drivers and deterrents affecting residents’ engagement in food conservation and operations, utilizing an adapted framework based on the theory of planned behavior. We employ partial least squares structural equation modeling to analyze responses from 390 residents. We find that perceived behavioral control, subjective norms, and attitudes significantly enhance intentions to conserve food. Moreover, environmental concerns amplify both attitudes and perceived behavioral control, while green marketing communications and knowledge elevate attitudes, environmental mindfulness, and conservation actions. A connection to nature is substantiated as a reinforcing factor for pro-environmental attitudes and operations. Notably, attitudes are identified as a critical mediator among the examined constructs. This investigation enriches sustainability scholarship by introducing a positive behavior-focused approach, advancing the discourse on sustainable operations. It offers actionable insights for market-driven interventions, policy-making (such as China’s lastest national policies on food security and rural region revitalization in 2024), and educational endeavors to mitigate food wastage and reinforce food supply chain resilience globally.

1. Introduction

About one-third of the world’s greenhouse gas (GHG) emissions are caused by food systems [1,2]. Reducing consumer food waste is critical in achieving the United Nations’ Sustainable Development Goal of halving retail and household food waste by 2030 [3]. Around 931 million tonnes of food can be saved annually if this problem is addressed [4]. This can save over 800 million individuals facing hunger and food insecurity [5]. Also, the food industry would be spared, and most people could be fed around the globe [6,7]. As the world grapples with the pressing challenges of food insecurity, environmental sustainability, and resource conservation, an in-depth examination of individual behaviors becomes imperative [8,9]. Among the various factors influencing global food security, the issue of food wastage stands out as a critical concern [10]. To create sustainable food systems, households must be encouraged to reduce food waste [11]. Studies show that, despite consumers’ environmental concerns, their actual participation in reducing food waste remains significantly low [12]. This can be attributed to the fact that preaching “don’t waste” does not yield the desired results. As a result, a shift to preach “save it”, which is a more positive approach, can coerce the individual to act more positively. In that regard, studies that focus on food-saving behavior rather than food waste reduction are important. Surprisingly, such studies do not exist, and this study advocates for that. In this context, China, with its burgeoning population and rapid socio-economic development, emerges as a crucial focal point for research on residents’ food-saving behavior (FSB).

China, as a highly populous country, plays a pivotal role in shaping global consumption patterns and resource utilization. The nation’s rapid urbanization and rising income levels have significantly altered dietary habits and lifestyle choices among its residents. Consequently, understanding the dynamics of food-saving behavior in China becomes paramount for devising effective strategies to address food wastage on a global scale. Over half of the foods wasted in China can be attributed to the dumping of excess food upon consumption. Also, the consumption phase of the food life cycle contributes significantly to food waste and accounts for up to 17% of the total waste generated in China [13]. The Chinese culture of dining breeds food waste in all areas, such as homes, restaurants, and canteens. This gives rise to the importance of assessing the drivers and barriers to food-saving among Chinese residents to combat the menace. If addressed, approximately 35 million tonnes of food can be saved in China each year [5]. Moreover, research on food-saving behavior is limited in China and the world at large. Among the few existing ones include [14], which researched adolescent food-saving behaviors. A greater part of the related existing literature has focused mainly on food waste behaviors, which is a negative assessment of food conservation practices. Some of these studies include [15,16,17], which delved into household food waste behaviors. In these and other studies, respondents and consumers were assessed on what they were wasting, which may not necessarily yield the desired conservation results. Instead, leading them directly to a positive path to stay conscious of their food-saving habits and activities may produce a better outcome. This study, therefore, identifies this gap and makes a great effort to assess residents’ behavior toward food saving. In this quest, the intricacies and factors that influence them to make better decisions toward food savings can be ascertained.

This research seeks to delve into the multifaceted aspects of residents’ food-saving behavior in China, examining the underlying drivers, cultural influences, and behavioral factors that contribute to or impede sustainable food consumption practices. The study aims to shed light on the intricate interplay between traditional values, bonding with nature, knowledge, and the evolving environmental consciousness among Chinese residents. This study stands out in its approach by focusing on the positive measurement of “food saving” among residents in China, in contrast to the predominant emphasis on “food waste reduction” in the existing literature. While numerous studies have explored interventions and policies aimed at curbing food waste, few have investigated the proactive behaviors that contribute to food conservation at the individual level. By examining the factors driving residents’ food-saving practices, this research fills a critical gap in understanding the positive dimensions of sustainable food consumption. Furthermore, the study aims to delineate the distinct motivations and barriers associated with food-saving behavior, providing a nuanced perspective that goes beyond the conventional focus on waste reduction. This novel approach not only contributes a fresh perspective to the existing discourse on food sustainability but also offers valuable insights for policymakers, educators, and practitioners striving to cultivate a culture of responsible food consumption in China and beyond.

This research can be significant because of its potential to inform policy interventions, educational initiatives, and public awareness campaigns aimed at saving the food industry. By identifying the critical determinants of residents’ food-saving behavior, policymakers, environmentalists, and community leaders can develop targeted strategies to foster a culture of sustainability and responsible food consumption. Furthermore, this research contributes to the broader global discourse on sustainable development by offering insights into the challenges and opportunities associated with changing consumption patterns in a rapidly evolving society. Lessons learned from the Chinese context can inform global efforts to address the growing nexus between food security, environmental conservation, and individual behaviors. In summary, this research endeavors to unravel the intricate web of factors influencing residents’ food-saving behavior in China, providing a nuanced understanding of the challenges and opportunities for fostering sustainable consumption patterns. Through this exploration, the study aims to contribute valuable insights into effective strategies to reduce food wastage, promote environmental sustainability, and enhance global food security.

2. Review of Relevant Literature and Development of Hypotheses

2.1. Theory of Planned Behavior

According to [18], three factors (perceived behavioral control, subjective norms, and attitude) form the theory of planned behavior (TPB). Combining attitude (ATT), subjective norm (SN), and perceived behavioral control (PBC) results in the formulation of behavioral intention. A person’s attitude is deemed as a favorable or negative assessment of how well they perform a specific behavior [19]. It is the product of behavioral beliefs and assessment of results, where views on behavior are personal convictions about the effects of certain behavior. On the other hand, the assessments of the outcome indicate whether the overall assessment of the potential consequences of the activity is favorable or unfavorable [20]. The societal pressure an individual faces to act in a particular way is known as the subjective norm [19]. It is emphasized to be a naturally occurring social component, and its effects are a product of motivational compliance (MC) and normative belief (NB). Normative belief is a person’s perspective of what others value in them in a particular scenario, whereas motivation to comply is someone’s desire to follow these people’s advice [20]. The third aspect, perceived behavioral control, indicates the perceived level of effort needed for a person to behave in a particular way [19]. In this, perceived power (PP) and control beliefs (CB) act as a couple to determine the outcome. CB is a person’s perception of the likelihood that a particular element will exist and can potentially affect (either positively or negatively) the action under consideration. How the person assesses the impact of these circumstances on the conduct in question is regarded as PP [20]. In the TPB domain, a person’s readiness to act in a particular way is shown by their behavioral intention, regarded by [21] as the immediate precursor of behavior. A person’s intention or motivation to engage in a particular conduct will be stronger when they have a more positive ATT, SN, or PBC. The TPB model has been employed in numerous research to assess both intention and behavior related to environmental sustainability.

Food saving, which is being studied, is also regarded as environmentally sustainable behavior [22]. Previous research in this area has made use of the TPB and demonstrated its applicability to a larger stretch of food waste reduction. For instance, ref. [23] used the TPB to investigate leftover sharing behaviors. Also, ref. [24] used the theory to assess food waste behavior in Turkey. Moreover, ref. [25] inculcated emotions into the TPB to investigate their impact on food waste reduction. Ref. [26] also investigated household food waste in Brazil using the TPB and many others. Unfortunately, the literature on food-saving behavior is non-existent. Among these studies conducted in the food waste reduction space, the majority of them found that attitude, PBC, and SN have a positive influence on intention and behavior. In that domain, the study assumes that when residents make a favorable assessment of food saving, feel socially pressured to save food, and believe they have the control to do so, they will save food. As a result, the study develops the following hypotheses using the antecedents of TPB.

Hypothesis 1 (H1). 

Attitude positively influences residents’ food-saving intentions.

Hypothesis 2a (H2a). 

Subjective norm is a positive predictor of attitude.

Hypothesis 2b (H2b). 

Subjective norms have a positive influence on residents’ food-saving intentions.

Hypothesis 3 (H3). 

PBC positively influences residents’ food-saving intentions.

2.2. Model Development

The TPB model has been deemed simple by many scholars, which affects its predictability potential and ability to capture a broader range of behaviors. Specifically, it has been argued that it has a non-cognitive constraint, even though it is frequently employed to predict intention and conduct. Scholars contend that because of this, its potential can be strengthened by including other variables [27,28]. Recent psychological research that expanded the theory has found that the TPB’s predictive power increases when additional constructs are included [29,30]. Likewise, this study believes that an analysis of this nature requires environmental variables and constructs that will meet the objectives of the research and hence introduced three additional variables (green information and knowledge, environmental concern, and natural bonding) into the TPB to assess residents’ food-saving behaviors. This also forms a part of the study’s novelty as the theory, together with the environmental variables, has not been collectively used to assess food-saving behavior in the literature. Therefore, it will serve as a theoretical foundation for future studies. These variables are described further.

2.3. Green Information and Knowledge

In this study, “green information and knowledge” is defined as the information and understanding that citizens have on the adverse effects of environmental issues and the potential mitigation activities and strategies that may be used to promote environmental sustainability [31]. Considerations include environmental conditions, climate change, environmental ideologies, and the ecological implications of food waste [32]. Knowledge and information on sustainability aid in sustainable living [33] and achieving SDGs. Green knowledge and information continue to be crucial yet understudied links that are currently poorly understood [34]. If residents are poorly informed, have incomplete knowledge, or are unaware of environmentally friendly options, then trying to act in an environmentally friendly way could be pointless [32]. People frequently misunderstand the knowledge they currently possess since sustainability knowledge is still evolving. Studies show that a lack of understanding or information about it may stampede a particular activity’s acceptance. For instance, ref. [35] stressed that the psychological barrier to the adoption of electric vehicles can exist when individuals lack knowledge and understanding about them. Hence, the roles of green information and knowledge should be tactically embedded within the TPB and other frameworks to unravel novel theoretical insights into the adoption of food saving.

Studies including [32,34,36,37] have demonstrated the positive effects of green information and knowledge on attitudes, behavior, and beliefs regarding ecologically friendly activities. According to [31], people adopt good energy-saving behaviors when they are well-informed. Similarly, those who are more knowledgeable about the environment exhibit higher levels of pro-environmental behavior, such as wasting less food, frequently using public transportation, sorting waste, and purchasing goods with less disposable packaging [30,38,39,40]. Scholars have noticed the positive effects of green information and knowledge in influencing intentions and behavior [31,32,34]. Drawing upon these arguments, the study suggests that the green information and knowledge residents possess will have a direct impact on their attitude and food-saving behavior. The following is therefore hypothesized:

Hypothesis 4a (H4a). 

Green information and knowledge have a positive influence on attitude.

Hypothesis 4b (H4b). 

Green information and knowledge have a positive influence on environmental concerns.

Hypothesis 4c (H4c). 

Green information and knowledge have a positive influence on citizens’ food saving.

2.4. Environmental Concerns

The full spectrum of views, feelings, knowledge, attitudes, beliefs, and behaviors relating to the environment has been dubbed “environmental concern” [41]. Environmentally conscious people are more inclined to try to alter their habits and safeguard the environment [32]. It plays a crucial role in defining people’s intentions and is regarded as one of the factors most strongly affecting consumer behavior [42]. Ref. [43] asserted that citizens with extreme concerns for the environment are bound to act pro-environmentally. Some of these include sustainable behaviors [44], sustainable consumption [32], electric motorbike switching [45], organic food purchases [46], and food waste [17,23,26]. Also, citizens’ environmental concerns can affect how they consume food [47]. Based on past studies, the study believes that when people are much more concerned about the environment, they will develop a positive attitude and food-saving behavior. Also, this awareness will influence their control over saving food. Thus, the following hypotheses are developed:

Hypothesis 5a (H5a). 

Concern for the environment positively influences attitude.

Hypothesis 5b (H5b). 

Concern for the environment positively influences PBC.

2.5. Natural Bonding

People’s love for the tangible world and the natural wonders they experience can be regarded as natural bonding [48]. When people love nature, they want to do everything in their capacity to keep it as such and not destroy it. It is expected that a person will form some emotional connections with nature because it is an essential part of their living environment [30,49]. There is evidence in the literature that connections made between persons who are emotionally tied to one location that has natural characteristics are more dependable while they are in that location as opposed to unconnected regions and are unnatural. The “green mind” concept contends that people who take good care of their bodies, minds, and selves are more likely to have a stronger connection to nature and treat it with care [50]. Environmentally responsible behavior is significantly impacted by natural bonds [51]. According to studies, those who feel an emotional connection to nature are more likely to undertake sustainably responsible behavior [52,53], and such individuals are dedicated to protecting the natural environment they grew up in or were exposed to [54]. Refs. [30,55] found that natural bonding impacted recycling behavior. Ref. [56] also discovered that park visitors who clean up other people’s trash in the park are thought to have a close connection to nature. Similar to these assertions, people who relate themselves to the environment and aim to protect it would have positive attitudes and reduce food waste that harms nature consequently. Therefore, these hypotheses are developed.

Hypothesis 6a (H6a). 

Natural bonding positively influences attitudes toward food savings.

Hypothesis 6b (H6b). 

Natural bonding positively influences food-saving behavior.

2.6. Intention Behavior

Following the TPB reasoning, the study forecasts that residents’ food-saving intention positively predicts their food-saving behavior. The study believes that food-saving behavior will be more prevalent if there is a strong intention to do so. Hence, the following hypothesis is proposed.

Hypothesis 7 (H7). 

Residents’ food-saving intention has a positive influence on food-saving behavior.

The study is based on these developments to develop a comprehensive framework that guides its assessment, as seen in Figure 1.

3. Data and Methods

3.1. Data Collection and Sampling

Residents of Jiangsu province in China were selected for this study. The province presents a unique setting for such studies. It boasts of indigenous Chinese culture, food, and eateries. As a result, food waste is a significant concern for the people and government of that area. In 2019, the province contributed 18.1 million tons to China’s total discharged waste of 242.06 million tons [57], including food waste. As a result, the province has prioritized waste management from households to industry and national level. The strategies aimed at creating zero-waste cities across Jiangsu province begin with an assessment of its residents’ food-saving behavior. These and other reasons necessitated the choice of this area in this study, and the results will make a meaningful contribution toward it. Before the final survey, the study performed a pretest to assess the reliability of the data and constructs. Thirty-five respondents who frequently use WeChat, a popular Chinese App, were surveyed for the pilot study. The pretest made sure the questionnaire’s interpretation and flow were correct, as well as its reliability and validity. Afterward, the study used www.sojump.com (accessed on 15 Febuary 2023) to collect data on food saving from residents of Jiangsu province. The platform provided access to a broad audience pool, allowing us to recruit participants from diverse backgrounds relevant to our research objectives. This platform and method is a reliable and convenient data collection tool used by many scholars [58]. The platform has millions of users across China, with more than one million users participating in daily surveys. The study opted for the paid package to ensure accurate and more valid responses. A stratified random sampling method was employed to ensure randomness in participant selection. Specifically, potential participants were categorized based on relevant demographic or contextual factors (e.g., age, gender, and geographic location). They then randomly selected participants from each stratum to ensure representation across different groups. These questions require them to give their opinions on a 5-point Likert scale, anchoring “strongly disagree” and “strongly agree”. Respondents’ consent was sought before taking data. They were also assured that their data were solely for research works. A total of 400 responses were targeted, but 390 valid responses were received and used for the study’s analysis. To figure out the necessary sample size, we applied Equations (1) and (2) below. The 390 survey responses that were received are appropriate for the study’s analysis, according to the formula’s 384 required samples.

$$\mathrm{Required} \mathrm{sample} \mathrm{size} (\mathrm{N})=\frac{Z\times SD\times \left(1-SD\right)}{{ME}^2}$$

(1)

$$\mathrm{N}=\frac{{\left(1.96\right)}^2\times 0.5\times \left(1-0.5\right)}{{\left(0.05\right)}^2}=384$$

(2)

where Z (z-score²) refers to the selected confidence level, N indicates the projected sample size, SD is the standard deviation, and ME denotes the margin of error.

3.2. Constructs Measurements

To develop more accurate and robust measurement items to suit the study’s objectives, past studies served as the foundation for designing research measurements. This created a foundation for developing a more suitable measurement that can capture the variable under study. The TPB measurements (attitude (5 items), subjective norms (3 items), and PBC (3 items)) were adapted from [59,60] and captured questions related to respondents’ attitudes toward social norms and control regarding food-saving behavior. On the additional variables, the study also sought studies that used them in different and similar areas and modified them to suit the study’s objectives. Specifically, environmental concerns had four items adapted from [61] and captured questions related to respondents’ consciousness of the environment and how it can influence their decisions toward food savings. Five (5) items were also used to measure green information and knowledge and were adapted from the works of [60,62]. This also captured the knowledge and information respondents have regarding food savings, its benefits, and its consequences. Natural bonding had four items capturing how respondents’ relationship with nature influences their food-saving behavior and were modified from the works [30]. Finally, food-saving intentions (FSI) and behavior were adapted from the works of [3,26,47], measuring how respondents are willing to save food and the possible food-saving practices they have been engaging in. These can be found in Appendix A. Overall, three components make up the questionnaire. The responders were first given a quick greeting, information about the study’s goal, and a guarantee of anonymity. The demographic profile questions in the second portion covered 6 characteristics such as gender, age, marital status, degree of education, financial status, and family size. All of the research constructs were included in the last section.

3.3. Statistical Analysis

PLS-SEM (partial least squares structural equation modeling) was utilized for the study’s statistical analysis, as it is frequently used in modern research. It has proven fruitful in predicting and determining causality among variables that employ the kind of data used in this study. PLS-SEM eliminates the limitation on the assumptions made regarding data and its normal distribution works very well when dealing with data that are not normally distributed [63]. Moreover, it conducts two major assessments, namely, the measurement model and the structural model [64]. As a result, it was selected to aid the study objectives, and the software used was SmartPLS 3.0.

4. Study Results and Discussion

4.1. Respondents’ Information

Table 1. Respondents’ demographics.

Profile		Freq.	%
Gender	Male	165	42.3
	Female	225	57.7
Age	18–28	77	19.7
	29–38	139	35.6
	39–48	117	30
	49–59	56	14.4
	59 and above	1	0.3
Marital status	Single	84	21.5
	Married	273	70
	Divorced	33	8.5
Education	High School and Below	50	12.8
	Junior College	91	23.3
	Vocational	219	56.2
	Bachelor	30	7.7
Income	Postgraduate	73	18.7
	Below 2500 RMB	33	8.5
	2500–5000 RMB	122	31.3
	5001–7500 RMB	115	29.5
	7501–10,000 RMB	47	12.1
Family size	2 or less	6	1.5
	3	96	24.6
	4	191	49
	5 or more	97	24.9
	Total	390	100

displays the demographics. The 390 respondents who answered the questionnaire made up the total sample size. Most of the respondents were women, representing 57.7% of the total respondents, and men formed 42.3%. Age-wise, 65% of the participants were between 29 and 48 years, 19.7% were those aged between 18 and 28 years, and those between 49 and 59 years were 14.4%. Seventy percent of the participants were married, compared to 21% who were single and 8.5% who were divorced. Amongst the people we interacted with, about 56% had vocational education. Regarding socio-economic characteristics, about 60.8% of them earn between 2500 and 7500 RMB. As for the family size, 288 respondents belonged to a family with at least four members (73.9%), 96 belonged to a family of three, (24.6%), and 6 belonged to a family with two members or less (1.5%).

4.2. Testing Measurement Models

Table 2. Construct validity and reliability.

Variable	Items	Factor Loadings (FL)	Cronbach’s Alpha (CA)	Composite Reliability	rho_A	Average Variance Extracted (AVE)
Attitude	ATT1	0.806	0.907	0.907	0.931	0.729
	ATT2	0.864
	ATT3	0.868
	ATT4	0.871
	ATT5	0.857
Environmental concern	EC1	0.795	0.862	0.866	0.906	0.707
	EC2	0.868
	EC3	0.842
	EC4	0.857
Green Information and Knowledge	GIK1	0.753	0.858	0.869	0.898	0.638
	GIK2	0.863
	GIK3	0.781
	GIK4	0.820
	GIK5	0.772
Natural bonding	NB1	0.853	0.850	0.852	0.899	0.689
	NB2	0.814
	NB3	0.842
	NB4	0.811
Food-saving behavior	FSB1	0.883	0.883	0.889	0.919	0.740
	FSB2	0.815
	FSB3	0.872
	FSB 4	0.871
Perceived behavioral norms	PBC1	0.817	0.811	0.818	0.888	0.725
	PBC2	0.881
	PBC3	0.856
Food-saving intention	FSI1	0.810	0.847	0.850	0.897	0.685
	FSI2	0.847
	FSI3	0.811
	FSI4	0.842
Subjective norms	SN1	0.768	0.708	0.719	0.836	0.629
	SN3	0.787
	SN3	0.823

Note: Cronbach’s alpha = CA, average variance extracted = AVE, factor loading = FL, and composite reliability = CR.

exhibits the results of CR (composite reliability), CA (Cronbach’s alpha), Rho A, AVE (average variance extracted), and FL (factor loadings). These metrics are employed to direct and confirm the selection of constructs and the accuracy of survey data. The CA and CR values have values higher than 0.7, exceeding the 0.7 cutoff limit (Hair et al., Sarstedt 2016), demonstrating an extreme level of reliability. All AVE values that assess the constructs’ convergence validity are significantly higher than the minimum value of 0.5 suggested by [65]. The FL statistics exceed the crossing-loading with other constructs and are above the 0.6 criterion [66].

The constructs’ discriminant validity was evaluated according to [67] and the correlation matrix according to [64]. The square roots of the AVEs for every construct are greater than the off-diagonal components in their corresponding rows and columns, as can be seen in

Table 3. Results of discriminant validity.

Heterotrait–Monotrait Ratio (HTMT)—Matrix
Items	ATT	EC	GIK	NB	FSB	PBC	FSI	SN
ATT
EC	0.477
GIK	0.373	0.586
NB	0.364	0.537	0.292
FSB	0.347	0.509	0.371	0.605
PBC	0.261	0.359	0.483	0.277	0.296
FSI	0.483	0.339	0.585	0.532	0.493	0.329
SN	0.307	0.343	0.304	0.420	0.301	0.226	0.321
Fornell–Larcker criterion
Items	ATT	EC	GIK	NB	FSB	PBC	FSI	SN
ATT	0.854
EC	0.423	0.841
GIK	0.334	0.511	0.799
NB	0.321	0.460	0.252	0.830
FSB	0.310	0.447	0.327	0.527	0.860
PBC	0.225	0.302	0.406	0.232	0.252	0.852
FSI	0.424	0.970	0.507	0.453	0.431	0.277	0.828
SN	0.251	0.270	0.242	0.327	0.239	0.176	0.252	0.793

Note: Attitude = ATT, GIK = green information and knowledge, subjective norm = SN, environmental concern = EC, natural bonding = NB, perceived behavioral control = PBC, food-saving intention = FSI, and food-saving behavior = FSB.

. These values are boldfaced along the diagonal. Particularly, each construct had good discriminant validity, as the square root of the AVE indicated. As shown in Table 3, another test for discriminant validity was the heterotrait–monotrait ratio (HTMT). The HTMT is a more reliable standard for discriminant validity according to [64]. If the result is below 0.9, it is regarded as valid. Strong discriminant validity was shown as the highest at 0.605.

4.3. Structural Model Assessment

The structural model’s predictive abilities were assessed by examining the coefficients of determination using the SmartPLS 3.0 software. It is measured by the coefficient of determination (R² value). R² values of at least 0.20 are considered strong in fields like consumer behavior [64]. The results reveal an overall explanatory power in the current study, and the constructs explain 23.1% and 33.8% of food-saving purchasing intention and behavior, respectively, as displayed in

Table 4. Model fit.

Measurement Items	Threshold	Outcomes
SRMR	<0.08	0.052
NFI	>0.80	0.877
R2 (GL)		0.338
R2 (GLI)		0.231

and Figure 2. This indicates a considerable predictive power together with the other variables. Once more, the study used model fit indices to verify the model’s dependability. The SRMR (Standardized Root Mean Square Residual) and the NFI (normed fit index) are within the acceptable range at 0.052 and 0.877, respectively. The sample data and the model fit very well. This information is present in Table 4.

4.4. Hypotheses Results and Validation

Table 5. Hypotheses testing.

Hypothesis	Pathway	Coefficients	Standard Deviation	T Statistics	p Values	Confidence Intervals		Outcome
						2.5%	97.5%
H1	ATT -> FSI	0.353	0.043	8.078	0.000 ***	0.265	0.438	Validated
H2a	SN -> ATT	0.106	0.049	2.096	0.036 *	0.008	0.205	Validated
H2b	SN -> FSI	0.135	0.047	2.823	0.005 **	0.039	0.228	Validated
H3	PBC -> FSI	0.177	0.046	3.778	0.000 ***	0.084	0.264	Validated
H4a	GIK -> ATT	0.142	0.054	2.637	0.008 **	0.033	0.245	Validated
H4b	GIK -> EC	0.514	0.037	13.967	0.000 **	0.440	0.582	Validated
H4c	GIK -> FSB	0.132	0.048	2.780	0.005 **	0.037	0.223	Validated
H5a	EC -> ATT	0.264	0.057	4.628	0.000 ***	0.153	0.375	Validated
H5b	EC -> PBC	0.306	0.045	6.703	0.000 ***	0.216	0.393	Validated
H6a	NB -> ATT	0.131	0.056	2.337	0.019 *	0.021	0.243	Validated
H6b	NB -> FSB	0.416	0.040	10.314	0.000 ***	0.337	0.494	Validated
H7	FSI -> FSB	0.178	0.053	3.346	0.001 ***	0.073	0.281	Validated

Note: * p < 0.05., ** p < 0.01, and *** p < 0.001. Attitude = ATT, GIK = green information and knowledge, subjective norm = SN, environmental concern = EC, natural bonding = NB, perceived behavioral control = PBC, food-saving intention = FSI, and food-saving behavoir =FSB.

shows the results of the hypotheses. Attitude (β = 0.353, p < 0.001), subjective norm (β = 0.135, p < 0.05), and perceived behavioral control (β = 0.177, p < 0.001) had a significant favorable influence on food-saving intention, validating H1, H2b, and H3, respectively. Subjective norms (β = 0.106, p < 0.05), green information and knowledge (β = 0.142, p < 0.01), environmental concern (β = 0.264, p < 0.001), and natural bonding (β = 0.131, p < 0.05) had a positive significant effect on attitude, validating H2a, H4a, H5a, and H6a respectively. Also, H4b was validated as green information, and knowledge significantly and positively influenced environmental concern (β = 0.514, p < 0.001). The positive relationship between environmental concern and PBC was established (β = 0.306, p < 0.001), which validated H5b. Food-saving behavior was found to be significantly influenced by green information and knowledge (β = 0.132, p < 0.01), natural bonding (β = 0.416, p < 0.001), and food-saving intention (β = 0.178, p = 0.001). These findings validated H4c, H6b, and H7. In conclusion, all the hypotheses developed for the study have been confirmed. They can also be confirmed in Figure 2.

4.5. Mediating Roles

We conducted a mediation role study to determine the constructs that partially mediate others. This is to ensure the indirect effects of these variables on the others. Taking inferences from the guidelines in [68] regarding mediating roles, these roles played by major constructs are exhibited in

Table 6. Mediating effects.

Pathway	Original Sample (β)	T Statistics	p Values	Confidence Intervals
				2.5%	97.5%
EC -> ATT -> FSI	0.092	3.110	0.002 **	0.042	0.160	+√
GIK -> EC -> ATT	0.134	4.277	0.000 ***	0.077	0.200	+√
GIK -> ATT -> FSI -> FSB	0.009	1.976	0.048 *	0.002	0.019	+√
NB -> ATT -> FSI	0.046	2.330	0.020 *	0.007	0.086	+√
EC -> PBC -> FSI	0.053	2.458	0.014 *	0.018	0.101	+√
GIK -> EC -> PBC -> FSI	0.027	2.276	0.023 *	0.009	0.055	+√
GIK -> EC -> PBC	0.154	5.336	0.000 ***	0.103	0.216	+√
GIK -> ATT -> FSI	0.050	2.528	0.012 *	0.011	0.090	+√

Note: * p < 0.05., ** p < 0.01, and *** p < 0.001. Attitude = ATT, GIK = green information and knowledge, environmental concern = EC, natural bonding = NB, perceived behavioral control = PBC, food-saving intention = FSI, food-saving behavior = FSB, and +√ = positive partial mediation.

. Attitude is regarded as one of the major constructs in this study, and the mediation study confirmed its importance in mediating other variables. The results revealed that attitude partially mediates people’s concerns for the environment and their food-saving intention. It mediated GIK and FSI, as well as NB and FSI. Moreover, together with FSI, they partially mediate GIK and FSB. The control residents have over their food-saving behavior was also found to partially mediate EC and FSI. Again, PBC mediated GIK and FSI. Finally, people’s concern for the environment played a partial mediation role between GIK and PBC.

4.6. Discussion

In this study, green information and knowledge, environmental concerns, and natural bonding were added to the TPB to provide new evidence of food-saving intention and behavior. According to the results, attitude had a very strong positive influence on food-saving intention, signifying that when people have a favorable assessment of saving food and its contribution to environmental sustainability, they are more willing to save food. The attitude–intention relationship has been established in various studies [69,70,71,72] and has been confirmed in this study. Also, their subjective norms were found to influence attitudes and intentions. This impact is such that people are willing to save food based on recommendations from loved ones or close ones. If the people they consider in their lives affirm the relevance of saving food to promote environmental sustainability, their attitudes will be positively influenced together with their intentions. This corroborates with past studies [69,70] that affirmed the significance of SN in citizens’ ATT and intentions. Regarding the last TPB construct, the outcome shows that PBC influences food-saving intention positively. This effect was among the strong ones detected in this study. This indicates that when people have no difficulty saving food, they will develop an intention to save food. Various studies have confirmed that perceived behavioral control is a good predictor of intention [73,74]. This is confirmed in this study as it depicted a significant positive result. All the TPB constructs showed a substantial impact on food-saving intention. This confirms the applicability of TPB in determining the residents’ food-saving intention in the Chinese context.

Green information and knowledge are identified to influence environmental concerns, attitudes, and food-saving behavior. The result is an indication that when people receive the right information and gain insights about food savings and environmental sustainability, their concerns for the environment will improve. Also, when they gain these insights and become aware of the importance of saving food to achieve SDGs, their attitudes toward it will improve. This will eventually influence their behaviors as shown in this study. This finding exposes the practicality and importance of information and knowledge in determining food-saving behaviors. This finding can also be attributed to the province’s exposure to education and knowledge as it boasts of being China’s educational hub. As a result, the knowledge and information residents have regarding SDGs are heightened, contributing to their positive impact on their food-saving behavior. Other studies may not yield the same results due to limitations in information and knowledge about environmental sustainability. This corresponds to past studies [32,34,37] that found information and knowledge to influence green behaviors. Environmental concerns are identified in this study as significantly influencing attitudes toward food-saving and PBC. It is a clear indication that people are concerned about the environment, which is influencing their decision to save food. Due to these concerns, they form these attitudes, beliefs, and abilities toward food saving. This is confirmed in the mediation analysis as PBC, and attitude provided a partial mediation role between environmental concerns and intention. This has affirmed previous studies [43,59,60,75] that revealed that awareness and consciousness of the environment can positively affect ATT, PBC, and intentions.

In this study, natural bonding was an important construct that aimed to establish whether or not bonding with nature or the environment can influence people’s attitudes and food-saving behaviors. The results revealed that natural bonding positively influences attitudes toward food-saving. Also, the highest and strongest connection among the various relationships of the study was found between natural bonding and food-saving behavior. This indicates that residents love and enjoy nature and continuously want to see it as sustainable, hence saving food. In other words, not wasting food green reminds them of the importance of keeping nature green and improving environmental sustainability through their positive behavior. This finding is in line with studies [30,53] that found natural bonding to be a great predictor of green behaviors. Finally, the study intended to establish not only food-saving intentions but also actual behavior. According to the results, residents’ intentions translate into their actual food-saving behavior. The greater or stronger the intention to save food, the more they save it. This relationship was a significant one, indicating a strong green intention–behavior relationship.

5. Conclusions and Policy Implications

5.1. Conclusions

In this study, samples were taken from residents in China to assess their food-saving behavior using the TPB, environmental concerns, green information and knowledge, and natural bonding as antecedents. The findings revealed that ATT, SN, and PBC significantly influence food-saving intentions. Environmental concerns positively influence both attitudes and perceived behavioral control. Green information and knowledge were identified to influence attitudes, environmental concerns, and food-saving behavior significantly. It was also established that a bond with nature positively influences attitude and food-saving behavior. Finally, the pivotal role of attitude is confirmed in this study as it partially mediates environmental concerns and intentions, green information and knowledge and intentions, and natural bonding and intentions. In the food-saving space, this study provides important findings that will concretize and enhance future studies in the area. These results will make a meaningful impact on the campaign in Jiangsu province toward waste management and the overall Chinese dream. Also, policymakers will benefit from its findings and implications outlined to promote food-saving and environmental sustainability.

5.2. Theoretical and Practical Implications

5.2.1. Theoretical Implications

Theoretically, the study outlines these implications. First, the role of TPB in food-saving behavior has been established in the Chinese context. Due to generalizability issues, studies in different regions are required for new evidence, and this study has fulfilled that requirement. As a result, it will provide a foundation for future studies that intend to assess different areas using the TPB or other variables. It also strengthens the role TPB plays in green behaviors. Second, the importance of natural bonding has been established in this study. Natural bonding is a variable that has gained popularity in recent pro-environmental studies. Scholars can then incorporate its role in TPB and food-saving behavior into other theories for new evidence. Finally, the pivotal mediation role of attitude has been established in this study. Attitude as a variable in the TPB domain is known to influence intention. However, it has justified its partial mediation role between these major environmental antecedents and established its relationship with food saving. This can encourage many future studies to test other variables according to their research objectives.

5.2.2. Practical and Managerial Implications

Based on the findings in this study, some critical actions and measures can be taken to boost food saving and improve environmental sustainability. To begin with, attitude plays a pivotal role in this study. It influenced intention as well as mediated natural bonding, environmental concerns, and green knowledge and information. This is evidence of the impact citizens’ attitudes can have on food saving. A negative attitude will result in food waste and vice versa. As a result, policymakers must prioritize improving citizens’ attitudes toward food saving. This must be carried out alongside campaigning for SDGs. The SDGs are usually centered on higher-level governance, leaving the majority of citizens at the lower level and in rural areas with no or limited knowledge about the course. The more citizens are familiar with SDGs, the more they will form good attitudes toward green living. This familiarization can be achieved by decentralizing SDGs to the lowest echelons in societies. Most campaigns are carried out using contemporary methods which are not available in rural areas. When they become acclimatized to the global green campaign, they will develop a positive attitude going forward that will affect their future lifestyles in terms of food saving.

Furthermore, subjective norm is one of the most valuable variables in this study because of its impact on attitude and intention. The results indicated the importance of families and dear ones in one’s decision toward food saving. Policymakers can hop on this advantage to boost citizens’ interest in food-saving. Most citizens look up to outstanding personalities in society and endorse any course these people endorse. These personalities can be identified to champion campaigns that are food-saving centered. These people include movie stars, musicians, veterans, and famous people with a great fanbase. They must be trained and engaged to present these campaigns in a nonchalant way, which will make people feel it requires less effort to save food. They can create trends on social media apps with their fans. When people feel it is a difficult task to do, they will hesitate. But when they realize it can be a daily life activity without stress or much effort, they will also hop on the trend.

Moreover, the study established the importance of knowledge and information in saving food. Most people waste food because they have an abundance of it and have no idea of the consequences of wasting food. Also, most people are not aware of the global hunger issues and the effects of food waste on the environment. This presents an avenue for policymakers to embark on extensive food-saving campaigns. Citizens must be well informed and educated on the SDGs and their importance to the world at large, especially Goal 2 (zero hunger). When people obtain the overall knowledge and information about the need to save food to increase sustainability, they will do so. This will go a long way in affecting their daily dietary needs to practice sustainable and responsible consumption as stipulated in SDG-12. They will subconsciously develop a habit and will do so in the future.

Finally, due to the critical role natural bonding is found to play in food-saving behaviors, producers must produce with nature in mind. People who connect with nature are very particular about designs and symbols that relate to them. In that regard, producers must collectively help the government in this campaign by developing packages that enable customers to eat or consume responsibly. Inscriptions on how to save food or save for later use can be imprinted on packages. Also, the packages can be designed as easy self-storage packages that can be saved for some time if consumers cannot consume them all at once. This will increase consumption consciousness and promote food saving among consumers.

5.3. Limitations and Future Research Direction

The study believes that future research can take note of these outlined limitations for novel results. Some of the identified limitations in this study are as follows. First, respondents were selected from Jiangsu Province in China and may have given their opinions based on background, culture, and local experience. Except for regions that share the same characteristics as China or Jinagsu province, generalizability may be an issue for others. Future studies may use other regions for new findings. Second, the study focuses on urban regions. Future studies can be conducted on rural regions to help with regional food supply chain security and regional economic revitalization. For example, since the beginning of 2024, China has adopted new policies for rural regions on food supply chain security, village governance, and rural economic development. Third, although the sample used in this study is acceptable, we believe a much higher sample size can impact the result. Based on this, future studies with a higher budget can expand the sample size in a country with a higher population size for new evidence. Finally, the study believes other theories and constructs can be inculcated into this study to provide new findings in the area, and future studies can consider them.