The Impact of Knowledge Management Capabilities on Innovation Performance from Dynamic Capabilities Perspective: Moderating the Role of Environmental Dynamism

As an important path to enhance the innovation performance of enterprises, knowledge management has received much attention in recent years. However, most of the existing literature on knowledge management and innovation performance is based on a static perspective, and ignores the influence of dynamic changes in the environment. This study intends to explore the relationship between dynamic knowledge management capability and innovation performance as well as examine the moderating effect of environmental dynamism. The questionnaire survey approach is used in this study and the data is collected from 253 sample enterprises in China. To estimate the proposed relationships in the theoretical model, this study adopts hierarchical Multiple Regression (MR) and Moderated Multiple Regression (MMR) methods. The results show that all dimensions of dynamic knowledge management capability have different degrees of positive influence on innovation performance. Moreover, it was also confirmed that there are different moderating effects of environmental dynamism on the relationship between the dimensions of knowledge management capability and innovation performance. This study can contribute to enriching the theoretical research of dynamic knowledge management capability and innovation performance, and offer scientific guidance for decision making to efficiently enhance the enterprise’s knowledge management level and innovation performance. Moreover, the findings can also provide valuable insights for enterprises to make use of KM capabilities to enhance innovation performance


Introduction
Faced with the diversity of customer needs and the rapidity of technological changes, many enterprises have considered innovation as one of the most important factors for coping with fierce market competition. Innovation performance depends on the degree of synergy between market and technology, which relies on knowledge management (KM) to achieve. Therefore, KM has been of great interest to academics as a key breakthrough to effectively enhance innovation performance. A scientific study to reveal the influence mechanism of KM on innovation performance is much needed.
According to the knowledge innovation theory, innovation is the integration of old and new knowledge to meet market demands, which is essentially a series of processes of KM. This reflects the strong link between KM capability and innovation performance [1]. A large body of literature has explained the mechanism between them, and the results showed that effective KM is the driving force for enterprises to carry out innovative activities and improve innovation performance [2,3]. The relationship between KM and innovation performance is becoming a hot issue, especially distinguishing the impact of different stages of KM on innovation performance [4], or exploring the impact of different KM approaches on the innovation [5]. However, there are some limitations in the existing literature: firstly, the research subject is limited. Some research selected certain classical activities of KM, such as knowledge sharing [6] and knowledge transfer [7,8], to explore the relationship between them and innovation performance. This cannot reveal the inner link between them in a systematic and comprehensive way. Secondly, the research perspective is limited. Existing literature mainly analyzes KM capabilities based on a static perspective. In fact, KM is a complex and continuous process and accompanied by some changes. It is obviously difficult to examine the impact of dynamic evolution of KM capabilities on innovation performance.
Meanwhile, as an open system, companies need to constantly interact with their external environment. When they conduct routine business activities, they are inevitably affected by the environment. Researchers have found that environmental dynamism can affect the relationship between KM and innovation performance [9,10]. Moreover, recent empirical studies suggested that dynamic environment can positively influence innovation performance in different ways through KM strategies [11]. The relationship of them has attracted much attention of scholars. However, there is relatively little literature that can provide support for deeper insights into the moderating effect on different processes of KM and innovation performance.
Based on the above-discussed literature, this study intends to fill in the research gap of traditional static perspective research on KM, and construct a new theoretical model by integrating KM and dynamic capability to examine their impact on innovation performance. Knowledge absorption, knowledge transfer, and knowledge application are the three key components of dynamic KM capabilities. In addition to exploring the impact of the external environment in depth, this study will explore how environmental dynamism moderates the relationship between KM capabilities and innovation performance. Thus, the following research questions were addressed in this study. RQ1: How do the different dimensions of KM (knowledge absorption, knowledge transfer, and knowledge application) affect innovation performance? RQ2: How does environmental dynamism (market dynamism and technological dynamism) moderate between KM and innovation performance?
The findings can enrich the theoretical research of KM capabilities and innovation performance, and provide valuable insights for enterprises to make use of KM capabilities to enhance innovation performance in practice.
The structure of the study is as follows. The theoretical foundations and literature review is presented in Section 2. The hypotheses are developed in Section 3. The methods used in this study are explain in Section 4, followed by providing the results of data analysis in Section 5. Further discussions about results are in Section 6. The final section presents the conclusions, which includes theoretical contributions, practical implications, the limitations of this research, and future research directions.

Knowledge Management from a Dynamic Capability Perspective
The knowledge-based view holds that KM can facilitate innovative practices by transforming knowledge assets into new products and services based on a series of management activities. It can foster the exchange and sharing of knowledge required for innovative activities, thereby stimulating the generation of new ideas and ultimately improving innovation performance. Existing research also provides theoretical support for this, for example, Singh et al. proposed that knowledge sharing can help companies achieve innovative goals and thus improve organizational performance [12]; Attia and Eldin found the positive impact of KM capabilities on organizational performance [13]; Mohamad et al. also believes that KM contributes to firm innovativeness [14]. In general, most literature on KM is based on a static perspective, which considers it as a continuous process, creating value and enhancing effectiveness by proactively absorbing and applying knowledge. Nevertheless, as an open system, the management activities are subject to changes in the external environment, which requires that KM cannot remain unchanged.
Furthermore, dynamic capability is considered as a higher-order capability that helps enterprises adapt to environmental changes as soon as possible and gain sustainable competitive advantage [15]. More specifically, it is a capability that integrates and resets internal and external resources to respond quickly to environmental changes by continuously updating resources and optimizing allocation methods. Capability can be regarded as the collection of knowledge in resource-based research perspectives. Similarly, dynamic capability is the integration of knowledge as a key resource to attain competitive advantage. In essence, it is the process of transforming and evolving knowledge to achieve a leap in value. This has attracted the attention of many scholars. For example, Verona et al. proposed that dynamic capability is knowledge-based and that knowledge absorption, integration, and reconfiguration are the basis of it [16]. Nielsen considered that dynamic capability is a specific and typical set of KM activities, including knowledge development, knowledge reintegration, and knowledge application [17]. Zollo et al. argued that dynamic capability is essentially a capability to maintain organizations' operating efficiently through knowledge creation activities [18]. Hilliard et al. also revalidated the interaction relationship between KM and dynamic capability [19]. Thus, it is obvious that dynamic evolution of knowledge provides an opportunity to combine KM with dynamic capability.
Based on knowledge-based view and dynamic capability view [20,21], this study defines dynamic KM capabilities as the ability of enterprises to absorb, transform, and apply internal and external knowledge continuously, according to the perceived trend of environmental changes. This capability focuses on the enterprises which dynamically allocate knowledge based on external changes in order to provide support for innovation activities. In this way, it can create value from existing resources and quickly respond to external technological and market changes. The process is essentially a dynamic cycle of knowledge absorption, knowledge transfer, and knowledge application, which runs through the whole process of innovation activities [22,23].
With reference to the research of scholars such as Gold [24], knowledge absorption mainly includes three elements: namely technology, structure, and culture, which is the premise for enterprises to carry out dynamic KM activities. The knowledge base can be enriched and have resource advantage gradually through absorbing relevant knowledge. Regarding knowledge transfer, drawing on the SECI knowledge spiral structure proposed by Nonaka [25], it is considered that it mainly involves the transfer of explicit and tacit knowledge. In other words, it means integrating internal and external resources and transforming it into new knowledge, which can provide support for innovation activities. During the process of knowledge application, enterprises eventually transform knowledge into real productivity and create economic benefits by reconfiguring knowledge resources according to their strategic direction [26]. Additionally, this can also increase opportunities for technological innovation.

Innovation Performance
Innovation has been considered as the process by which companies successfully apply innovative ideas resulting from knowledge resources to develop new products, technologies, and services. Based on this, innovation performance can be defined as the outcome of innovation, and it reflects the degree of success achieved by firms in meeting its innovation objectives [27,28]. In general, it may appear in the form of the development of new technologies, the improvement of products, and gaining desired market share. In accordance with the existing literature, this study regards innovation performance as examining the enhancement of the capabilities and the realization of value in the process of innovative practices.
Scholars have recognized that innovation is essentially an interactive process of reallocation of knowledge resources towards market demand [29]. Furthermore, it is an effective means of generating innovative ideas with the integration of new externally acquired knowledge and existing internal knowledge [30]. This is beneficial to facilitating the successful implementation of innovative practices. As discussed above, innovative practice relies on the ability of firms to manage knowledge. In other words, dynamic knowledge capabilities have a positive influence on innovation performance.

Environmental Dynamism
Based on the 'environment-behavior' research paradigm, the external environment is considered as an important context for an organization's survival and development. Additionally, its activities are influenced and constrained by the dynamic changes in the environment. Moreover, previous studies have confirmed that environmental dynamism is a basic determinant of innovative behaviors [31].
Environmental dynamism refers to the rate and unpredictability of changes in the environment, including technologies, product demand, and customer preferences [32]. Rapid changes and uncertainty make it more difficult for firms to predict future market and technology trends. Additionally, they also create discrepancies between existing knowledge and the knowledge needed for innovation [33]. As a consequence, in order to obtain a sustainable competitive advantage, companies have to absorb the knowledge resources they need from the external environment. Consistent with this view, research has implied that the benefits which derive from strong knowledge management capabilities are expected to be even greater in dynamic environment [34,35]. This suggests that environmental dynamism has an impact on knowledge management capabilities and innovation performance.

Knowledge Absorption Capability and Innovation Performance
The knowledge base is often the starting point for corporate innovation, which is a process of continuous accumulation of knowledge and needs to be supported by rich knowledge resources. As a result, the richness of the knowledge base is related to the level of innovation performance [36]. Knowledge absorption capability has an impact on innovation performance mainly through information technology, climate atmosphere, and organizational structure. In detail, information technology can provide a transmission medium for knowledge exchange by supporting formal or informal communication. From this, it facilitates the sharing of knowledge across different boundaries and increases the frequency of innovation activities [37,38]. Corporate culture can stimulate employees' enthusiasm and motivation for innovation by creating an innovative atmosphere, which in turn contributes to an increase in the overall level of innovation [39]. Organizational structure can help enterprises optimize or reconfigure processes according to the development trend of the environment [40]. This makes it possible that internal structure matches the external environment, slowing down the impact brought by environmental changes. In that case, companies will be able to shorten the response time and thus accelerate the speed of innovation. Thus, knowledge absorption capability can effectively promote organizational learning and product development activities, and contribute to enhancing performance in terms of frequency, level, and speed of innovation. From the above discussion, this study proposes the following hypothesis:

Hypothesis 1 (H1).
Knowledge absorption capability has a positive impact on innovation performance.

Knowledge Transfer Capability and Innovation Performance
Knowledge transfer is a dynamic process of internalizing the knowledge into strategic resources through organizational learning. It is a key process of enriching new innovative resources, which has a significant impact on innovation performance [41]. In fact, innovation is a dynamic adaptive behavior based on the perception-response model. During this process, companies transform new knowledge with old knowledge, increase the depth of knowledge and then flow purposefully to support innovative activities. Obviously, the wider the scope of knowledge transfer, the better innovation performance. In order to carry out product or technology innovation successfully, it is crucial for companies to use knowledge transfer capability to sublimate and convert knowledge into heterogeneous resources. That is to say, the stronger the knowledge transfer capability, the bigger the possibility of possessing heterogeneous knowledge, the easier it is to generate a wealth of innovative ideas, and the more likely that enterprises enhance performance levels [42]. From the above discussion, this study proposes the following hypothesis: Hypothesis 2 (H2). Knowledge transfer capability has a positive impact on innovation performance.

Knowledge Application Capability and Innovation Performance
Knowledge application is the ability to put accumulated and transformed knowledge into innovative practices [43], and enhance innovation performance by deploying the knowledge required for multiple forms of innovation to fuel the output of new products and technologies. Given that the essence of knowledge application is the continuous transformation of intellectual capital into innovative outcomes, this requires companies to pay attention to the following points when drawing on knowledge application capability. On the one hand, companies need to master the capability to assess the opportunities and threats brought about by environmental changes, and make the most of knowledge resources for innovative activities. On the other hand, it is extremely important to integrate and transform knowledge from different organizations and departments. With the help of this process, firms can achieve the goal of stimulating a constant flow of innovative ideas and laying the foundation to win more opportunities [44]. In short, the ability to apply knowledge is not only a prerequisite for enterprises to achieve their goals, but also a guarantee to enhance innovation performance. Knowledge that has been absorbed and transformed can only generate economic benefits if it is applied to business activities. This can foster new innovative opportunities and increase the possibility of commercialization of knowledge. From the above analysis, knowledge application capability can increase the success rate of innovation solutions and improve innovation performance. Therefore, this study proposes the following hypothesis:

Hypothesis 3 (H3).
Knowledge application capability has a positive impact on innovation performance.

The Moderating Effects of Environmental Dynamism
Environmental dynamism can affect the content of KM and the direction of innovation, which in turn generates different performance. Additionally, it is bound to interfere with the process of making the leap from knowledge to innovation. In view of this, environmental dynamism has been considered as an important contextual variable affecting corporate innovation. It is worth noting that environmental dynamism is a multidimensional concept and each dimension may have a distinguished influence. Hence, this study plans to incorporate environmental dynamism as a moderating variable, which refers to the researches of other scholars [45,46], and explore its impact in terms of both market dynamism and technological dynamism.
Market dynamism represents the degree of change in the behavior or needs of companies' stakeholders, such as customers, competitors, and partners. When market environment is more stable, customers' consumption preferences are less likely to change. Moreover, competitors' behavior is relatively predictable, and the competitive advantages possessed by enterprises can mostly be maintained for a long period of time. At this time, enterprises can respond to environmental changes with their existing knowledge without updating the knowledge base. They have little incentive to innovate with knowledge, and the levels of KM capabilities are relatively low. Therefore, the positive influence of the dimensions of dynamic knowledge management capability on innovation performance will be weakened. On the contrary, when the level of environmental dynamics is high, customers' consumption preferences change rapidly and a competitor's behavior is less predictable, making it difficult for firms to provide new products continuously. For this reason, firms need to innovate to meet customers' changing needs, and are generally more motivated to utilize KM capabilities [47].
Based on the above analytical logic, the changing market environment may result in the loss of the established knowledge strengths. This can stimulate firms to expand and update their knowledge base and integrate various knowledge resources through KM activities to support innovative activities. The impact of the dimensions of dynamic knowledge management capabilities on innovation performance will be enhanced when the market environment is dynamic. The hypotheses are as follows.
Hypothesis 4a (H4a). Market dynamism positively moderates the relationship between knowledge absorptive capability and innovation performance.
Hypothesis 4b (H4b). Market dynamism positively moderates the relationship between knowledge transfer capability and innovation performance.
Hypothesis 4c (H4c). Market dynamism positively moderates the relationship between knowledge application capability and innovation performance.
Technological dynamism reveals the speed of change in the development of new products and technologies in industry to which the company belongs. When the pace of technological change is slow, companies will be able to anticipate future trends based on stable information and rely on previous technological architectures to cope with changes in the environment. There is little willingness to use KM for innovative outcomes at this moment. Additionally, the impact of the dimensions of dynamic knowledge management capabilities on innovation performance will be weak. When the pace of technological change is fast, companies have to constantly update the knowledge base in order to avoid a rapid depreciation of existing technological knowledge [48]. Especially companies which were faced with unpredictable evolutionary paths and shortened product life cycles, there is a strong willingness to make use of KM and respond to technological innovation in time [49].
Based on the above analytical logic, the positive impact of the dimensions of dynamic KM on innovation performance will increase with technological changes constantly emerging. The hypotheses are as follows.
Hypothesis 5a (H5a). Technological dynamism positively moderates the relationship between knowledge absorptive capability and innovation performance.
Hypothesis 5b (H5b). Technological dynamism positively moderates the relationship between knowledge transfer capability and innovation performance.
Hypothesis 5c (H5c). Technological dynamism positively moderates the relationship between knowledge application capability and innovation performance.
In summary, this study constructed a theoretical research model about the relationship between dynamic knowledge management capabilities, innovation performance, and the influence of environmental dynamism, as shown in Figure 1.

Research Approach
This study focuses on the relationship between dynamic knowledge management capabilities, environmental dynamism, and innovation performance. Based on an empirical questionnaire, the survey analysis approach, which is commonly used by scholars, was used for data collection. The reasons for adopting this method are as follows: the questionnaire survey approach can record the respondents' responses truthfully and quickly collect data from the targeted population. Moreover, it can also obtain a large amount of objective information at a relatively low cost.

Questionnaire Development
The aim of this study is to investigate the relationship between the different dimensional components of dynamic knowledge management capabilities on innovation performance and the moderating role of environmental dynamism. The questionnaire used was mainly based on the classical 5-point Likert scale, which is extensively used in literature [50,51]. The initial questionnaire was formed by modifying the items as appropriate to the purpose of the study. Additionally, the scales were firstly translated and back-translated to increase the accuracy of items. To further guarantee the reliability and validity, a pilot study was conducted. In the pilot test, we selected some experts and entrepreneurs in the relevant fields, and their opinions were used to improve the initial questionnaire. Afterwards, the questionnaire was distributed on a small scale for pre-testing, and the individual questions were modified again according to the analysis results and feedback from the pilot study's respondents. Finally, the questionnaire was developed.
The questionnaire contains a total of 33 measurement items, including basic information about the respondents and measurements of variables. In addition, except for the control variables, the questionnaire was measured through a 5-point Likert scale, which ranges from 1, "strongly disagree" to 5, "strongly agree". All items that we used in the questionnaire are given in Appendix A.

Sample and Procedures
The survey sample is mainly high-tech enterprises in the central region of China, covering industries such as advanced manufacturing, electronics, IT, communications, biomedicine, and new materials. The reason is they are knowledge-intensive industries with frequent technological and product innovations that fit the research theme. The data was collected using a questionnaire in two ways.
Firstly, the data were obtained through interviewing sample companies and distributing paper questionnaires to some managers. Secondly, electronic questionnaires were assigned to sample enterprises outside of Henan Province and other locations using questionnaire platforms and emails to survey respondents. In order to ensure the validity of the data, the questionnaires were completed by middle and senior managers who are familiar with the overall situation of the companies.
A total of 420 questionnaires were distributed and 326 were collected. After eliminating invalid questionnaires, 253 valid questionnaires were finally obtained, for an effective response rate of 60.24%. In addition, considering that the questionnaire collection method may affect the independence of samples, this study performed a test on the valid questionnaires recovered. The results show that there is no significant difference among the samples.

Variables and Measures
The research is composed of three independent variables (knowledge absorption, knowledge transfer, and knowledge application), two moderating variables (market dynamism and technological dynamism), and dependent variable (innovation performance).

Dynamic Knowledge Management Capabilities
Adopted from Lane et al. [52,53], this study designed six items to measure knowledge absorption capability, five items were designed to measure knowledge transfer capability, and six items were designed to measure knowledge application capability in terms of knowledge integration and configuration.

Innovative Performance
Management innovation and technological innovation as the key types of innovation can affect innovation outputs. Accordingly, innovation performance is supposed to measure in a combination of both management and technology. Similarly to Gemünden et al. [54], this study designed six items to evaluate the degree of innovation in processes and products and services in order to measure the innovation performance of the firm with the help of KM capabilities.

Environmental Dynamism
Market dynamism and technological dynamism are two dimensions of environmental dynamics. The former mainly indicates the changes in clients' demand and competition intensity, and the latter refers to the possibility that significant changes will occur about product performance improvements and technology innovations. The measurement was based on the scale used by Jansen et al. [55] and Martínez-Pérez et al. [56]; six items were designed to measure market dynamism and technological dynamism.

Control Variables
Referring to previous literature [57,58], firm age, firm ownership, firm size, and industry were used as control variables in this study to distinguish their effects from those of the independent variables on the dependent variable. In particular, the age was calculated from the time the firm was established and ends at the time of receiving the questionnaire. Firm size was measured by the number of employees. The dummy variables were set for the firm ownership, where 1 represents state-owned enterprises and 0 represents non-state-owned enterprises. Similarly, the industry was set as a dummy variable, where 1 represents high-tech enterprises and 0 represents traditional enterprises.

Demographic Analysis
In order to understand the basic distribution of the respondents, this study analyzed their demographic characteristics (please see Table 1). The result shows that 28.25% of the enterprises are in the IT industry and 42.29% are private enterprises. Additionally, the majority of the enterprises have existed from 3 to 5 years (41.90%), and 41.51% have 100 to 300 employees.

Reliability and Validity
Reliability reveals the trustworthiness of the survey respondents in completing the questionnaire. This study used SPSS 22.0 to analyze the data and assess the reliability of the multi-item constructs according to Cronbach's alpha. The results show that the Cronbach's alpha of each individual construct, which is all above the threshold of 0.7, ensured that the reliability of the adopted scale met the criteria. This indicates that the scale for the variables is within an acceptable level.
Validity measures the accuracy of the measurement items, including content validity and construct validity. About content validity, the scale used in this study can be considered to have good content validity because it is based on several previous established scales, and it has been modified as appropriate according to specific research contexts. In addition, construct validity can be divided into convergent validity and discriminant validity. On the one hand, this study used AMOS 25.0 to conduct confirmatory factor analysis (CFA). The convergent validity was tested by calculating the average variance extracted (AVE) and construct reliability (CR) value of each variable (please see Table 2). Table 1 shows satisfactory results for the convergent validity by indicating that the AVE of each variable is above the threshold of 0.5 and CR of each variable is greater than 0.7. The loadings of the questions corresponding to the variables are shown in Table 2 which shows that all items had high factor loading (>0.5). On the one hand, the square root of the AVE values of each variable was calculated to test the discriminant validity (please see Table 3). They are greater than the correlation coefficient between the variables, which indicates the questionnaire has good discriminant validity.

Common Method Bias
Considering that the questionnaire data originates from the subjective judgment of participants, common methodological bias may exist [59]. In order to control them within a reasonable range, this study adopted both ex ante procedural control and ex post statistical test to weaken its possible influence. In terms of ex ante procedural control, we mainly took the following approaches: The question items referred to mature scales of previous research as far as possible to ensure the clarity of the question formulation. Additionally, the variables were measured using several items and the key variables were staggered. Meanwhile, multiple methods were used to collect data and respondents can complete the questionnaire anonymously. In terms of ex post statistical test, the study analyzed the question items using "Harman's one-factor test". The study accounted for the first factor for 27.05%, and this shows no single factor was found to dominate. In summary, there is no serious common method bias.

Descriptive Statistics and Correlation Analysis
The means, standard deviations, and correlation coefficients of the variables are shown in Table 3. As can be seen from the results, the relationships between knowledge absorption ability (r = 0.633, p < 0.01), knowledge transfer ability (r = 0.502, p < 0.01), and knowledge application ability (r = 0.686, p < 0.01), as well as innovation performance are all significantly and positively correlated. Market dynamism is significantly correlated with innovation performance (r = 0.478, p < 0.01) and the relationship between technological dynamism and innovation performance is positively correlated (r = 0.570, p < 0.01). This indicates that there is a correlation between the variables studied, which initially verifies the previous hypothesis and provides support for further studies.

Testing of Hypotheses
To avoid the problem of multicollinearity, the key variables were standardized before regression analysis and tested again using the variance inflation factor (VIF) [60], and the results show that the VIF values of each variable is distributed between [1,6], indicating that the multicollinearity is within a reasonable control range.

Main Effect Analysis
This study used hierarchical regression analysis to test the above theoretical hypotheses. In detail, the control variables and independent variables were gradually put into the model respectively. According to the results of the analysis, it can be judged whether the independent variables have an influence on the dependent variable, the direction of influence and the degree of influence, and thus, the main effect can be verified.
Model 1 represents the regression analysis of industry, firm type, firm age, and firm size on innovation performance. It is the basis model that includes only control variables and dependent variable. From model 2 to model 4, they are the core regression steps to test the influence of dynamic knowledge management capabilities on innovation performance. Model 2 is based on Model 1, in which the "knowledge absorptive capability" was added to the model to explore the relationship between "knowledge absorptive capability" and innovation performance under the influence of control variables. Model 3 introduced "knowledge transfer capability" into Model 2 to examine the effect of knowledge transfer capability on innovation performance. Model 4 further added "knowledge application capability" to test whether there is a significant effect between knowledge application capability and innovation performance. In this way, the analysis and validation of the main effects were completed step by step, and the regression results are shown in Table 4. The results demonstrate that the variance explained by knowledge absorption ability, knowledge transfer ability, and knowledge application ability on innovation performance is 13.30%, 15.80%, and 27.20% respectively. Additionally, knowledge absorption capability positively affects innovation performance (β = 0.106, p < 0.05), therefore, hypothesis H1 is supported. Knowledge transfer capability has a positive effect on innovation performance (β = 0.183, p < 0.05), and hypothesis H2 is supported. The positive relationship between knowledge application capability and innovation performance is significant (β = 0.546, p < 0.001), and hypothesis H3 is supported. In conclusion, dynamic knowledge management capabilities have a significant positive effect on innovation performance.

Moderating Effect Analysis
Based on the standardization of key variables, this study constructed interaction terms for each dimension of environmental dynamism and dynamic knowledge management capabilities to test the moderating effect. The results are shown in Table 5. The above models test the moderating effects of market dynamism and technological dynamism on the relationship between each dimension of dynamic knowledge management capability and innovation performance respectively.
Firstly, the moderating effect of market dynamics was tested by adding the interaction terms of it and each dimension of dynamic knowledge management capability to Model 5. Model 6 suggests that the interaction term can explain 3.50% of the variance in innovation performance. The moderating effect of market dynamism on the relationship between knowledge absorption capability and innovation performance is not significant (β = 0.087, p > 0.05). However, market dynamism positively moderates the relationship between knowledge transfer capability and innovation performance (β = 0.168, p < 0.05). In contrast, it negatively moderates the relationship between knowledge application capability and innovation performance (β = −0.186, p < 0.05), which is opposite to the direction of Hypothesis. This means hypothesis H4b is supported that market dynamism positively moderates the relationship between knowledge transfer capability and innovation performance. The hypothesis H4a that market dynamism positively moderates the relationship between knowledge absorptive capability and innovation performance do not pass the test. And the hypothesis H4c that market dynamism positively moderates the relationship between knowledge application capability and innovation performance is consistent with this.
Secondly, the interaction term between technological dynamism and three dimensions of dynamic knowledge management capability was introduced into the model to test the moderating effect. Model 8 shows that the interaction term can explain 6.40% of the variance in innovation performance. Technological dynamism positively moderates the relationship between knowledge absorption capability (β = 0.109, p < 0.05), knowledge transfer capability (β = 0.122, p < 0.05), and innovation performance. However, technological dynamism negatively moderates the relationship between knowledge application capability and innovation performance (β = −0.145, p < 0.01). This means H5a and H5b are supported and H5c did not pass the test. Model 9 is the total model, which again validates the conclusions drawn above.

Discussions
This study established a theoretical model with dynamic knowledge management capabilities, innovation performance, and environmental dynamism from the perspective of dynamic capability. Relying on the above research framework, we explored the mechanism of dynamic knowledge management capabilities which influences innovation performance and the moderating effect of environmental dynamism. The hypotheses were verified by using the data obtained from the questionnaire, and the following are the details of the above hypotheses and discussion of results.
Dynamic knowledge management abilities have a significant contribution to innovation performance. This suggests that knowledge absorption, knowledge transfer, and knowledge application all have positive effects on innovation performance. Additionally, knowledge application capabilities have the strongest impact on innovation performance, followed by knowledge transfer and knowledge absorption. The conclusion is in line with the findings of Shin et al. [61] and Thomas et al. [62]. This study again clarifies the mechanism of KM on innovation performance and deepens the understanding of dynamic capability and knowledge-based view. At the same time, it also provided an explanation for the long-standing debate in the field of strategic management on whether dynamic capability can positively influence firm performance in the Chinese context. Furthermore, there is a differentiated interactive moderating effect on the relationship between dynamic knowledge management capabilities and innovation performance. The findings are consistent with the point proposed by Marsh et al. and Rodrigo et al. [63,64], which reveals the "complexity" of environmental dynamism. In particular, the moderating effect of market dynamism on the relationship between knowledge absorption and innovation performance is not significant, while technological dynamism positively moderates knowledge absorption and innovation performance. Both market dynamism and technological dynamism positively moderate knowledge transfer and innovation performance, but negatively moderate knowledge application and innovation performance.
Additionally, it should be noted that the moderation effects of two of the pathways proposed are slightly inconsistent with the hypothesis. More specifically, the first one is that this study hypothesizes that market dynamism positively moderates the relationship between knowledge absorptive and innovation performance, but the empirical results reveal that the moderating effect is not significant. Considering customers' product preferences are complex and changeable in the extremely dynamic market, the reason for this phenomenon may be that it is difficult for enterprises to collect data and to pinpoint the direction of demand in a short period of time. Besides, knowledge absorption involves various aspects such as information technology, climate atmosphere, and organizational structure. These factors often have a "hysteresis effect" when responding to external environmental changes, which makes it difficult for the influence of market dynamism on knowledge absorption capability and innovation performance to appear in a timely manner. It leads to the moderating effect of market dynamism as insignificant [64]. These factors often have a "lag effect" when responding to external changes, which prevents the impact of market dynamism on knowledge absorptive capability and innovation performance from being demonstrated in a timely manner. These circumstances described above result in an insignificant moderating effect.
The other path is that the study hypothesizes that technological dynamism positively moderates the relationship between knowledge application and innovation performance, but the empirical results demonstrate that the direction of regulation is negative. The possible reason is that this study takes high-tech enterprises as the sample, and such firms are in this environment: diverse customer needs, rapid technological changes, and short product cycles [65]. However, the long duration of the cycle of applying knowledge to innovative activities makes it difficult for innovation performance to compensate for the resulting costs in the short term. Moreover, coupled with the limited knowledge application capability of some companies, it leads to the negative impact of the dynamism of the environment on knowledge application capability and innovation performance.

Conclusions
This study integrated the knowledge-based view and the dynamic capability view. A theoretical model of dynamic knowledge management capabilities was constructed with knowledge absorption, knowledge transfer, and knowledge application as the core dimensions. Based on this, we examined the influence of dynamic knowledge management capabilities on innovation performance respectively. Moreover, the environmental dynamism was introduced as a moderating variable into the model to deeply investigate the interactions between the variables. The main conclusions are as follows.
Dynamic knowledge management capabilities have a positive impact on innovation performance. In detail, knowledge absorption, knowledge transfer, and knowledge application all contribute to innovation performance, among which knowledge application has the strongest positive effect on innovation performance, followed by knowledge transfer and knowledge absorption.
There are partial moderating effects of environmental dynamism in the relationship between dynamic knowledge management capabilities and innovation performance. More specifically, the moderating effect of market dynamism in the relationship between knowledge absorption and innovation performance is insignificant, while technological dynamism positively moderates the relationship. Market dynamism and technological dynamism positively moderate the relationship between knowledge transfer and innovation performance, while negatively moderating the relationship between knowledge application and innovation performance.

Theoretical Contributions
The theoretical contributions of this study are the following: Firstly, we integrated dynamic capability into the theoretical framework of knowledge management and proposed a dynamic knowledge management capability model. In this context, the interaction between knowledge management capabilities and innovation performance was deeply explored from the perspective of dynamic capability. This study can extend static knowledge management, which mainly emphasizes the internal "knowledge base", to the dynamic knowledge management, which is "process-oriented". It can also make up for the shortcomings of the static research perspective and facilitate the intersection and penetration of dynamic capability theory with the fields of KM and innovation performance.
Secondly, the existing literature has generally attached importance to the impact of a particular process of KM on innovation performance [66], with less comprehensive consideration of the impact of different stages of KM. However, as an open system, it is obvious that there are limitations to explain the impact of a particular process of KM in isolation. To this end, this study integrated knowledge absorption, knowledge transfer, and knowledge application, and explored the impact of each stage on innovation performance based on dynamic interactive feedback process. Moreover, the results can provide a more systematic and comprehensive theoretical explanation of the influence of KM capabilities on innovation performance from the dynamic capability perspective.
Finally, in conjunction with the power-change theory, the moderating roles of market dynamism and technological dynamism were examined separately. Although many scholars have studied the relationship between environmental dynamism and innovation performance in the past [67,68], they are less likely to explore the moderating effect on the impact between KM capabilities and innovation performance. Therefore, this study took environmental dynamism into account and further clarified its moderating effect on innovation performance during the dynamic development of KM. There is no doubt that this study can make contributions to exploring the role boundaries of dynamic KM capabilities.

Practical Implications
The findings of this study have the following implications for how enterprises can systematically utilize dynamic knowledge management capabilities to enhance innovation performance.
On the one hand, enterprises should attach more importance to the construction and enhancement of dynamic knowledge management capabilities. The reason is that dynamic knowledge management capabilities can help enterprises absorb and expand potentially valuable knowledge into the knowledge base from the external environment. It makes it easier for enterprises to optimize the allocation of knowledge resources, promote crossborder integration, and thus strengthen the ability to adapt to the environment. Meanwhile, dynamic knowledge management capabilities enable enterprises to respond to market new demands quickly, and carry out innovative behaviors such as product development based on these changes. This allows enterprises to establish a competitive advantage with uniqueness and initiative, which can create a barrier against competitors, and thus achieve the goal of long-term growth.
On the other hand, enterprises should pay close attention to the impact of environmental changes and develop innovative solutions that are in line with them. Environmental dynamism may dilute a firm's innovative resources and cause it to lose its competitive advantage, which in turn interferes with innovation performance. In order to reduce the negative impact of environment on innovative activities, enterprises are supposed to overcome the traditional path dependence and be sensitive to the trends. Consequently, enterprises can ensure that their KM strategies match environmental changes, so that they can develop scientific decision-making solutions and achieve a continuous improvement of innovation performance.

Limitations and Future Research Directions
Although the findings of this study can provide some help to business managers, there are still some limitations due to the complexity of the issue.
To begin with, the research object focused on a certain province, and the regional coverage of the sample was relatively limited. Therefore, future research can try to expand the scope of the sample data collection and conduct a comparative analysis between regions.
Secondly, this study examined only the moderating effect of environmental dynamism on the relationship between KM capabilities and innovation performance. Additionally, other variables can be selected as moderating variables in future studies to enrich the mechanism between dynamic knowledge management capabilities and innovation performance, such as individual characteristics, organizational culture, and organizational structure. Moreover, the subject of knowledge management is people, therefore, future research could also consider factors such as human resource management as moderating variables.
Finally, this study used cross-sectional data to explore the relationship between variables, which fails to provide a comprehensive picture of the influence process between them. Future research can consider selecting panel data covering different periods to reveal the evolution pattern between them more deeply and precisely.
Author Contributions: Writing-review, funding acquisition, project administration, supervision, L.F.; conceptualization, formal analysis, investigation, methodology, and writing-original draft, Z.Z.; writing-review, funding acquisition, project administration, supervision, J.W.; writing-review and editing, K.Z. All authors have read and agreed to the published version of the manuscript.

Conflicts of Interest:
The authors declare no conflict of interest.