Farmers’ Adoption of Agricultural Nature-Based Solutions in Northeast China: An Extended Theory of Planned Behavior Approach

Abstract

The academic field has limited studies on implementing nature-based solutions (NbSs) in Chinese agriculture, especially regarding households’ and farmers’ attitudes toward the acceptance of agricultural NbS technologies. Based on this, this study aims to fill this knowledge gap by examining the factors behind farmers’ agricultural practices of NbS adoption in Northeast China. Accordingly, this paper constructs an extended theoretical framework of planned behavior (TPB) and empirically tests it using partial least squares structural equation modeling (PLS-SEM) on data from 298 farmers in Inner Mongolia. It was found that incorporating institutional trust (IT) into TPB improves the prediction of Chinese farmers’ intention to de-farm behavior. Unlike previous studies, this study identified two indirect pathways influencing farmers’ intention to adopt agricultural NbS technologies: from institutional trust through attitude to intention, and from institutional trust through perceived behavioral control to intention. The findings broaden the research perspective in this area, provide empirical evidence for the application of NbSs in agricultural practices in China and other developing countries, and demonstrate in practice that NbS has been promoted as an essential tool to maximize nature’s ability to provide ecosystem services and create direct economic benefits for farmers through increased yields and reduced costs.

1. Introduction

As the global population has increased and economic prosperity has improved in recent decades, a significant gap has emerged between the current food production and the amount needed to feed everyone in 2050 [1]. On the one hand, China’s agricultural productivity has seen rapid growth in recent decades, largely due to the intensive use of agricultural inputs, particularly chemical products like fertilizers [2]. Compared to the 1960s, the average yield of China’s major grain crops—maize, rice, and wheat—has nearly doubled, reaching 5845 kg per hectare by 2023 [3], which is 20.70% higher than the world average [4]. Fertilizers enhance crop growth by providing essential nutrients such as nitrogen, phosphorus, and potassium, while insecticides and herbicides help prevent pests, minimize losses, and ensure both yield and quality in agricultural production [5]. On the other hand, the cornerstones of agricultural production systems, healthy land and soil, are already under immense pressure. Indeed, around 33% of the Earth’s land area and up to 52% of its agricultural land are experiencing varying degrees of degradation, leading to abandonment by land managers on millions of hectares annually [6,7]. Additionally, the global agri-food system is responsible for approximately one-third of total anthropogenic greenhouse gas (GHG) emissions [8,9]. In 2020, the breakdown of emissions within agri-food systems revealed that nearly half originated from farm-gate activities (7.4 Gt CO₂eq), followed by pre- and post-production processes (5.6 Gt CO₂eq), and land-use change (3.1 Gt CO₂eq) [10]. These perpetuate a feedback loop that worsens the climate crisis and amplifies numerous risks such as loss of biodiversity and food security [11,12,13,14]. In China, excessive utilization of agricultural chemicals has resulted in pollution from non-point sources and severe environmental challenges [15,16,17,18,19]. Furthermore, the overuse of chemical fertilizers and the lack of crop rotation techniques have led to severe soil degradation, reduced soil organic carbon (SOC) sequestration, and decreased crop yields [20,21,22,23,24].

Nature-based solutions (NbSs) are increasingly being advocated in both research and policy circles as one of the possible means to support future food systems and facilitate a sustainable shift in agricultural production [25]. NbSs encompass strategies aimed at protecting, managing, or restoring natural environments while also enhancing ecosystems to address societal challenges. These approaches include utilizing nature for adaptation purposes, implementing natural climate solutions, and incorporating green infrastructure [26]. The primary goal of all NbS initiatives is to enhance nature’s capacity to provide essential ecosystem services that effectively address human concerns such as adapting to climate change, ensuring food security, and mitigating the risks associated with disasters [27]. Many NbSs are directly applied in agricultural production and management, such as conversion agriculture and cropland nutrient management. Farmers or producers primarily implement these technologies, which may bring direct economic benefits, such as increased production or reduced costs, along with broader social benefits [24,28]. As shown in Figure 1, agricultural NbS technologies can play a crucial role in ensuring food security [29], addressing climate change [30], and enhancing natural biodiversity [31] in the long term.

The emergence of the NbS concept in China’s mainstream research and policies is very recent. The promotion of environmentally friendly agriculture has been actively supported by the Chinese government, which has implemented a range of policies and guidance programs as part of its national agricultural planning and development. These initiatives encompass measures such as optimizing the use of chemical fertilizers and pesticides, adopting sustainable planting and raising cycles, and implementing an action plan to preserve farmland on black soil in north-east China [32,33,34]. Since 2019, China has started introducing NbSs in scientific research on research and environment management [35], and researchers identified universal implementation key elements for NbS best practices worldwide based on certain Chinese NbS cases [36,37]. In addition, China’s Ministry of Science and Technology initiated the ‘EU-China Nature-based Solutions for Nutrient Management in Agriculture and Sustainability Transformation’ project in 2023 in collaboration with the Horizon Europe program. Despite the efforts in policy advocacy, there is currently little research analyzing the application of NbS in China’s agriculture, especially from the perspective of households’ and farmers’ intentions towards agricultural NbS practices [38]. Insufficient research on NbS in China can lead to a disconnect between the concept and practice of green and sustainable agriculture [39]. Additionally, existing research lacks the incorporation of institutional trust in government, scientists, etc. into studies of the diffusion of green agricultural technologies such as NbS. This limits the integration of micro-farmer socio-economic with macro-ecological and environmental fields and weakens China’s discourse in the field of NbS on the international stage.

Therefore, it is crucial to assess the level of interest among Chinese farmers in adopting NbS technologies for agricultural production and determine the factors that impact their willingness. Social psychology research techniques are extensively employed to ascertain the driving forces behind human behaviors and their influencing factors, enhance comprehension of farmers’ decision-making processes, and provide guidance for policy formulation [40]. This research contributes to the existing literature by examining the causal connections between TPB constructs concerning farmers’ intentions to adopt NbS technology in maize and soybean cultivation, while also investigating the significance of institutional trust within this process. By adopting this approach, a more comprehensive comprehension of farmers’ decision-making is achieved, thus providing valuable insights for policymakers aiming to promote NbS further in China and other developing nations. Hence, the objective of this research is to address the existing gap in studies by investigating the factors that drive Chinese farmers to embrace NbS in agricultural production. Specifically, this study aims to make contributions in three key areas: (1) evaluating the suitability of the theory of planned behavior for analyzing farmers’ intentions towards adopting NbS technology in agriculture; (2) enhancing the explanatory power of the original TPB model by incorporating variables related to institutional trust and identifying key factors influencing farmers’ readiness to embrace NbS technology; and (3) exploring how institutional trust indirectly influences intentions by investigating its effects on perceived behavioral control and farmer attitudes.

The following sections of this paper are categorized into five segments: The subsequent section provides an introduction to the theoretical basis and suppositions, encompassing the theory of planned behavior alongside an extended rendition that integrates institutional trust. Additionally, it outlines the research hypotheses and delineates the proposed research framework’s design. The third part provides further details on sampling, research areas, measurement, as well as reliability and validity testing. In the fourth part, descriptive statistics are presented along with the results obtained from data analysis using structural equation modeling. Moving on to the fifth part, it discusses how these study findings relate to existing research. Finally, in the last section, we present our research conclusions and acknowledge any limitations encountered during this study.

2. Theoretical Background and Research Hypotheses

2.1. Theory of Planned Behavior

The theory of planned behavior is also an established socio-psychological principle applied as the main conceptual theoretical basis for investigating farmers’ inclinations in the adoption of agricultural NbS practices. TPB was originally introduced by Ajzen and Fishbein [41] and later developed into full-fledged theory through the addition of perceived behavioral control [42]. As such, the theory holds that TPB has the following five elements: attitude, subjective norms, perceived behavioral control, intention, and actual behavior. In contrast to conventional analysis approaches, the theory of planned behavior provides improved predictability of behaviors with a reduced number of components that are easier to identify. It has become one of the most extensively utilized approaches for analyzing farmers’ intentions and behaviors along with their respective theoretical determinants [43]. According to TPB’s postulation, individuals primarily base their decisions on engaging in specific behaviors on their intentions. These intentions are impacted by three pivotal elements: disposition, subjective consensus (social influence), and perceived ability to act (ease or difficulty) [42]. Moreover, TPB has garnered extensive support in recent studies examining individuals’ behavioral intentions, such as visiting green hotels, low-carbon agriculture adoption, and pro-environment behaviors [44,45,46,47,48].

Despite the acknowledged advantages it provides, there are a couple of disadvantages associated with previous applications of the TPB in relation to agricultural practices. Firstly, prior research has not extensively investigated the specific directions of influence between TPB constructs and the underlying rationales for these associations [49,50]. Secondly, when examining agriculture, past studies have frequently disregarded the potential impact that trust variables can exert on farmers’ attitudes, subjective norms, and perceived behavioral control [51]. Lastly, most research only focuses on directly examining how extended variables affect intention without fully considering their indirect effects on the original TPB variables [52]. By making reasonable extensions to the TPB framework, its inherent limitations can be effectively addressed and its theoretical explanatory power enhanced. Recent studies have widely employed models based on the extended TPB to better understand farmers’ behavioral decisions in pro-environment agricultural practices [53]. Moreover, most studies evaluating the relationship between this variable and farmers’ behaviors have found it to strongly predict NbS practices and other pro-environmental behaviors [43,53,54,55].

The initial crucial factor that influences an individual’s intention to behave in the TPB model is their attitude (or mindset). Attitude refers to the emotions, either positive or negative, that individuals experience when they engage in a specific behavior. If people develop a favorable mindset toward a specific behavior, then the likelihood of performing that behavior is high [56,57,58]. Indeed, different studies have underscored that mindsets are a main explaining variable for the possibility of adoption of environment-friendly farming practices by farmers, such as organic farming, NbS practices, and green agricultural production [46,59,60,61,62]. Similarly, in the context of agricultural NbS technologies, it can be inferred that if farmers believe that NbS technologies, such as crop rotation, conservation tillage, and other techniques, are beneficial for improving soil quality, increasing crop yields, and protecting the ecological environment, they will develop a positive attitude and may intend to adopt these practices. Thus, it is hypothesized that:

Hypothesis 1. 

The positive attitude towards agricultural NbS favors farmers’ intention to participate in NbS practice.

Subjective norms play a vital role in influencing an individual’s inclination to take action. It relates to the perceived impact exerted by significant individuals on whether one should or should not participate in a specific behavior. This implies that an individual’s conduct is primarily influenced by the favorable or unfavorable evaluations of certain influential individuals who hold importance for that person [44]. Subjective norms (SN) are a social factor encompassing the perceived societal pressure to embrace particular technologies or behaviors [63]. The greater the perception of subjective norms, the higher the likelihood of an individual engaging in a specific behavior [42,53,64,65]. This principle also applies to adopting NbS technology in agricultural production [55]. If farmers perceive that most people around them believe they should adopt NbS technology, they will experience pressure and have intentions to implement it. Therefore, it is hypothesized that:

Hypothesis 2. 

The subjective norms of farmers positively and significantly influence their intention to adopt nature-based solutions practices.

Perceived behavior control is an individual’s perception relating to the degree of influence one perceives their own volition having on the dislike action in question [42]. Training, resources, market risks, and extra costs involved when presenting farmers with new technologies are likely to influence their willingness to adopt specific behaviors [59]. Therefore, the concept of PBC pertains to individuals’ perception of the ease and difficulty associated with carrying out a specific behavior [58]. In accordance with this definition, individuals who possess greater control over these factors are more inclined to engage in certain behaviors. Consequently, if the use of new technologies in agricultural production were easily performed and individuals that used them had the required knowledge and abilities, the former would be more likely to form an intention to try them out [28,66,67]. According to the above argument we can derive the following hypothesis.

Hypothesis 3. 

Perceived behavior control positively affects farmers’ intention to adopt nature-based solutions practice.

2.2. Extended of Institutional Trust in TPB

In this research, an institutional trust variable was added to the TPB model. Other variables have been included in extended forms of this model as direct determinants of either intention or behavior in analytical studies that try to enhance the predictive ability of the model. Some of these variables include moral norms, past behavior, environmental values, and risk perception, which can also be included to form an extended form of the TPB model. More importantly, the institutional trust variable itself needs to be better represented in the TPB model of farmer decision-making, considered in models like Technology Acceptance Model (TAM) or consumer behavior model [52,68]. Some studies primarily include institutional trust as the moderating variable towards the effect of attitude on intention [51,64,69,70]; few of them consider direct or indirect effects of institutional trust on individuals’ behavior intention [71]. Of interest is how institutional trust affects intentions and whether its effect on intentions can be mediated by attitude, subjective norms, and perceived behavioral control.

Indeed, institutional trust is defined as an attitude towards a particular institution (or organization, public sector, etc.) characterized by a positive expectation that the institution will perform its functions appropriately [72,73,74]. Besides the direct influence of institutional trust (IT) on willingness, IT is a moderator in the relationship between farmers, the government, and public scientists, enabling farmers to develop positive attitudes toward specific behaviors, thus enhancing behavioral willingness [71]. Furthermore, the integration of IT in agricultural practices enhances farmers’ confidence and belief in their ability to carry out specific actions. This positive influence on perceived control over behavior subsequently leads to an increase in farmers’ intention to adopt nature-based solutions in agriculture [52,75]. Consequently, the integration of IT into core TPB builds up a more enlightening consideration of the farmers’ intentions towards the adoption of NbS. Conforming to this, we thus propose the following hypothesis:

Hypothesis 4. 

Farmers’ intention to adopt nature-based solutions practices is significantly and positively influenced by the level of trust they have in institutions.

Hypothesis 5. 

The relationship between farmers’ attitudes and their intention to adopt nature-based solutions practices is positively moderated by institutional trust.

Hypothesis 6. 

Institutional trust has a positive moderating effect on the relationship between farmers’ perceived behavioral control and their intention to adopt nature-based solutions practices.

According to the previous discussion, hypothetical relationships and the theoretical research framework are presented in Figure 2.

3. Methodology

3.1. Study Area

Horqin Right-Wing Front Banner is located in Northeast Inner Mongolia, China (Figure 3). This area covers 17,000 square kilometers, encompassing 14 townships and 228 villages. It lies within the northeast black soil region, one of the world’s three most significant black soil regions. It is acknowledged as a crucial grain-producing area in China and the most extensive base for high-quality commodity grain production. In 2023, the grain sown area in this region remained stable at 390,000 hectares, with a total grain output exceeding 1.55 million tons. Conservation tillage practices were implemented on 98,000 hectares, while socialized services for substitute farming covered 63,333 hectares. The area dedicated to soybean seed production accounts for over 50% of the national total [76]. In order to improve soil quality, enhance soil fertility, protect the ecosystem, and increase maize and soya yields, the region has been promoting agricultural NbS in several townships for many years.

3.2. Sampling Method and Survey

This paper proposes a combination of typical sampling and stratified random sampling to minimize sampling bias and enhance the representativeness of the study sample. Although several counties in Northeast China are well-suited for promoting agricultural NbS technology, the lack of clear support from many local governments could introduce sample selection bias. To address this, typical sampling techniques were employed to choose the sample counties. Horqin Right-Wing Front Banner was selected because the proportion of maize and soybean cultivation (77%) closely matches that of Northeast China (78%), with comparable agricultural production conditions [77]. Despite research constraints limiting the sample area to Horqin Right-Wing Front Banner, it is believed that this choice provides a level of representativeness for the promotion of NbS technology in the region. For stratified random sampling, three agricultural towns and one agro-pastoral town in Horqin Right-Wing Front Banner were selected based on their agricultural production conditions and ecological zone diversity. The selected towns include E’ti, Barigastai, E’ergetu, and Dashizhai (Figure 3). In each town, at least four villages were chosen, and from each village, a minimum of 20 sample farmers were randomly selected, resulting in a total sample of 300 households.

The survey was conducted in November and December 2023 by students with professional training in agricultural economics. The questionnaire consisted of three sections. The first part collected demographic information and household or cooperative characteristics. The second section focused on inputs related to various aspects of maize and soybean production. The third part included questions to assess the original and extended constructs of the theory of planned behavior. A total of 300 questionnaires were completed, with 298 valid responses, yielding an effective rate of 99%.

3.3. Measurement

The survey instrument consists of 16 items categorized into five groups: (1) three items for assessing intentions related to NbS; (2) three items for measuring attitudes; (3) four items for perceived behavioral control; (4) three items for subjective norms; (5) three items for institutional trust. The scale of Intention was adapted from other studies that quantified agricultural technology adoption using TPB [46,53]. It employed a five-point Likert scale (1 = extremely weak–5 = extremely strong) to assess farmers’ intention of using several common agricultural NbS techniques in Northeast China. To measure Attitude, PBC, SN, and IT, a standard Likert scale (1 = strongly disagree–5 = strongly agree) was used in the study. Specifically, the Attitude scale assessed farmers’ mindset and emotions toward various dimensions of NbS technologies (e.g., profit, sales, and societal impact). The PBC scale evaluated farmers’ ability to control technology-related risks, master key technological aspects, and obtain external support. The SN scale included three items derived from rural interpersonal relationships in China and social network theory. The IT scale consisted of three sub-items focusing on farmers’ trust in government and public scientists. Utilizing this rating system helps address statistical concerns and facilitates efficient data analysis [78]. An important aspect worth mentioning is that established measures from previous research were employed for TPB variables and institutional trust in this study. Detailed information about the questionnaire’s contents can be found in

Table 1. Research measurement concepts and variables.

Construct	Measurement Items	Source
Intention	I am willing to adopt NbS, such as crop rotation on my land.	Ajzen (1991) [42]; Senger et al. (2017) [46]; Li et al. (2021) [53]; Varyvoda et al. (2024) [55]
	I tend to adopt agricultural NbS such as conservation agriculture when I grow maize or soybean.
	I would like to participate NbS nutrient management such as organic fertilizer substitution on my land.
Attitude	Optimal use of agricultural NbS technologies on our farm is profitable.	Ajzen (1985) [56]; Savari and Gharechaee (2013) [58]; Senger et al. (2017) [46]; Xia et al. (2023) [62]
	The adoption of agricultural NbS technology is beneficial for sales and consumers.
	The adoption of agricultural NBS technology can contribute to society and food security.
Perceived Behavior control	I can address the risks of increased costs associated with NBS technologies.	Ajzen (1991) [42]; Daxini et al. (2018) [66]; Gao et al. (2017) [59]; Savari and Gharechaee (2013) [58]
	I believe that scientists can help solve the problems of NbS technology applications.
	I think the government will likely assist with the application of NbS technology.
	I believe I can master NbS technology and contribute to national food security.
Subjective norms	When I propose adopting NbS technology, neighboring farmers will show a positive attitude.	Ajzen (1991) [42]; Fan et al. (2015) [64]; Rezaei and Seidi (2019) [65]; Varyvoda et al. (2024) [55]
	When I adopt NbS technology, valuable individuals will offer affirmative.
	The farmers whom I respect encouraged me to adopt NbS technology; I will comply.
Institutional Trust	I believe that government regulation of agricultural NbS technology is reliable.	Moon et al. (2017) [69]; Blackstock et al. (2010) [75]; Canova et al. (2020) [52]; Daxini et al. (2019) [66]
	I believe the government cares about the public interest when making decisions about NbS technologies.
	I believe agricultural scientists can make the right decisions about NbS.

.

3.4. Validity and Reliability of Instrument

To ensure the scientific integrity of the indicators, an expert panel consisting of professors from various disciplines including economics, psychology, and agricultural sciences, as well as local agricultural extension workers, carefully reviewed the draft survey and questions prior to conducting interviews with farmers. Their constructive comments were adopted for revision in order to finalize the questionnaire.

Table 2. Discriminant validity.

Construct	Attitude	IT	Intention	PBC	SN
Attitude	0.823
IT	0.241	0.778
Intention	0.428	0.484	0.802
PBC	0.326	0.502	0.400	0.661
SN	0.024	0.065	0.197	0.033	0.627

Notes: Values on the diagonal are the square root of the average variance extracted (AVE) while the off diagonals are correlations.

displays the discriminant validity results obtained using the suggested method. The findings indicate that each construct’s AVE value exceeded its corresponding cross-correlation, thus confirming satisfactory discriminant validity.

3.5. Data Analysis

Data were analyzed using SPSS 26 (IBM Corp., Armonk, NY, USA) and SmartPLS 4 software (SmartPLS GmbH, Hamburg, Germany). Structural equation modeling was conducted with SmartPLS to investigate the relationships between independent and dependent variables. This statistical technique is adept at managing multiple interrelated variable relationships within a model, thereby identifying the key factors influencing farmers’ decision-making processes [79]. The measurement model was initially examined, followed by testing the structural model [80]. Furthermore, a resampling technique employing 5000 iterations was employed to evaluate the statistical significance of path coefficients and loadings. The Partial Least Squares (PLS) approach focuses on estimating predictions in statistical models with structures that offer causal explanations [81,82]. This method addresses the common dichotomy between explanation and prediction often emphasized in academic research and is well-suited for studies with small data samples [83,84].

4. Results

4.1. Descriptive Statistics

The findings pertaining to the social-economic attributes at both individual and household levels for the households surveyed have been consolidated in

Table 3. Demographic characteristics of surveyed farmers.

Variable	Category	Frequency	Percent (%)
Farm type	Ordinary farmers	243	81.54
	Family Farm	22	7.38
	Agricultural cooperatives	33	11.07
Age	Under 35 years	14	4.70
	35–50 years	139	46.64
	Above 50 years	145	48.66
Gender	Male	294	98.66
	Female	4	1.34
Education	Under 6 years	122	40.94
	7–9 years	143	47.99
	More than 9 years	33	11.07
Cultivation size	≤3.35 Ha	72	24.16
	3.35 Ha–6.67 Ha	93	31.21
	>6.67 Ha	133	44.63
Farm decision	Yes, and full-time farming	264	88.59
	Yes, and part-time farming	25	8.39
	No farm decision	9	3.02
N	298

. The classification of these indicators is based on relevant literature [47,53]. And the classification threshold of cultivation size was obtained by using the quartile method based on the situation of this data. Regarding the farm type, the majority (81.54%) were ordinary farmers, and about one-quarter (18.45%) were family farms and cooperatives. The mean age of the household heads was 50.11 years. For the gender of household heads, 98.66% were male. The average year of education of household heads was 7.46 years. Moreover, in terms of cultivated land size, about one quarter (24.16%) were less than 3.35 ha, and less than one-third (31.21%) were between 3.35 ha and 6.67 ha while approaching half (44.63%) were more significant than 6.67 ha. Also, the mean value of land size was 30.74 ha. In addition, results showed that most household heads (88.59%) were full-time farmers and key decision-makers (Table 3). The research sample was basically consistent with the actual demographic and agricultural statistics of Northeast China in terms of age, education level, land size, and agricultural labor input, as reported in the China Rural Statistical Yearbook (2023), demonstrating the reliability of the sampling method.

According to the survey data (

Table 4. Descriptive statistics of key technologies adoption.

Technologies Adoption	Before 2022		2022		2023
	Mean	S.D.	Mean	S.D.	Mean	S.D.
Rotation	0.205	0.40	0.369	0.48	0.396	0.49
Fallow cultivation	0.013	0.11	0.017	0.13	0.023	0.15
No-tillage	0.121	0.33	0.252	0.43	0.332	0.47
Sub soiling	0.201	0.40	0.208	0.41	0.272	0.45
NPKS	0.081	0.27	0.101	0.30	0.101	0.30
Organic Fertilizer	0.111	0.31	0.232	0.42	0.252	0.43
N	298

Notes: NPKS stands for straw returning to the field; Source: self-produced.

), in the study area, conservation tillage techniques and agricultural nutrient management practices—such as crop rotation, no-tillage, deep tillage, and organic fertilizer substitution—have been the most widely adopted NbS technologies over the past two years and are showing an increasing trend. In 2023, nearly 40% of surveyed maize and soybean growers adopted crop rotation technology, around 35% implemented no-tillage practices, and close to 30% utilized deep tillage and deep loosening techniques before sowing. These agricultural NbS technologies are closely aligned with the needs of farmers in Northeast China to address soil quality issues and are the primary agricultural technologies considered in this study.

4.2. Assessing the Status of Variables in Structed Equation

The mean score and standard deviation for each variable were computed using a structured equation model. The findings are displayed in

Table 5. Descriptive statistics of variables in structed equation.

Construct	$\overline{\mathbf{X}}$	S.D.	Interpretation
Intention	3.445	1.385	Moderate
Attitude	3.480	1.129	Moderate
Perceived Behavior control	3.739	1.098	High
Subjective norms	3.231	1.058	Moderate
Institutional Trust	3.667	1.150	Moderate
N	298

Notes: 1–2.33 = low; 2.34–3.67 = moderate; 3.68–5.00 = high.

, indicating that farmers in this particular area hold a favorable inclination towards the adoption of agricultural NbS technologies ($\overline{\mathrm{X}}$ = 3.445). Farmers showed a slightly positive attitude towards NbS technologies ($\overline{\mathrm{X}}$ = 3.480). They claimed to have a high perceived behavior control when adopting NbS technologies ($\overline{\mathrm{X}}$ = 3.739). Moreover, they tended to see subjective norms about NbS technologies as moderate ($\overline{\mathrm{X}}$ = 3.231). Furthermore, the farmers admitted to possessing a reasonable degree of confidence in both the government and scientists ($\overline{\mathrm{X}}$ = 3.667) (Table 5).

4.3. Measurement Model

In this section, the adequacy of two measurement models, namely the general and extended models, was assessed using confirmatory factor analysis (CFA). The results indicated that both models adequately fit the data (

Table 6. Displays the outcomes obtained from fitting measurement models.

Constructs	Measurement Items	Original TPB			Extended TPB
		λ	t	Reliability and Validity Statistics	λ	t	Reliability and Validity Statistics
Intention	IN1	0.900	59.935	AVE: 0.802	0.890	48.644	AVE: 0.802
	IN2	0.917	67.506	CR: 0.924; α: 0.877	0.920	75.399	CR: 0.924; α: 0.877
	IN3	0.869	39.592		0.876	43.255
Attitude	AT1	0.915	60.761	AVE: 0.823	0.910	51.419	AVE: 0.823
	AT2	0.890	46.579	CR: 0.933; α: 0.893	0.895	55.327	CR: 0.933; α: 0.893
	AT3	0.916	48.496		0.916	50.962
PBC	PBC1	0.850	32.663	AVE: 0.656	0.814	25.224	AVE: 0.661
	PBC2	0.827	17.607	CR: 0.883; α: 0.826	0.858	25.019	CR: 0.886; α: 0.826
	PBC3	0.675	9.674		0.692	10.537
	PBC4	0.872	28.864		0.875	29.927
SN	SN1	0.739	7.823	AVE: 0.627	0.739	7.886	AVE: 0.627
	SN2	0.802	11.430	CR: 0.834; α: 0.706	0.802	11.969	CR: 0.834; α: 0.706
	SN3	0.833	12.977		0.833	12.632
IT	IT1	-	-	-	0.881	43.485	AVE: 0.778
	IT2	-	-	-	0.921	74.261	CR: 0.913; α: 0.857
	IT3	-	-	-	0.843	33.607

Notes: Original TPB: SRMR = 0.07; D_G1 = 0.401; D_G2 = 0.203; NFI = 0.905; RMS_Theta = 0.08. Extended TPB: SRMR = 0.06; D_G1 = 0.521; D_G2 = 0.260; NFI = 0.911; RMS_Theta = 0.09.

). Based on the CFA findings, standardized factor loadings (λ) for all selected items related to each targeted construct in both measurement models were found to be significant at a 1% significance level (p < 0.01). Moreover, Cronbach’s alpha coefficients were used to measure the reliability of our research instruments, obtaining values between 0.7 and 0.9 (Table 6). These results support the one-dimensionality of the indicators used in each measurement model and confirm their suitability for measuring the research constructs.

Recently, concerns have been raised regarding the reliability of the Fornell–Larcker criterion in ensuring discriminant validity. Consequently, a new criterion called heterotrait–monotrait ratio of correlations has emerged.

Table 7. Heterotrait–monotrait (HTMT) results.

Construct	Attitude	IT	Intention	PBC	SN
Attitude	-
IT	0.273	-
Intention	0.483	0.554	-
PBC	0.384	0.595	0.462	-
SN	0.066	0.103	0.245	0.088	-

presents values for Average Variance Extracted, Composite Reliability (CR), and Cronbach’s Alpha (α). All extracted values surpass the recommended thresholds of 0.5, 0.6, and 0.7 respectively [79,81,85], indicating sufficient convergent validity and reliability for all latent variables in our proposed model.As demonstrated in Table 7, all HTMT values were below the threshold (HTMT.85), thereby confirming no issues with discriminant validity [86,87]. Moreover, upon reviewing the model fit indices, both the original TPB and extended TPB models exhibited favorable fits. This assessment is further supported by meeting the criteria thresholds: SRMR < 0.10, D_G1 > 0.05, D_G2 > 0.05, NFI > 0.90, and RMS_Theta ≤ 0.12; indicating acceptable levels of fit were achieved. Additionally, it appears that the extended TPB model displayed a more appropriate level of fit compared to its initial counterpart in this study.

4.4. Structural Model

In this section, the measurement models were first validated, followed by the utilization of two structural models—the TPB model and its extended version—to examine the study hypotheses.

4.4.1. Original TPB Model

The initial TPB model was employed for analysis purposes. As evident from the outcomes of the preliminary model (Figure 4), all paths exhibit statistical significance. Path analysis coefficients are presented in

Table 8. Result of research structural model (bootstraps = 5000).

Hypotheses	Original TPB			Extended TPB
	$\mathsf{\gamma }$	t-Value	Result	$\mathsf{\gamma }$	t-Value	Result
H1: Attitude → Intention	0.326	5.375 ***	Confirm	0.301	5.133 ***	Confirm
H2: SN → Intention	0.176	3.736 ***	Confirm	0.163	3.764 ***	Confirm
H3: PBC → Intention	0.303	5.501 ***	Confirm	0.127	2.090 ***	Confirm
H4: IT → Intention	-	-	-	0.337	5.496 ***	Confirm
H5: IT → Attitude	-	-	-	0.241	3.743 ***	Confirm
H6: IT → PBC	-	-	-	0.502	8.948 ***	Confirm

Notes: *** indicates that the path coefficient is significant within a 99% confidence interval.

. The results indicate that farmers’ intention to adopt agricultural NbS technologies is essentially determined by Attitude, γ = 0.326, p < 0.001. The second and third most influential factors on farmers’ intention were found to be Perceived Behavioral Control and Subjective Norms in that order, with γ = 0.303, p < 0.001 and γ = 0.176, p < 0.001, respectively. Based on the foregoing analysis and results obtained from the data of the original TPB structural model, hypotheses H1, H2, and H3 are thus supported. According to Table 8, findings indicate that the key variables associated with effectiveness are: Attitude, Perceived Behavioral Control, and Subjective Norms.

4.4.2. TPB Model with Institutional Trust

The results of estimating the extended TPB structural model, which includes the institutional trust variable, are illustrated in Figure 5 and detailed in Table 8. Like the original structural model, all indicators demonstrated a good fit. The results indicated that including the institutional trust variable explained an additional 15.7% variance in behavioral intention compared to the original model, suggesting an enhanced predictive capability (Figure 5).

The newly developed model explains 51.4% of the variance in the intentions of farmers to adopt NbS technologies, which falls within an acceptable range [83,88,89]. As can be seen from the t-values, attitude, perceived behavioral control, subjective norms, and institutional trust all have positive significant effects on farmers’ adoption intentions, thus supporting hypothesis H4. Furthermore, the variable of institutional trust has a stronger influence in the extended theory of planned behavior model compared to the original theory of planned behavior structural model (Table 8).

In addition to these direct links, some indirect drivers also determined the intentions of farmers. Logically, presence of trust in institutions positively influenced attitude (γ = 0.241, p < 0.001), which means that farmers with higher levels of institutional trust were likely to perceive the NbS technologies as beneficial, thus confirming hypotheses H5. Moreover, this study established a high positive correlation between institutional trust and perceived behavioral control with γ = 0.502, p < 0.001. This may be an indication that farmers who had heightened trust in institutions were more likely to develop the relevant competencies and reduce costs associated with the adoption of NbS technologies, hence supporting Hypothesis H6 (Figure 5; Table 8).

5. Discussion

A psychological-social model was developed and applied to examine farmers’ intentions using data collected from 298 respondents. The objectives of this paper are threefold: (1) to assess the effectiveness and applicability of the TPB in understanding farmers’ intentions towards adopting NbS technologies; (2) to enhance the explanatory capacity of the original TPB model by incorporating institutional trust as a variable; (3) to explore potential indirect effects of institutional trust on intentions through its influence on perceived behavioral control and farmers’ attitudes. The findings indicate that including institutional trust as a variable improves the model’s explanatory power, revealing two indirect pathways through which it influences adoption intentions towards NbS.

Results show there is a significant relationship of farmers’ attitudes with their behavior and intention regarding the adoption of agricultural NbS. This agrees with other studies conducted by Han (2015) [45], Senger et al. (2017) [46], Ledesma et al. (2018) [90], Hou and Hou (2019) [47], Cakirli et al. (2020) [61], and Xia et al. (2023) [62]. This is a factor considered very important in attitude according to the theory of planned behavior, very close to the intentions of farmers [91]. Therefore, when it comes to NbS technologies or other environmentally friendly practices, attitude plays a pivotal role in influencing behavior through its direct association with intentionality [62]. The underlying mechanism behind this phenomenon may be that farmers’ belief in green and ecological practices fosters a positive perception of green agricultural technologies [46,59], thereby increasing their inclination towards adopting sustainable technologies like NbS [55]. Furthermore, while attitude was found to be the most influential factor in the original model, its impact on intentionality decreased slightly in the extended model, consistent with other studies based on an extended TPB framework [58,60]. One possible explanation for this could be that compared to the newly introduced variables, attitude has a relatively weaker effect on farmers’ intentions.

Subjective norms, which represent external societal pressures, have emerged as a significant independent predictor of farmers’ inclination to adopt NbS practices. This suggests that farmers who perceive greater social support or evaluation regarding NbS are more inclined to embrace these practices. Furthermore, this observation has been acknowledged in the existing literature on NbS behavior and other environmentally conscious technologies by Fan et al. (2015) [64], Rezaei and Seidi (2019) [65], Li et al. (2021) [53], Yin et al. (2022) [63], and Xia et al. (2023) [62]. The development of subjective norms is closely intertwined with the social network formed through interactions between farmers and their environment, as well as between the general public and governmental entities [63]. There are three existing divisions of social network types, and interpersonal social network (ISN) is an important research object in the context of agricultural technology adoption [92]. Consistent with previous studies [63,93], this study reveals that subjective norms based on the ISN concept positively impact farmers’ willingness to adopt NbS technologies. Rural China exhibits characteristics of an acquaintance society, characterized by a dense and small social network structure dominated by neighbors and relatives. This structure amplifies the influence of others’ evaluations on technology adoption [94,95]. Therefore, under the subjective norms of social pressure and accessible information sources, farmers judge whether adopting NbS technologies is appropriate.

The findings further indicate that farmers’ intention towards implementing agricultural NbS practice is significantly influenced in a positive manner by their perceived behavioral control. It indicates that control beliefs, especially those related to technical operations, technical challenges, and risk prevention, have significant value in reinforcing intentions [96]. NbS eco-friendly technologies, such as conservation tillage, require professional knowledge, skills, and attention to detail. Consequently, they are among the more complex green agricultural production practices [28]. Consequently, in the event that farmers are unable to surmount the challenges posed by technical risks and insufficient technical training, their motivation may dwindle, leading to potential failure in making favorable choices [66,67,97]. In addition, the PBC component of this research fully considers whether farmers can receive help from the government and agricultural scientists when encountering difficulties in adopting NbS technologies. The analysis confirms the validity of this consideration. A possible explanation for this phenomenon is that the government and agricultural scientists currently play multiple roles in rural China’s top-down technology diffusion system, such as providing technical training, technical guidance, and government credit endorsement [98,99,100]. This support enhances farmers’ confidence in adopting NbS technologies.

The results from the extended TPB model reveal that institutional trust emerged as the most critical direct predictor affecting intentions toward NbS practices. Institutional trust refers to the confidence placed in public institutions, including local government, agricultural extension services, and agricultural scientists [74]. This finding suggests that individuals with higher institutional trust are more likely to behave more responsibly with environmentally friendly behavior such as NbSs [72,101,102] and believe that its decisions and policies benefit them [103]. A possible reason for this phenomenon is that farmers believe that the agricultural technologies promoted by governments and scientists can be trusted to help them deal with technical problems reliably [72,104,105].

Unlike other research that regards institutional trust as a moderating variable affecting factors such as attitude, perceived behavior control, and subjective norms [51,72], this study examines the direct impact of institutional trust on farmers’ willingness to adopt NbSs and uncovers new pathway where institutional trust indirectly influences willingness by affecting attitudes and PBC. Although the technical content of this research varies, the results are similar to the pathways through which institutional trust influences individual willingness in studies on pro-environmental behaviors, such as consumers’ willingness to buy organic food and farmers’ willingness to adopt nutritional management practices [52,74]. From the perspective of attitude as a mediating variable, farmers’ institutional trust plays a crucial role in their overall positive evaluation of NbS technology. The impact of institutional trust on attitudes shows that the more farmers trust the government and scientists, the more positive their attitudes toward NbSs are, which in turn positively affects their willingness to adopt NbSs. Relevant studies suggest that advice sources from trusted agricultural experts and scholars can improve farmers’ attitudes towards specific behaviors [75]. Furthermore, in the study of perceived behavioral control as a mediating variable, higher levels of institutional trust can increase farmers’ confidence in successfully implementing NbS technology. As mentioned before, unlike European countries, the Chinese government and scientists are vital in promoting and extending agricultural technology [99,100]. This is especially pertinent in developing countries, where many farmers may lack advanced education in new agricultural technologies and have limited capital [99,100].

6. Conclusions

This study makes a substantial contribution to the current understanding of the predictors for the adoption of agricultural NbS technologies among maize and soybean farmers in northeastern China. The current research was an exploratory step toward predicting farmers’ willingness in Northeast China to adopt NbS practices, represented by crop rotation and conservation cultivation, in staple crop production. This study employs the theory of planned behavior model to identify key factors influencing farmers’ intentions to adopt agricultural nature-based solutions in Northeast China, while also investigating the direct and indirect roles of institutional trust in this process. The results of this study suggest that the expanded theory of planned behavior demonstrates greater accuracy in forecasting the behavioral intentions of Chinese farmers, with potential enhancement through fostering trust in institutions. Farmers’ intentions were positively influenced by their attitude, subjective norms, perceived behavioral control, and institutional trust. Compared to previous studies, this research has found two indirect pathways through which institutional trust can affect farmers’ intention to adopt an agricultural NbS technology: ‘Institutional trust → attitude → intention’ and ‘Institutional trust → perceived behavioral control → intention’. Second, it is observed in this study that the appropriateness of using TPB for studying NbS adoption would be fitting, while at the same time extendable to other environmentally friendly agricultural technology practices.

The conclusion of this research offers valuable insights for promoting NbSs in agriculture and shaping relevant policy support in Northeast China. Firstly, it is essential for the Chinese government and agricultural scientists to eliminate any barriers to trust, foster strong relationships with farmers, and create a supportive environment for the adoption of new technologies. The government should work to enhance farmers’ trust in agricultural institutions by implementing transparent and effective policies, sharing information openly, and encouraging farmer participation in the policy-making process. Secondly, given the importance of Attitude and Perceived Behavioral Control in influencing technology adoption, China’s agricultural extension system should design training and outreach programs tailored to the specific characteristics of the technologies and farmers’ needs. This is particularly important for promoting NbS practices such as conservation tillage and crop rotation, presenting them in ways that are both familiar and practical for farmers, and emphasizing the economic and ecological benefits of these practices. Finally, recognizing the role of SN, it is crucial to establish social guidance and demonstration mechanisms that can amplify the influence of early adopters. Local governments in China should consider offering incentives such as technology subsidies or easier market access to further encourage the widespread adoption of NbS technologies among farmers.

However, this study has several limitations, which should be noted. Firstly, this study was conducted in a single location, Horqin Right-Wing Front Banner, in Northeast China, using cross-sectional data from maize-soybean farmers. The findings may not apply to other regions in China or other developing countries. Future research should expand the study area and crop range. Secondly, the analysis needed to address the gap between NbS willingness and actual behavior. Future studies should focus on the consistency between willingness and behavior by incorporating variables representing farmers’ actual adoption of NbSs. Lastly, although the extended TPB model has moderate predictive power for adopting NbS behaviors, other influencing factors, such as socio-demographic characteristics and risk preferences, may also play a significant role and should be considered in future research.