Adoption of eco-friendly soil-management practices by smallholder farmers in Shandong Province of China

Logit analysis of factors influencing farmers adopt decision making

1. Study area

Shandong Province is located on the eastern coast of China,with a total area of 137,800 km2 and 75,107.61 km2 cultivated area. It is located between 34°22rsquo; and 38°23rsquo;N and 114°19rsquo; and 122°43rsquo;E, in a temperate-zone monsoon climate with four distinctive seasons. Topographically the area is a plain with an average altitude of 57.6 m above mean sea level, and its annual temperature is 13.8°C with August as the hottest month (28.2°C) and January as the coldest month (minus;1.6°C). The annual precipitation is approximately 727.9 mm, and there is high rainfall from May to September. As a major agricultural province, this region has two main crops grown annually. Winter wheat (Triticum aestivum L.) and summer maize (Zea mays) are the two main crops, which are grown on 3,525,210 ha and 2,874,213 ha, respectively, followed by rice (Oryza sativa L., 130,694 ha), millet (Setaria italica, 15,168 ha) and Chinese sorghum (Sorghum bicolor (L.) Moench, 6558 ha).

The total grain output in Shandong in 2013 was 45.28 Mt, which was the highest of all provinces in China. However, the increase in grain yield was highly dependent on extensive agricultural inputs. The amount of fertilizer applied reached 4.72 Mt, which was surpassed by only Henan Province (CSSB 2013). Farmland in Shandong is the worst in the entire countrybecause its soil mainly lacks nitrogen (N) and phosphorus (P), and land desertification is also a serious problem (Guo et al. 2007). As estimated, the amount and density of agricultural pollution in this region are very extensive, and soil loss due to chemical fertilizer is the major reason. Thus, it is urgent to encourage farmers in Shandong to adopt eco-friendly soilmanagement practices to ease these environmental issues.

2. Conceptual framework and empirical model

The conceptual framework in this paper is based on the conceptual models developed by Negatu and Parikh (1999)as a user-technique characteristic model. According to this model,the characteristics of a technology or practice and farmer perception of the characteristics play the central role in the adoption decision and diffusion process under agroecological, socio-economic and institutional contexts. The model also emphasizes the importance of the involvement of farmers in the technology development process and the roles of relevant policies and strategies in facilitating that involvement. Thus, factors affecting technology adoption in this study are170 grouped into farmer and farm household characteristics, farm biophysical characteristics, farm financial/management characteristics and social/institutional characteristics (Knowler and Bradshaw 2007).

The theory of the maximization of utility is applied to the adoption by a farmer of an advanced technology or practice(Sidibeacute; 2005). When the utility derived from a new technology exceeds that of the older one, a farmer will adopt the technology. According to Greiner et al.(2009) and Greiner and Gregg(2011), the motivations of farmers to adopt conservation practices include three types: conservation and lifestyle motivation, economic/financial motivation and social motivation. The goals of a farmer in China in adopting a new technology are mainly to earn high income or yield; thus, the utility in thisstudy is gaining more profit. The relationship between utility185 (Uij) and characteristics (Xi) affecting adoption for the ith farmer is as follows (Eq. 1):

Uij=Xialpha;j εij j ｛0，1｝ and i =｛1， 2，3hellip;，n ｝ (1)

where j indicates the adoption or non-adoption of the technology, and i indexes the factors affecting the adoption; (Uij)represents the utility the farmer obtains from the technology, and (Xi) represents a vector of the factors affecting the adoption. It is assumed that a farmer selected the agricultural technology that could bring him the largest utility, and the ith farmer could adopt the jth technology if Ui1 f Ui0; Bi indexes the adoption decision (Eq. 2):

（2）

The logical function is applied to model the dependent variable (Eq. 3), the parameters are estimated by the likelihood ratio test and the statistical analysis used is a binary logistic regression of SPSS 21.0 (SPSS Inc., Chicago, IL) as follows:

（3）

where mu;i=εi1- εi0, and F（Xibeta;） is the cumulative density function for mu;i, P denotes the probability of the ith farmer to adopt the new technology, and beta; predicts the log odds of the dependent variable.

The model (Eq. 4) used to analyze the factors affecting the adoption of eco-friendly soil-management practices can be expressed as follows:

（4）

The independent variables representing the characteristics affecting the adoption of the practices can be found in Table 1.

Table 1.Definition of variables used in the logit model of adoption of eco-friendly soil-management technologies or practices in Shandong, China.

Variable Description of variable

Adoption of eco-friendly soil-management Dummy, 1 if adoption, 0 otherwise practices (ADOPTION)

Age (AGE) Continuous variable, year

Education (EDUCATION) Unordered categorical variable, less than elementary school = 1, elementary school = 2, junior middle school = 3, high school = 4, further education = 5

Awareness of farmland degradation (DEGRADATION)

Dummy, 1 if severe, 0 otherwise

Knowledge about environmental regulations (REGULATIONS)

Dummy, 1 if know, 0 otherwise

Risk type (RISK)

剩余内容已隐藏，支付完成后下载完整资料

中国山东省个体农户采用生态友好型土地管理实践

农户采纳决策影响因素的Logit分析

1 研究区域与农户调研

作为农业大省，山东是我国粮食主产区之一，全省总耕地面积为75107.6平方公里。该省种植的主要农作物为小麦和玉米，种植面积分别为3525210公顷和2874213公顷，另外水稻，谷子和大豆种面积分别为130694公顷，15168公顷和6558公顷。2013年山东总的粮食产量为4528.2万t，居于全国各省前列，仅次于黑龙江和河南。该省粮食产量的增加主要依靠农业生产资料的过度投入，如2013年山东化肥施用量达472.7万t，居于全国前列。然而山东耕地缺N/P元素，水土流失和土地沙漠化程度较为严重，土地质量全国最差。由于山东农业集约化程度较高，农业面源污染排放总量和排放强度均较大，而且农业生产中化肥流失是造成农业污染的主要原因之一，因此鼓励农户采用生态友好型农业生产技术迫在眉睫。

2.模型理论与假设

本研究理论框架主要以Negatu和Parikh（1999）提出的农户-技术特征模型为理论基础，具体为在农业生态、社会经济和相关制度的背景下，农业生产技术特征及农户对此特征的认知水平对技术采纳决策和传播过程起了重要的作用，该模型强调了农业生产技术发展的过程中农户参与的重要性，并且政府相应的支持政策能够鼓励农户的参与行为。因此本研究将影响农户技术采纳决策的因素分为农户特征、农田特征、农业生产财政和管理特征与其他外源性因素四类。

本研究利用最大效用理论来解释农民对农业新技术的采用决策，当采用新农业生产技术产生效用大于采用旧的农业生产技术时，农户会选择采纳新的农业生产技术。据Greiner等（2009, 2011）, 农户采用保护性农业生产技术的动机包括保护生态环境、经济收益动机和社会动机。由于我国农户进行农业生产的主要动机是获取高的产量和收益，因此本研究最大效用即为收益最大。影响第i个农户采纳决策的效用（）和影响因素特征（）之间的关系如下：

（1）

其中，表示不采用或采用农业生产技术，表示采纳决策影响因素种类；表示采纳农业生产技术的效用，表示采纳决策影响因素向量。本研究假设选择能给他们带来最大效用的生产方式，如果，第个农户将选择采纳生产技术，表示采纳决策，如下（2）：

（2）

Logical回归函数用来进行变量模型回归（3），模型系数有极大似然比检验来估计，二元Logit模型检验由SPSS21.0 (SPSS Inc., Chicago, IL) 完成。

（3）

其中，是积累密度函数，表示第i个农民的采取新的农业生产技术概率，表示因变量比数比。模型（4）用来分析分析生态友好型农业生产技术影响因素影响程度，如下：

（4）

模型中解释变量代表影响农户采纳生态友好型土壤管理技术的因素如表6-3所示。

表6-3 模型中解释变量及其定义

Table 6-3 Definition of variable used in the Logit model of adoption of the eco-friendly soil management technologies or practices in Shandong, China

土地养分管理技术是我国目前发展可持续农业，保障国家粮食安全主要推广的农业生产技术，主要包括施用有机肥、保护性耕作、秸秆还田、作物轮作、种植绿肥和测土配方施肥。Drsquo;Souza等（1993）表明当农民采用七项农业生产技术中采用四项即可认为其采用生态友好型农业生产技术。因变量用来表明农业生产者是否是生态友好型农业生产技术的采用者，各影响因素对生态友好型农业生产技术采用情况假设如表6-4。

表6-4 模型变量统计性描述和假设

Table 6-4 Descriptive statistics for the Logit model of adoption of the eco-friendly soil management technologies or practices in Shandong, China

农户特征影响因素主要包括年龄（AGE）、学历（EDUCATION）、土地退化认知（DEGRATION）、环保法规认知（REGULATIONS）、风险类型（RISK）、农业生产技术认知（PERCEPTION）和采纳态度（ATTITUDE）。由于年轻的、学历较高的农户年龄更容易接触了解新的农业生产技术，并把它应用于生产实践 ^[201]，因此本研究假设年龄对采纳决策有负向影响，学历对采纳决策有正影响。由于农民如果意识到农田土壤退化会引起农产品产量下降和食品安全问题，土地退化认知对采纳决策有正向影响。Shiferaw和Hol

剩余内容已隐藏，支付完成后下载完整资料

资料编号：[26206]，资料为PDF文档或Word文档，PDF文档可免费转换为Word