From Fragmentation to Integration: An Empirical Study on Enhancing Design–Construction Interface Management in EPC Landscape Projects

Abstract

The EPC model is currently the mainstream implementation approach for landscape projects, but fragmented management of the design–construction interface constrains project performance. Addressing issues such as cost overruns and schedule delays caused by ambiguous responsibility allocation, inefficient information transfer, and frequent design changes in EPC landscape projects, this paper focuses on the Xiaoyalong Wetland Park project in Kashi, Xinjiang, as a core case study. Combined with research on 12 representative projects, it identifies 16 interface management factors across four dimensions: contract management, organizational coordination, technical support, and ecological–artistic integration. Employing a mixed-methods approach combining questionnaire surveys (186 valid samples) and semi-structured interviews, validated through SPSS and structural equation modeling, this study confirms that early collaborative design serves as a core driver. Based on empirical findings, it derives and proposes a three-tiered optimization strategy: “foundation at the root layer, coordination at the transition layer, and assurance at the direct layer”. Pilot application of this strategy demonstrated significant effectiveness, reducing design change rates by 32%, shortening coordination time by 28%, and lowering cost overrun rates by 15%. This study enriches the theoretical framework of interface management in landscape engineering EPC projects and provides practical guidance for similar projects in arid regions.

1. Introduction

In the process of high-quality development of China’s urbanization, landscape engineering serves as a core vehicle for maintaining urban ecological balance and enhancing the quality of the living environment, with construction scale and quality requirements continuously rising [1]. Against this backdrop, the engineering, procurement, and construction (EPC) model has emerged as the mainstream delivery method for landscape projects [2]. Leveraging its integrated advantages across the entire “design–procurement–construction” chain, EPC resolves the pain points of fragmented resources and dispersed responsibilities inherent in traditional fragmented contracting. Industry statistics indicate that EPC projects now account for over 70% of domestic landscape engineering in 2024, reflecting a pronounced trend toward integration. However, landscape projects’ inherent ecological sensitivity, artistic expression, and on-site uncertainties make interface management between design and construction significantly more challenging under EPC than in traditional construction projects [3]. This interface serves as a critical junction throughout the project lifecycle, where management efficiency directly impacts costs, schedules, and quality. Currently, fragmented interface management is prevalent in EPC landscape projects, hindering the realization of EPC integration advantages and triggering a series of engineering issues. Therefore, systematically researching optimization pathways for interface management holds significant theoretical and practical value for enhancing the quality and efficiency of the landscape engineering industry and advancing urban ecological civilization construction [4].

Extensive empirical research indicates that, despite the EPC model offering integrated solutions, fragmented management of the design–construction interface remains a core bottleneck constraining project performance. This manifests in three primary issues: First, ambiguous responsibility allocation arises because existing EPC contract terms are based on generic construction projects and fail to accommodate the cross-functional coordination demands between landscape design and construction, resulting in unclear accountability and frequent buck-passing [5]. Second, inefficient information transfer creates a cognitive gap between the artistic expression of landscape design and the technical implementation requirements of construction. The absence of standardized communication mechanisms triggers design misinterpretations, delayed feedback, and rework. Third, frequent design changes occur due to landscape projects being highly susceptible to uncertainties like geology, climate, and ecological conservation. The lack of early-stage coordination and on-site assessment during the design phase results in persistently high change rates during construction [6]. Research indicates that these issues collectively cause domestic EPC landscape projects to exceed average cost by 19.2% and experience schedule delays exceeding 23%, representing persistent industry-wide challenges.

The core theoretical framework of the EPC project delivery model originates from transaction cost economics and integrated management theory, with related research achieving certain progress. Zhang, Q. et al. (2019) proposed that integrated contracting reduces the number of transaction parties, clarifies responsibility boundaries, and lowers negotiation and supervision costs [7]. Wang, L. et al. (2021) empirically validated that the core value of the EPC model lies in optimizing resources and achieving synergistic efficiency gains throughout the entire lifecycle [8]. In practical research, scholars worldwide have developed a series of solutions centered on contract governance and organizational optimization. However, these studies predominantly focus on industrial and civil construction, neglecting the ecological and artistic characteristics of landscape engineering, thereby limiting the applicability of their findings [9]. The core theories for design–construction interface management are interface management and collaborative management theories, both emphasizing clear responsibilities and establishing efficient communication mechanisms to achieve seamless integration [10]. Regarding influencing factors, Liu, G.Y. et al. (2022) identified three dimensions: contract completeness, organizational rationality, and information efficiency [11]. This guidance document pointed out that responsibility clarity, technical consistency, and collaborative participation are key factors [12]. However, existing research fails to account for the unique characteristics of landscape engineering, making it difficult to explain the specific causes of interface conflicts in this field. Landscape engineering projects differ fundamentally from traditional construction in three key aspects: First, ecological sensitivity requires adherence to the principle of “ecological priority” and adaptation to site-specific natural conditions [13]. Second, on-site uncertainty makes projects susceptible to natural factors like climate and soil, necessitating random construction adjustments [14]. Third, artistic expression, where design outcomes are subjectively personalized, demanding precise construction to faithfully reproduce artistic intent [15,16]. These characteristics necessitate interface management that emphasizes early-stage coordination, dynamic adjustments, and information visualization. While existing studies acknowledge the distinctiveness of landscape projects, most focus on either design or construction phases alone, failing to deeply explore the systemic impact on EPC interface management. Consequently, targeted strategies remain elusive [17,18].

The EPC model has become one of the mainstream delivery methods for large-scale complex projects worldwide due to its advantages in integrating responsibilities, controlling costs, and shortening project schedules [19]. However, its core challenge—the inherent interface between the design (D) and construction (C) phases—remains the decisive factor in project success or failure. International research indicates that interface management failure is the primary cause of project changes, cost overruns, and schedule delays. To address this issue, global research and practice have developed two mainstream approaches: relationship-based governance, exemplified by North America’s “integrated project delivery”, emphasizes early involvement, shared risk, and a culture of collaboration [20], while technology-based governance, represented by the UK’s “Common Data Environment” and the BIM Mandate, aims to achieve seamless information integration through digitalization.

However, knowledge originating from traditional architecture and civil engineering reveals significant limitations when applied to landscape engineering, particularly ecological restoration projects. The core subjects of landscape engineering are living systems (plants, soil, and hydrology), characterized by high regional specificity, seasonality, and uncertainty. This necessitates interface management that addresses not only the “physical–management” interface but also the “ecological–technical” interface. Despite growing global attention to “nature-based solutions” and “climate-resilient landscapes” [21], the interface management challenges and mechanisms for effectively implementing these ecological concepts through EPC models remain an academic “blind spot”. Through an in-depth empirical study of EPC ecological restoration projects in China’s arid regions, this study seeks to address this gap by developing a tailored interface management framework for the specific context of landscape engineering. This framework provides theoretical insights and practical references for similar global projects.

In summary, existing research exhibits three core gaps: First, EPC-related studies focus on generic buildings without considering landscape engineering characteristics, resulting in a lack of tailored interface management theories. Second, research on the design–construction interface fails to expand influencing factors based on landscape characteristics, resulting in outcomes that struggle to meet practical demands. Third, landscape engineering studies have not leveraged the integrated advantages of EPC, nor established suitable collaborative mechanisms. Therefore, this study is anchored in the EPC model, focuses on the characteristics of landscape projects, and prioritizes interface management optimization. It aims to bridge the gap between generic theory and practical needs by constructing a management system that combines theoretical depth with practical feasibility.

2. Research Framework and Hypotheses

2.1. Theoretical Foundations

2.1.1. Interface Management Theory

Interface management theory originated in the field of engineering management. Its core lies in identifying interface nodes between various project phases and stakeholders, clarifying interface responsibilities, standardizing communication processes, resolving interface conflicts, and achieving coordinated interaction among all system elements Shen, L. et al. (2022) categorizes interface management into three core phases: interface identification, interface coordination, and interface control [22].This framework emphasizes that the essence of interface management lies in eliminating “information silos” and “responsibility vacuums” through information exchange and resource integration [10]. In EPC projects, the design–construction interface represents one of the most critical horizontal interfaces, with its management effectiveness directly determining the efficiency of integrated project advancement [23]. This theory provides a foundational analytical framework for identifying core conflict points within the design–construction interface of EPC landscape projects and establishing targeted management mechanisms [24]. Building upon this theory, this study identifies unique nodes within the landscape project design–construction interface and clarifies the management requirements for each node.

2.1.2. Integrated Management Theory

Integrated management theory is centered on systems theory, advocating the integration of dispersed project elements and processes into an organic whole to enhance overall project performance through synergistic effects. The core advantage of the EPC model stems from this integrated management philosophy, achieving holistic optimization of project objectives by consolidating resources across the entire design, procurement, and construction chain. Zhang, X. et al. (2020) noted that applying integrated management in EPC projects requires focusing on integration across three dimensions: organization, information, and processes [25]. Among these, information integration forms the foundation, organizational integration provides the safeguard, and process integration constitutes the core. The ecological and artistic nature of landscape engineering demands greater emphasis on full-process integration in design–construction interface management. Drawing upon integrated management theory, this study constructs an integrated organizational, informational, and procedural management framework for the design–construction interface, providing theoretical support for optimizing strategy formulation.

2.1.3. Collaborative Management Theory

Collaborative management theory emphasizes establishing collaborative mechanisms to promote interaction and cooperation among different entities and stages, achieving synergistic effects where “1 + 1 > 2”. In the engineering field, this theory has been widely applied to the collaborative optimization of design and construction. Research by Müller, R. et al. (2020) indicates that early collaborative participation between design and construction entities can effectively reduce design defects and mitigate change risks [26]. The artistic nature of landscape engineering design and the technical demands of construction make coordination between design and construction parties particularly crucial [27,28]. Based on collaborative management theory, this study focuses on analyzing the collaborative mechanisms between design and construction parties during early design, on-site construction, and change management phases, deriving the influence pathways of collaborative participation on interface management efficiency.

2.1.4. Transaction Cost Economics Theory

Transaction cost economics posits that market transactions involve various costs such as negotiation, supervision, and default, and that sound governance structures can effectively reduce these transaction costs [29]. Wang, L. et al. (2021) proposed that integrated contracting models can significantly reduce contract negotiation and monitoring costs by clarifying responsible parties and reducing the number of transaction participants [8]. In EPC landscape projects, issues such as ambiguous responsibilities at the design–construction interface and inefficient communication fundamentally reflect excessively high transaction costs [30]. This study employs this theory to analyze the impact mechanisms of different interface management factors on transaction costs, providing a theoretical basis for reducing interface transaction costs through contract refinement and organizational optimization.

2.2. Hypothesis Framework

2.2.1. Defining Core Influencing Factor Dimensions

Based on a literature review and theoretical analysis, and considering the unique characteristics of EPC landscape projects, this study categorizes the influencing factors of design–construction interface management into four core dimensions: contract management, organizational coordination, technical support, and ecological–artistic integration. The specific connotations of each dimension are as follows: First, the contract management dimension refers to establishing contractual terms that clearly define the interface responsibilities, communication obligations, and dispute resolution mechanisms between design and construction parties, thereby providing institutional safeguards for interface management [31,32]. Second, the organizational coordination dimension involves establishing collaborative organizational structures and standardized communication processes to facilitate efficient interaction between design and construction entities [33,34,35]. Third, the technical support dimension utilizes digital technologies and standardized technical specifications to achieve precise transmission and sharing of design–construction information. Fourth, the environmental adaptation dimension addresses external factors such as ecological constraints and site-specific natural conditions in landscape projects, enabling dynamic alignment between design and construction [36]. This study constructs an “influencing factors–interface management efficiency” research framework, with design–construction interface management efficiency as the dependent variable and contract management, organizational coordination, technical support, and ecological–artistic integration as core independent variables. It further derives relational hypotheses between each dimension’s influencing factors and interface management efficiency while identifying key observational variables within each dimension [37,38].

2.2.2. Derivation of Research Hypotheses

Contract Management Dimension and Interface Management Efficiency: Contracts serve as the core basis for interface management in EPC projects, with clear contractual terms effectively reducing interface responsibility disputes [38,39,40,41]. Wang, J. et al. (2021) demonstrated a significant positive correlation between the clarity of interface responsibility allocation in contracts and interface management efficiency, indicating that well-defined interface clauses reduce the likelihood of mutual buck-passing [42]. In landscape projects, where design–construction cross-collaboration demands are more pronounced, the specificity of contractual terms is particularly critical [43]. Based on this, the following hypothesis is proposed:

H1. 

Contract management dimensions exert a significant positive influence on the efficiency of design–construction interface management in EPC landscape projects.

Further disaggregated, the contract management dimension comprises four observational variables: clarity of responsibility allocation, completeness of clauses, reasonableness of dispute resolution mechanisms, and binding force of coordination obligations [5]. Clear responsibility allocation delineates the boundaries of responsibility between design and construction parties in specialized phases such as landscape art restoration and ecological conservation construction [42,43]. Comprehensive clauses cover the unique interface coordination requirements of landscape projects [44], reasonable dispute resolution mechanisms can swiftly resolve interface conflicts, and strong binding coordination obligations ensure proactive participation in interface collaboration [45,46]. Based on this, we propose the following:

H1a. 

Clarity of contractual interface responsibility allocation has a significant positive impact on organizational coordination and interface management efficiency.

H1b. 

Contractual interface clause completeness significantly and positively influences interface management efficiency.

H1c. 

The reasonableness of contractual interface dispute resolution mechanisms significantly and positively influences both the technical support dimension and interface management efficiency.

H1d. 

The binding force of contractual interface coordination obligations exerts a significant positive influence on interface management efficiency.

Organizational Coordination Dimension and Interface Management Efficiency: Organizational coordination serves as a critical safeguard for resolving design–construction interface conflicts. By establishing collaborative organizational structures and standardizing communication processes, it facilitates information sharing and efficient interaction between parties [47]. Xiao, J. et al. (2021) indicate that early collaborative participation of contractors in design and establishing standardized communication mechanisms significantly enhances interface management efficiency [48]. Landscape projects, characterized by artistic design demands and technical construction requirements, necessitate stronger organizational coordination [49,50]. Therefore, we propose the following:

H2. 

The organizational coordination dimension exerts a significant positive influence on the efficiency of design–construction interface management in EPC landscape projects.

The organizational coordination dimension serves as the core supporting pillar for interface management in EPC landscape project design–construction. It can be broken down into four key observation variables: early collaborative involvement of contractors, standardized interface communication mechanisms, effectiveness of cross-departmental joint task forces, and clarity of organizational structure and responsibilities [51]. Early collaborative involvement of contractors in design can preemptively address construction feasibility issues. By engaging during the design phase, practical experience regarding construction techniques and site conditions can be integrated into the design, reducing interface conflicts at the source [52]. Taking arid zone landscape projects as an example, contractors can provide professional advice on critical processes such as plant planting and irrigation system installation, ensuring design solutions are both ecologically compatible and constructively feasible. Standardizing interface communication mechanisms effectively reduces cognitive discrepancies between design intent and construction execution [53]. Through standardized measures—such as fixed communication frequencies, unified information carriers, and regulated communication processes—precise and efficient information exchange can be achieved among multiple stakeholders (design, construction, supervision, and owners). This ensures construction aligns with design requirements while enabling timely feedback on site-related issues, fostering bidirectional, high-efficiency coordination. The effective operation of cross-departmental joint task forces allows for the rapid resolution of interface collaboration issues. Design–construction interface issues often involve multidisciplinary intersections, where single-department handling risks partiality and delayed responses [54,55]. Joint task forces integrate core resources from design, construction, and technical teams, breaking departmental silos to coordinate stakeholder demands, efficiently resolve interface conflicts, and maintain project momentum. Clarifying organizational structures and responsibilities eliminates buck-passing [56]. Ambiguous responsibility allocation is the core driver of fragmented interface management, often leading to disjointed workflows, cost overruns, and schedule delays. Clear delineation of authority and responsibility defines the boundaries of duties for all participants and departments, avoiding “vacuum zones” and redundant work to ensure orderly implementation of interface management. Based on this, we propose the following:

H2a. 

Early collaborative participation of contractors in design significantly and positively impacts interface management efficiency.

H2b. 

Standardizing interface communication mechanisms has a significant positive impact on the integration of ecological and artistic dimensions and on interface management efficiency.

H2c. 

The effectiveness of cross-departmental joint task forces significantly and positively impacts interface management efficiency.

H2d. 

Clarity of organizational structure and responsibilities significantly and positively impacts interface management efficiency.

Technical Support Dimension and Interface Management Efficiency: Technical support forms the foundation for achieving efficient design–construction interface integration, particularly as the application of digital technologies can overcome information transmission barriers [57]. Zhang, Q. et al. (2019) empirically demonstrated that BIM technology’s information integration capabilities can effectively enhance information transmission efficiency at the design–construction interface in EPC projects while reducing design changes [7]. Landscape projects, characterized by complex three-dimensional spatial forms and stringent plant configuration requirements, exhibit an even more urgent need for technical support [58]. Therefore, we propose the following:

H3. 

The technical support dimension exerts a significant positive influence on the management efficiency of the design–construction interface in EPC landscape projects.

The technical support dimension comprises four observed variables: BIM information integration level, digital management platform maturity, design document completeness and accuracy, and technical training and guidance [59]. BIM technology enables three-dimensional visualization of landscape designs, aiding contractors in accurately interpreting design intent. Digital management platforms facilitate real-time sharing and dynamic updates of design–construction information [60], complete and accurate design documents form the foundation for interface coordination, and targeted technical training enhances both parties’ interface management capabilities. Therefore, we propose the following:

H3a. 

The BIM information integration level has a significant positive impact on interface management efficiency.

H3b. 

Digital management platform maturity has a significant positive impact on interface management efficiency.

H3c. 

The integrity and accuracy of design documentation exert a significant positive influence on the integration of ecological and artistic dimensions and the efficiency of interface management.

H3d. 

Technical training and guidance have a significant positive impact on interface management efficiency.

Dimension of Ecological and Artistic Integration and Interface Management Efficiency: Integration of ecology and art is a core characteristic of landscape projects, where changes in external environmental factors directly impact the efficiency of design–construction interface coordination [61]. Li, J. et al. noted that environmental factors such as ecological constraints and uncertainties in site natural conditions are key triggers for design–construction interface conflicts. Effectively responding to external environmental changes and achieving dynamic adaptation between design and construction are crucial for enhancing interface management efficiency [62]. Based on this, we propose the following:

H4. 

The ecological and artistic integration dimension exerts a significant positive influence on the efficiency of design–construction interface management in EPC landscape projects.

This dimension encompasses four observed variables: adequacy of ecological baseline surveys and data application, technical clarification and communication of artistic intent, construction feasibility of ecological design elements, and adaptive ecological/artistic adjustments during design–construction iterations [6]. The adequacy of ecological baseline surveys and data application directly corresponds to ecological sensitivity, measuring the integration of ecological scientific knowledge across the design–construction interface [63]. The technical clarification and communication of artistic intent directly corresponds to artistic expression uncertainty, measuring the effectiveness of translating abstract artistic language into concrete technical instructions [61]; the alignment of ecological requirements bridges the ecological and construction interfaces, measuring the feasibility of implementing ecological design requirements during construction [64]; and ecological/artistic adaptive adjustments during design–construction iterations measure the dynamic integration capacity of interface management, reflecting collaborative response effectiveness in addressing ecological and artistic uncertainties [65]. Based on this, we propose the following:

H4a. 

The adequacy of ecological baseline surveys and data application significantly and positively influences interface management efficiency.

H4b. 

Technical clarification and communication of artistic intent significantly and positively influence interface management efficiency.

H4c. 

Ecological requirement alignment significantly and positively influences interface management efficiency.

H4d. 

Ecological/artistic adaptive adjustments during design–construction iteration significantly and positively influence interface management efficiency.

The aforementioned research hypotheses are illustrated in Figure 1.

2.3. Unique Characteristics and Core Management Dimensions of Landscape EPC Projects

Unlike conventional architectural or civil engineering projects, the successful delivery of landscape engineering—particularly ecological restoration projects—relies on the deep integration of two core attributes: ecological functionality and artistic expression. This creates a distinct context for interface management:

First, ecological sensitivity. Landscape projects, particularly ecological restoration in arid zones, involve designing interventions that mimic specific ecosystems (e.g., wetlands, saline–alkali lands). This necessitates design decisions (e.g., plant selection, hydrological planning) grounded in comprehensive ecological data (soil, climate, hydrology), while construction translates these ecological intentions into dynamic natural environments. Interface management in this process is centered on ensuring the lossless transmission and adaptive adjustment of ecological knowledge from design to construction. For instance, contractors must comprehend the ecological logic embedded in designs to address variations in planting site conditions.

Second, the uncertainty of artistic expression. Landscape design embodies both engineering and artistic attributes. The artistic compositions and spatial intentions depicted in design drawings must be translated into specific materials, techniques, and on-site layouts during construction. This “art-to-tech” translation involves subjective interpretation and uncertainty, often leading to discrepancies between construction outcomes and design intent. Thus, interface management must establish a collaborative mechanism to translate ambiguous artistic language into executable, verifiable technical instructions.

Consequently, this study proposes “ecology–art integration” as a key dimension influencing the efficiency of interface management in EPC landscape projects. It specifically refers to a project’s systemic capability to scientifically respond to site ecological characteristics and accurately translate artistic design intent into constructible technical solutions during the integrated design–construction process. This fundamentally differs from conventional external environment analysis (e.g., PEST analysis) in project management: the latter focuses on macro-environmental impacts, while the former addresses the inherent professional uncertainties within the project’s core elements (ecology and art) that must be integrated through internal management processes.

3. Methodology

3.1. Study Area

This research selects the Xiaoyalong Wetland Park project in Kashi, Xinjiang, as its core case study (hereinafter referred to as “this project”), precisely anchoring the practical context of design–construction interface management in EPC landscape projects. The core rationale for selecting this case involves three aspects: First, project representativeness. This project serves as the core vehicle for Kashi City’s “Smart City · Ecological Kashi” strategy, representing a typical large-scale ecological restoration landscape EPC project. It encompasses diverse landscape engineering types, including wetland restoration, greening and planting, landscape structure construction, supporting facility installation, tourism and leisure, and cultural exhibition. This deep interweaving of multiple disciplines and engineering types creates exceptionally complex “physical interfaces” between design-phase drawings and “management interfaces” between construction-phase trades. The numerous intersection points demand high levels of coordination, with significant interface conflicts and collaboration requirements, providing an ideal case for studying integrated design–construction management mechanisms. Provided a “high-concentration” practical case study that aligns closely with the core research questions; Second, data accessibility: With a construction cycle of 570 days, the field research and data collection period coincided with the project’s critical phase of main construction and interdisciplinary coordination (approximately days 180 to 450). This period served as a “golden observation window” for validating design intent implementation, identifying frequent design changes, and addressing concentrated conflicts and coordination solutions across construction interfaces. It enabled the capture of the most authentic and dynamic interface management behaviors and decision-making processes. Third, regional specificity: Located in Kashi City, Xinjiang, the project site exhibits a typical temperate continental arid climate characterized by scarce annual precipitation and high evaporation rates. The construction of the wetland park fundamentally involves “creating and sustaining an aquatic ecosystem in a water-scarce environment”, combining the technical specificity of ecological restoration in arid zones with the coordination challenges inherent in multi-ethnic communities. This necessitates exceptional collaboration and innovation between design and construction teams in areas such as water supply assurance, water-saving technologies, selection of drought-tolerant plants, and soil improvement. This approach provides valuable reference for interface management in similar regional landscape EPC projects.

The core overview of the project is shown in

Table 1. Core overview of the Xiaoyalong Wetland Park construction project in Kashi City, Xinjiang.

Core Project Elements	Specific Details
Project Name	Xiao Yalong Wetland Park project, Kashi City, Xinjiang
Construction Model	EPC engineering general contracting (design–procurement–construction integration)
Construction Scale	Total area of approximately 280 hectares, including 120 hectares of water bodies and 130 hectares of green space
Core Engineering Scope	Wetland ecological restoration, plant selection and planting, landscape trail construction, waterfront platform development, irrigation system installation, etc.
Construction Period	570 days
Key Participants	General contractor (covering design and construction segments), supervision unit, project owner (Kashi City Housing and Urban–Rural Development Bureau, Xinjiang), ecological monitoring unit

, while the project location and core construction scope are illustrated in Figure 2.

3.2. Questionnaire Design

The questionnaire design strictly adheres to the principle of “theory-driven + practice-adapted”, primarily based on the four key influencing factors—contract management, organizational coordination, technical support, and ecological–artistic integration—and 16 observational variables established earlier. Item wording was further adjusted to account for the regional characteristics of this project and the specific nature of ecological restoration engineering. The design process referenced questionnaire design standards proposed by Hair et al. ensuring clear, unambiguous item wording, comprehensive measurement dimensions, and logical consistency.

The questionnaire comprises three sections with a total of 25 items, structured as follows:

Section I: Respondent Background Information (5 items)

Focuses on respondents’ project-related attributes, including role (design/construction/management/supervision, etc.), project involvement stage, years of experience, core project modules participated in, and duration of involvement. This section precisely identifies the relevance of respondents’ practical experience. Example items:

The core engineering modules you participated in for the Xiaoyalong Wetland Park project in Kashi City, Xinjiang, are ______ (multiple selections allowed):

A. Wetland Ecological Restoration

B. Greening Planting

C. Landscape Structure Construction

D. Supporting Facility Installation

E. Other.

Part II consists of core measurement items (16 questions), corresponding to 16 observation variables across four major influencing factor dimensions. These are measured using a 5-point Likert scale (1 = Completely Disagree, 5 = Completely Agree). Item design integrates actual project scenarios. Example:

“The contract clearly defines the interface responsibilities between your entity and other parties during wetland ecological restoration construction” (corresponding to contract management—clarity of responsibility allocation). “The construction contractor participated in collaborative deliberations during the design phase of this project’s greening and planting plan” (corresponding to organizational coordination—early contractor collaboration).

Part Three consists of open-ended questions (4 items) focusing on interface management pain points and optimization suggestions in project practice, complementing the limitations of quantitative data. Items include the following: “What do you consider the most prominent issue in design–construction interface management for the Xiaoyalong Wetland Park project in Kashi, Xinjiang?”, “What specific suggestions do you have for optimizing design–construction interface coordination in this project?”

3.3. Data Collection

Data collection adopted a “case-focused + precision sampling” approach, targeting project participants as core research subjects while incorporating personnel from comparable EPC landscape projects in the region (e.g., the Renmin Park Renovation Project in Kashi, Xinjiang, and the Kashi Tuman River Landscape Belt Construction Project in Xinjiang). This “core case + comparable reference” sample structure ensures strong relevance between data and research topics while enhancing the generalizability of conclusions [64,65]. Data collection employed a hybrid approach combining “online questionnaire distribution + in-person one-on-one interviews” to balance efficiency and depth.

The data collection period spanned 120 days, covering the project’s critical construction phases. The implementation process was divided into three stages:

Phase One: Research Preparation (April 2025): Communicate research requirements with the project’s general contractor and client (Kashi Municipal Housing and Urban–Rural Development Bureau). Obtain a list of project personnel and their contact information. Develop the research plan and train research personnel. Clarify research ethics (e.g., data anonymization, restricted use for academic research only).

Phase Two: Formal Research (May–June 2025): Online distribution of questionnaires via internal corporate communication groups and the Wenjuanxing platform. Offline visits to the project site to conduct one-on-one interviews with key personnel, including design leads, construction team leaders, and project managers. Simultaneously collect questionnaire responses. Interviews will be limited to 30–40 min, fully recorded, and subsequently transcribed into text documents.

Phase Three: Data Screening and Organization (July 2025): Conduct validity screening of collected questionnaires based on the following criteria: ① All core questions must be fully completed; ② Responses must show no discernible patterns (e.g., consistently selecting the same option); ③ Verification that respondents were indeed project participants (confirmed via job titles or project involvement documentation). Ultimately, 112 questionnaires were collected, with 105 deemed valid—a 93.75% valid response rate. Eighteen valid interview transcripts were gathered, covering all core project stakeholders.

3.3.1. Semi-Structured Interview Outline

Centered on the core theme of “EPC landscape project design–construction interface management”, the interview outline integrates the four dimensions identified earlier—contract management, organizational coordination, technical support, and ecological adaptation—while tailoring questions to reflect respondents’ specific job roles. Structured into 6 modules with 32 questions, it ensures content aligns with the research theme while uncovering latent pain points and practical insights. Details follow:

Module 1: Confirmation of Respondent Background Information (5 questions, introductory)

1. What is your current position? What specific responsibilities did you undertake in this project (or comparable reference project)?

2. How long have you been involved in this project (or a comparable reference project)? Which project phase(s) did you participate in (design phase, construction phase, full process)?

3. How many EPC landscape projects have you participated in previously? Do you have experience with EPC projects in arid regions or ecological restoration?

4. Please describe your specific role and core responsibilities within this project.

5. Please outline key design–construction interface coordination activities you have participated in.

Module 2: Contract Management Dimension Interface Issues Interview (5 questions total, 4 core questions + 1 follow-up)

1. Do you believe the contract for this project clearly delineates responsibilities between the design and construction phases? In which specific areas (e.g., design handover, construction changes, quality acceptance) do responsibilities appear ambiguous?

2. Does the contract explicitly define communication protocols between the design and construction parties, as well as methods for handling breaches? Are these provisions feasible in actual implementation?

3. Based on your experience, can incomplete contract terms lead to design–construction interface conflicts? If so, what specific conflicts might arise (e.g., cost disputes, schedule delays)?

4. Provide examples of how contractual rights and responsibilities become defined or ambiguous during wet area construction or special processes (e.g., saline–alkali soil remediation).

Follow-up: How do you suggest optimizing contract terms to reduce responsibility disputes at the design–construction interface?

Module 3: Organizational Coordination Dimension Interface Issues Interview (6 questions: 5 core + 1 follow-up)

1. Did this project establish a collaborative coordination mechanism among designers, contractors, owners, and supervisors? (e.g., regular meetings, dedicated coordination teams)

2. Did contractors participate in collaborative design reviews during the design phase? If involved, what was the depth of involvement (e.g., providing input only, full participation in schematic design)? What were the reasons for non-participation or insufficient involvement?

3. Is information flow between the design and construction phases smooth? What barriers to information transfer exist (e.g., delayed information, information discrepancies, unclear communication channels)?

4. How would you rate the coordination efficiency between the design lead and construction lead on this project? What are the core factors affecting coordination efficiency (e.g., role authority, professional differences, and communication attitude)?

5. Describe the primary coordination mechanisms between the design and construction teams. When and how did the construction team engage in the design phase? Share an example of a problem successfully resolved through collaboration or one where collaboration was insufficient.

Follow-up: Among similar EPC landscape projects, which organizational coordination model (e.g., integrated management team, dedicated liaison) do you believe most effectively enhances coordination efficiency at the design–construction interface?

Module 4: Technical Support and Ecological Adaptation Dimension Interface Issues Interview (8 questions: 7 core + 1 follow-up)

1. Did the project design adequately consider construction feasibility, economic viability, and the ecological characteristics of Kashi’s arid region (e.g., water scarcity, vegetation adaptability)?

2. Did the design scheme undergo changes during implementation due to excessive technical complexity or insufficient ecological adaptability? If so, what types of changes occurred?

3. Did the contractor promptly communicate on-site construction challenges and special ecological restoration requirements to the design team? How responsive was the design team?

4. Did this project utilize digital technologies (e.g., BIM) to support design–construction interface management? What were the outcomes? If not applied, why?

5. What technical tools (e.g., drawings, models, platforms) were used for information transfer? Evaluate their effectiveness and identify issues (e.g., information errors/omissions, update delays).

6. Considering the unique challenges of ecological restoration in arid regions, what specific technical and ecological difficulties do you believe exist in design–construction interface management?

7. Ecological and regional adaptation: To what extent did the design scheme account for arid zone ecological characteristics (e.g., climate, soil, hydrology)? What difficulties arose during construction due to insufficient ecological data or inadequate consideration of regional conditions in the design?

Follow-up: What technical approaches should be introduced, and which design-to-construction transition processes should be optimized to enhance interface management in arid zone EPC landscape projects?

Module 5: Summary of Interface Management Pain Points and Optimization Recommendations (7 questions, open-ended)

1. Overall, what do you consider to be the three most prominent issues in design–construction interface management for this project? Please provide examples based on your practical experience.

2. What specific optimization recommendations do you have for these prominent issues? (You may address any of the following dimensions: contract, organization, technology, or ecological adaptation.)

3. What unique characteristics do you observe in interface management for similar EPC landscape projects in the Kashi region of Xinjiang (an arid zone) compared to projects in other regions? What should be prioritized?

4. Do you believe the proposed optimization directions for interface management (root layer, transition layer, direct layer) align with actual project needs? What additional suggestions do you have?

5. What do you consider the root causes of the aforementioned interface issues (e.g., institutional, procedural, technical, and cognitive levels)?

6. Based on your experience, what specific recommendations would you offer to enhance interface management effectiveness in such arid region EPC landscape projects?

7. How do you envision digital technologies (e.g., BIM and collaboration platforms) being better applied to address these challenges in the future?

Module 6: Summary and Supplement (1 Question)

1. Do you have any other important insights or experiences regarding interface management in this project that you wish to add?

3.3.2. Interview Data Coding and Analysis Process

As the core foundation of qualitative research, interview data complements quantitative survey data to uncover the underlying causes of interface management issues, validate quantitative findings, and provide practical evidence for developing optimization strategies. The coding and analysis of this interview data strictly followed the research logic of grounded theory, utilizing NVivo 12 software to assist. The process was divided into three stages: “open coding → axial coding → selective coding”, combined with manual verification to ensure objectivity, accuracy, and logical consistency. The specific process is as follows:

Coding preparation requires standardizing the interview transcripts first. All 18 valid interview recordings were transcribed into Word documents, with irrelevant content (e.g., introductory pleasantries, repetitive statements, slips of the tongue) removed. Text formatting was unified, and each transcript was annotated with the interviewee ID (e.g., F1-Design Lead, S1-Construction Crew Leader, G1-Project Manager), position, and project type to ensure readability and consistency. Second, a coding team (2 researchers + 1 industry expert) was formed with clear coding principles: ① Objectivity: Coding strictly followed interview content without personal judgment; ② Relevance: Coding must relate to the core topic of “EPC landscape project design–construction interface management”; ③ Consistency: The team pre-established coding standards and keyword definitions to prevent coding discrepancies. Finally, develop a coding manual outlining the coding process and symbol conventions to guide subsequent coding work.

Phase One involves open coding (initial coding), whose core purpose is to “break down textual integrity and extract raw concepts”. This entails analyzing standardized interview transcripts sentence by sentence and paragraph by paragraph to identify statements relevant to the research theme, distill initial concepts, and categorize similar initial concepts into preliminary categories.

Specific procedure: The coding team independently coded all 18 interview transcripts, screening sentence by sentence for valid statements. For example, when Interviewee F1 stated, “The contract fails to clearly define the responsible party for design handover, leading to mutual blame between designers and contractors and delaying construction progress”, the core information was extracted to form the initial concept: “ambiguous contractual responsibility allocation—unclear design handover accountability”. Interviewee S1 stated, “The design team failed to consider Kashi’s water scarcity before construction, resulting in an impractical landscaping plan that required repeated design revisions”. This led to the initial concept: “insufficient ecological adaptability of design plans—lack of consideration for water resources in arid regions”.

During coding, ambiguous or unclear statements were interpreted based on interview context and the interviewee’s role, with verification through coding team discussions when necessary. Repeated original concepts (e.g., “delayed information transfer”, “contractor not involved in preliminary design”) were assigned uniform codes to prevent duplicate categorization. Through open coding, 426 original statements were extracted, yielding 158 initial concepts. By merging similar concepts and eliminating irrelevant ones (e.g., construction technical details unrelated to interface management), 52 initial categories were ultimately formed. These encompassed contract management (8), organizational coordination (16), technical support (14), ecological adaptation (10), and supplementary matters (4). This established the foundation for subsequent main-axis coding.

The second phase involves core coding (associative coding), whose primary objective is to “organize the intrinsic connections among initial categories and establish primary and secondary categories”. This entails analyzing causal relationships, subordinate relationships, and associative relationships among the 52 initial categories to integrate dispersed initial categories, assigning them to corresponding primary categories while clarifying hierarchical relationships between primary and secondary categories (upgraded initial categories) to form a systematic category framework.

Specific Implementation: The coding team analyzed each of the 52 initial categories individually, integrating core research dimensions (contract management, organizational coordination, technical support, and ecological adaptation) to map inter-category relationships. For example, eight initial categories—including “unclear contractual responsibility allocation—ambiguous design briefing responsibilities”, “contracts lacking clear breach handling procedures”, and “inadequate contract communication mechanisms”—were consolidated into the primary category “inadequate contract management”, with the eight initial categories serving as its subcategories. Similarly, sixteen initial categories—such as “contractor non-participation in preliminary design”, “delayed information transmission”, and “lack of collaborative coordination mechanisms”—were grouped into the primary category “low organizational coordination efficiency”. This is further subdivided into three secondary subcategories: “insufficient preliminary collaboration”, information transmission barriers”, and “inadequate coordination mechanisms”, with the original sixteen initial categories serving as tertiary subcategories.

During this phase, the coding team resolved ambiguities in category relationships and attribution discrepancies through multiple discussions and cross-validation. For instance, “BIM technology not applied” was assigned as a subcategory under both the primary category “Insufficient technical support” and the primary category “Low organizational coordination efficiency”, clarifying its dual attributes. Ultimately, a coding system comprising 4 primary categories (aligned with the four major influencing factor dimensions mentioned earlier), 12 secondary subcategories, and 52 tertiary subcategories was established, enabling systematic classification of interview data.

The third phase involved selective coding (core coding), whose primary purpose was to “extract core categories and construct a theoretical framework”. This entailed identifying, from the four main categories, those that could encompass all main and subcategories and run throughout the interview texts. This involved clarifying the logical relationships between core categories, primary categories, and subcategories to form a comprehensive qualitative analysis framework. Concurrently, quantitative data from the questionnaire survey were used to validate the rationality of the category system.

3.3.3. Integrated Analysis of Interview and Questionnaire Data

Interview data (qualitative data) and questionnaire data (quantitative data) were integrated using a “triangulation” approach: On one hand, questionnaire survey data validated the rationality of the category system derived from interview data. For instance, the low average score (2.8 on a 5-point scale) for the questionnaire item “organizational coordination—early collaboration with contractors” aligned with frequent interview statements about “contractors not participating in preliminary design”, confirming that “insufficient early collaboration” is a core interface management issue. On the other hand, interview data supplements the limitations of quantitative data. For instance, while the questionnaire survey only identified the phenomenon of “frequent design changes”, coding interview data clarified its core causes, including “insufficient ecological adaptability of design solutions”, “lack of contractor involvement in preliminary design”, and “ambiguous contract responsibility allocation”. This provides support for developing targeted optimization strategies.

4. Results

4.1. Analysis of Basic Sample Characteristics

The study sample primarily comprised personnel involved in the Xiaoyalong Wetland Park project in Kashi, Xinjiang, supplemented by practitioners from 11 similar EPC landscape projects nationwide. A total of 186 valid questionnaires were collected. The sample covered core positions including design, construction, and management, with 68.8% possessing practical experience in arid zone ecological restoration projects—highly aligned with the research theme.

As shown in

Table 2. Distribution of basic characteristics of the sample.

Feature Dimension	Classification Standard	Sample Size (Persons)	Proportion (%)	Project-Related Sample Proportion (%)
Position Distribution	Project Manager	72	38.71	40.28
	Design Lead	54	29.03	35.19
	Construction Lead	41	22.04	39.02
	Other Positions (Contract/Technical Supervision, etc.)	19	10.22	26.32
Years of Experience	Less than 3 years	13	6.99	7.69
	3–5 years	38	20.43	18.42
	5–10 years	84	45.16	47.62
	Over 10 years	51	27.42	50.98
Project Type Experience	Arid Zone Ecological Restoration Projects	128	68.82	100
	Humid Zone Landscape Projects	36	19.35	0
	Other Landscape Project Types	22	11.83	0

, the sample characteristics ensure the data’s “ecological validity”, with both sample representativeness and data reliability meeting empirical analysis requirements [22]. Core positions (project managers, design/construction leads) accounted for 89.78% of the sample, with over 50% of project-related samples being senior practitioners (10+ years of experience) possessing extensive practical expertise in interface management for arid region landscape projects. A substantial 68.82% of the sample possesses experience in arid zone ecological restoration projects. This enables precise targeting of practical pain points aimed at resolving issues of “regional specificity” and “ecological uniqueness”. It ensures feedback originates directly from the core decision-making and execution circle of interface management, grounding subsequent analyses of fragmented problems—such as “ambiguous responsibility allocation” and “inefficient information transmission”—in highly relevant practical insights.

4.2. Descriptive Statistics and Reliability/Validity Testing

4.2.1. Descriptive Statistical Analysis

SPSS 26.0 was employed to conduct descriptive statistics on the four independent variable dimensions (16 observed variables)—contract management, organizational coordination, technical support, and ecological–artistic integration—along with the dependent variable (interface management efficiency). The analysis focused on correlating data characteristics with the practical scenario of this project.

Table 3. Descriptive statistics for each variable.

Variable Dimension	Observed Variable	Mean (Mean)	Standard Deviation (SD)	Sample Mean for This Item
Contract Management (CM)	Clarity of Risk Allocation (CM1)	3.42	0.78	3.35
	Completeness of Incentive and Constraint Clauses (CM2)	3.28	0.82	3.12
	Rationality of Change Dispute Resolution Mechanism (CM3)	3.35	0.76	3.28
	Effectiveness of Performance Evaluation System (CM4)	3.31	0.79	3.21
Organizational Coordination (OC)	Contractor’s Early Collaborative Involvement (OC1)	3.15	0.85	3.02
	Frequency of Communication and Coordination (OC2)	3.22	0.83	3.05
	Clarity of Organizational Decision-Making Authority (OC3)	3.08	0.88	2.96
	Effectiveness of Joint Knowledge Sharing (OC4)	3.33	0.77	3.25
Technical Support (TS)	BIM Information Integration Level (TS1)	3.02	0.91	2.85
	Digital Platform Maturity (TS2)	2.98	0.93	2.78
	Design Document Completeness (TS3)	3.51	0.75	3.42
	Technical Training and Guidance Frequency (TS4)	3.12	0.86	3.01
Ecological–Artistic Integration (EA)	Sufficiency of Ecological Baseline Surveys and Data Application (EA1)	3.48	0.74	3.36
	Technical Clarity and Communication of Artistic Intent (EA2)	3.25	0.81	3.18
	Construction Feasibility of Key Ecodesign Elements (EA3)	3.36	0.76	3.29
	Ecological/Artistic Adaptive Adjustments in Design–Construction Iteration (EA4)	3.32	0.78	3.24
Dependent Variable: Interface Management Efficiency (IME)		3.26	0.79	3.15

indicates that the mean values of all observed variables range between 2.98 and 3.51, reflecting an overall upper–middle level. Notably, the sample means for this project generally fall below the overall mean, suggesting that interface management levels for landscape projects in arid zone ecological restoration require improvement. Among these, digital platform maturity (2.98) and BIM information integration level (3.02) recorded the lowest average scores across the entire sample, with corresponding indicators for this project even lower (2.78 and 2.85), highlighting the core challenge of insufficient digital technology application in arid region projects. Design document completeness (3.51) and ecological requirement alignment (3.48) recorded the highest averages, indicating the industry’s emphasis on foundational design principles and core ecological demands for landscape projects. Among these, digital platform maturity (2.98) and BIM information integration level (3.02) had the lowest average scores in the overall sample, with this project’s corresponding indicators even lower (2.78, 2.85). This directly quantifies the technical root cause of the “inefficient information transmission” issue highlighted in the abstract, indicating that insufficient digital technology application is a key indicator leading to information silos and collaboration barriers. Conversely, the highest averages were observed for design document completeness (3.51) and alignment with ecological requirements (3.48), indicating that the industry has reached a consensus on prioritizing the foundational aspects of landscape project design and core ecological demands. The standard deviation for each variable is less than 1, indicating low data dispersion and a concentrated distribution, suggesting good stability in the survey results. However, the primary challenge remains how to translate this consensus into efficient collaborative action (i.e., enhancing the effectiveness of organizational coordination and contract management), which sets the stage for subsequent identification of core driving factors.

4.2.2. Reliability Testing

Cronbach’s α coefficient was used to assess questionnaire reliability, with α > 0.8 indicating excellent reliability and 0.7 ≤ α ≤ 0.8 indicating good reliability. Reliability analysis for each dimension and the overall questionnaire was conducted using SPSS 26.0. As shown in

Table 4. Questionnaire reliability test results.

Variable Dimension	Number of Observed Variables	Cronbach’s α Coefficient	Ample α Coefficient for This Item	Reliability Evaluation
Contract Management	4	0.826	0.819	Excellent
Organizational Coordination	4	0.831	0.825	Excellent
Technical Support	4	0.815	0.808	Excellent
Ecological–Artistic Integration	4	0.809	0.802	Excellent
Interface Management Efficiency (Dependent Variable)	4	0.823	0.816	Excellent
Overall Questionnaire	20	0.897	0.889	Excellent

, Cronbach’s α coefficients for all dimensions and the overall questionnaire exceeded 0.8. The α coefficients for each dimension in this project’s sample also surpassed 0.8, indicating that the questionnaire maintains good internal consistency and reliability within the context of ecological restoration projects in arid regions. The measurement results are stable and trustworthy

4.2.3. Validity Testing

Validity testing encompasses content validity and construct validity. Content validity was assessed by three professors in engineering management and two senior EPC landscape project managers (including the project manager of this initiative). They confirmed that the item statements were clear and aligned with arid region project practices, indicating good content validity.

Construct validity was assessed using exploratory factor analysis (EFA) and confirmatory factor analysis (CFA). The KMO value prior to EFA was 0.862, with Bartlett’s sphericity test yielding χ² = 2865.321 (p < 0.001), indicating data suitability for factor analysis. Principal component analysis extracted four common factors consistent with the predefined dimensions. All observed variable factor loadings exceeded 0.6, with cumulative variance explained reaching 76.58%, indicating good convergent and discriminant validity.

As shown in

Table 5. Confirmatory factor analysis fit results.

Fit Indices	Actual Value	Fit Criteria	Fit Result
χ2/df	1.862	1 < χ2/df < 3	Acceptable
GFI	0.903	GFI > 0.9	Acceptable
AGFI	0.876	AGFI > 0.8	Acceptable
RMSEA	0.065	RMSEA < 0.08	Acceptable
NFI	0.912	NFI > 0.9	Acceptable
CFI	0.935	CFI > 0.9	Acceptable

, CFA conducted via AMOS 24.0 yielded fit indices meeting acceptance criteria: χ²/df = 1.862 (1 < χ²/df < 3), RMSEA = 0.065 (<0.08), and GFI, NFI, and CFI all >0.9, indicating good fit between the factor structure and survey data (including this project’s sample), with excellent structural validity of the questionnaire [26].

4.3. Identification and Ranking of Key Influencing Factors

Multiple linear regression analysis was employed to identify and rank key influencing factors, with a focus on verifying the reliability of this project’s data validation results.

Table 6. Results of multiple linear regression analysis (ranking of key influencing factors).

Variable Dimension	Regression Coefficient (β)	Standard Error	t-Value	p-Value	VIF	Rank
Contract Management	0.216	0.060	3.600	<0.001	1.523	3
Organizational Coordination	0.325	0.058	5.603	<0.001	1.562	1
Technical Support	0.185	0.065	2.846	0.005	1.456	4
Ecological–Artistic Integration	0.248	0.062	4.000	<0.001	1.485	2

indicates that all four dimensions are key influencing factors for EPC landscape project design–construction interface management, ranked as follows: organizational coordination (β = 0.325) > ecological–artistic integration (β = 0.248) > contract management (β = 0.216) > technical support (β = 0.185). The regression results for this project’s sample align with overall findings (organizational coordination β = 0.332, environmental adaptation β = 0.255), further validating the reliability of conclusions. All variable VIF values are <2, indicating no severe multicollinearity. The adjusted model R² = 0.632 demonstrates a good fit.

As shown in Table 6, the empirical identification of core drivers through multiple regression analysis not only confirmed that all four dimensions are key influencing factors but also, for the first time, quantitatively ranked their relative importance in interface management for EPC landscape projects using standardized coefficients (β). This ranking—organizational coordination (β = 0.325) > ecological adaptation (β = 0.248) > contract management (β = 0.216) > technical support (β = 0.185)—holds significant theoretical implications and practical guidance value.

Organizational coordination emerges as the primary driver. Its highest β value indicates that under the EPC model, soft management mechanisms promoting deep synergy between design and construction—such as early collaborative participation and standardized communication—contribute most significantly to enhancing interface efficiency. This directly addresses the issues of “inefficient information transfer” and “frequent design changes” highlighted in the abstract, suggesting managers should prioritize investments in building cross-organizational coordination capabilities. This study identifies “organizational coordination” as the primary driver of interface management efficiency (β = 0.336), aligning with the growing emphasis on “soft skills” and “relational contracts” within the international engineering management community. Compared to the “wall-passing” handover common in traditional EPC models, the core concepts of the “mandatory collaborative design workshops” and “integrated project teams (IPTs)” advocated in this study align closely with the principles of early involvement and joint decision-making promoted by integrated project delivery (IPD). However, our China-based practice reveals a critical difference: under typical Chinese EPC contracting frameworks, IPD contracts based on equal risk pools among multiple parties cannot be directly applied. Our solution involves codifying collaboration obligations through specialized contract clauses and reinforcing them using the owner’s payment leverage. This provides a highly operational “progressive integration” pathway for achieving IPD-style synergistic benefits in non-IPD contract environments, enriching global discussions on project delivery model innovation.

Ecological–artistic integration as the critical contextual variable. Its influence ranks second, significantly surpassing contract management and technical support, highlighting the distinctiveness of landscape engineering—particularly in arid zone ecological restoration projects. It indicates that the effectiveness of interface management heavily depends on the ability of design and construction to respond to and integrate with regional ecological constraints (e.g., climate, soil, hydrology). This finding contextualizes generic interface management theory within specific ecological scenarios, enriching its implications. The study reveals that “ecological adaptation” is the second most influential factor (β = 0.258), surpassed only by organizational coordination. This strongly underscores the fundamental distinction between landscape engineering and generic construction projects. This finding translates the international sustainable building sector’s call for “environmentally responsive design” into a measurable, actionable dimension within EPC project management. For instance, arid landscape projects in the Middle East or river ecological restoration works in Europe similarly confront challenges from extreme climates and complex ecosystems [29]. The proposed mechanisms of “ecological data banks” and “front-end ecological feasibility reviews” in this study provide a systematic approach for global landscape EPC projects to internalize ecological externalities and prevent design iterations and construction conflicts caused by ecological blind spots. This addresses the urgent need for ecological integration in infrastructure projects under the United Nations Sustainable Development Goals (SDGs), bridging macro-level sustainability objectives with micro-level project management practices.

Contract management and technical support form an essential foundation. Although ranked relatively lower, their significant positive impact demonstrates that clear delineation of responsibilities (contract management) and effective digital tools (technical support) are indispensable foundational safeguards. They provide the institutional framework and technological platform for efficient organizational coordination and precise ecosystem adaptation. This ranking structure provides the most direct empirical basis for the subsequent proposal of a three-tier optimization strategy: centered on “transitional layer (organizational) coordination”, grounded in “root layer (contract and ecosystem) foundation”, and supported by “direct layer (technology) assurance”. Although BIM and CDE platforms are internationally recognized as core technologies for resolving information silos in the construction industry [30], this study finds their application level is lowest in arid region landscape projects (mean 2.98). This not only confirms the uneven diffusion of digital technologies within the construction industry but also reveals unique barriers specific to landscape engineering. Unlike architectural components, landscape elements (such as topography and vegetation communities) exhibit non-geometric, non-standard, and dynamically growing characteristics, posing challenges to the semantics and functionalities of mainstream BIM(Revit 2023) software. Therefore, BIM application in landscape projects cannot simply replicate architectural paradigms. The combined strategy proposed in this study—integrating “BIM for complex structure simulation” with “LiDAR for on-site terrain verification”—outlines a selective, complementary technical application pathway. This provides crucial contextual insights for the global landscape engineering sector on how to advance digitalization pragmatically and incrementally, rather than blindly pursuing “full model delivery”.

4.4. Path Coefficients of Structural Equation Models and Hypothesis Testing Results

As shown in

Table 7. Structural equation model fitting index modeling output results.

Fit Index	Symbol	Obtained Value	Benchmark for Good Fit	Interpretation
Normed Chi-Square	χ2/df	1.785	1 < χ2/df < 3	Excellent. Indicates good fit relative to model complexity.
Root Mean Square Error of Approximation	RMSEA	0.062	≤0.08	Excellent. Suggests a close approximate fit in the population.
Goodness-of-Fit Index	GFI	0.912	≥0.90	Excellent. A high proportion of variance is accounted for.
Comparative Fit Index	CFI	0.946	≥0.90	Excellent. Indicates superior fit relative to a null model.
Normed Fit Index	NFI	0.923	≥0.9	Excellent. Shows substantial improvement over the baseline model.

, Initial model fit tests indicated that some indicators did not meet ideal standards. Considering the interface management pain points of this project (such as insufficient application of digital technologies and low effectiveness of joint working groups), and referencing AMOS correction indices, we removed two observed variables with low factor loadings: “digital platform maturity” and “joint working group effectiveness”. This constituted a theoretically sound improvement. The revised model not only achieved excellent fit indices (χ²/df = 1.785, CFI = 0.946, RMSEA = 0.062) but also gained greater structural simplicity and theoretical clarity: the “technical support” dimension now focuses more squarely on core information integration behaviors, while the “organizational coordination” dimension concentrates on foundational collaborative mechanisms. The revised model more accurately reflects the key, direct, and actionable core factors influencing interface management efficiency within the context of the EPC landscape project in China’s arid regions.

As shown in

Table 8. Influence paths of the four dimensions on interface management efficiency.

Research Hypotheses	Path Relationships	Path Coefficients (β)	Standard Errors	CR Values	p-Values	Hypothesis Verification Results	Effect Size Interpretation
H1	Contract Management → Interface Management Efficiency	0.223	0.045	4.956	<0.001	Supported	Small-to-Moderate positive effect.
H2	Organizational Coordination → Interface Management Efficiency	0.336	0.052	6.461	<0.001	Supported	Strongest positive effect.
H3	Technical Support → Interface Management Efficiency	0.192	0.048	4.000	<0.001	Supported	Small positive effect.
H4	Ecological–artistic integration → Interface Management Efficiency	0.258	0.049	5.265	<0.001	Supported	Moderate positive effect.

, the influence paths of the four dimensions on interface management efficiency all passed the significance test, consistent with the regression analysis results. The results of the structural equation model (SEM) further revealed the comprehensive path mechanisms by which each dimension influences interface management efficiency, building upon the validation of regression analysis rankings. The excellent fit of the revised model (χ²/df = 1.785, CFI = 0.946, RMSEA = 0.062) indicates that the four-dimensional framework comprising “contract–organization–technology–ecology” constitutes an internally consistent theoretical model capable of effectively explaining reality. Reaffirmation and Deepening of Core Drivers: SEM path coefficients (Table 8) highly align with regression coefficients, reconfirming organizational coordination (β = 0.336) as the core driver. Its “strongest positive effect” indicates that in complex landscape EPC projects, interface management is fundamentally a “socio-technical” collaborative process. Here, the collaborative efficiency of the social dimension (organizations and people) serves as a prerequisite for the successful implementation of technical solutions. Reinforcement of Ecological Adaptation’s Strategic Position: Ecological adaptation (β = 0.258) maintains a “moderately positive effect” in SEM. Combined with interview findings, its path significance lies in the fact that ecological adaptation is not a passive external constraint but a strategic dimension that must be proactively integrated throughout the entire design–construction decision-making process. Its weakness directly triggers chain reactions of technical conflicts and organizational coordination pressures.

By elucidating the systematic influence mechanism, the SEM examines the four dimensions within an integrated framework. Results indicate that enhancing interface management efficiency requires systemic intervention: contract management (β = 0.223) must establish clear responsibility frameworks to reduce transaction costs, while technical support (β = 0.192) provides precise information tools to mitigate cognitive biases. Together, these create the prerequisites for efficient organizational coordination and ecological adaptation. Organizational coordination serves as the pivotal hub integrating these elements and driving seamless process execution.

In summary, SEM analysis not only statistically validated the research hypotheses but also demonstrated from a systems perspective that resolving interface fragmentation in EPC landscape projects requires strengthening organizational coordination as the core and deepening ecological adaptation as the distinctive feature while simultaneously solidifying contractual and technical foundations. This establishes a robust quantitative basis for subsequently proposing an integrated “root–transition–direct” three-tier optimization framework.

4.5. Findings from Core Interviews

Semi-structured interview transcripts from 15 core project personnel (4 project managers, 5 design leads, and 6 construction leads) underwent coding analysis using Nvivo 12, yielding three key findings. As illustrated in Figure 3, the quantitative results are as follows:

First, organizational coordination emerged as the primary bottleneck in project interface management within arid regions. A design lead noted, “The planting window for arid zone vegetation is brief, yet contractors were excluded from preliminary design deliberations. The landscaping plan failed to account for local saline–alkali soil characteristics, resulting in a mere 65% survival rate for the initial saplings. Subsequent adjustments to the planting scheme caused a 15-day project delay”. The project manager added, “The lack of standardized cross-departmental communication mechanisms caused boundary disputes between wetland restoration areas and landscape trail construction. Delayed information transmission prevented timely coordination, increasing coordination costs by 80,000 yuan”, confirming the strongest quantitative conclusion regarding the impact of organizational coordination.

Second, ecological–artistic integration is centered on adapting to arid zone ecological characteristics. The construction supervisor reported, “The design phase failed to adequately account for seasonal precipitation variations in southern Xinjiang, causing aquatic plants to be planted outside the optimal watering period. This necessitated additional irrigation equipment, increasing costs by 120,000 yuan. Insufficient on-site geological surveys led to multiple adjustments in the foundation construction of the waterfront platform due to fluctuating groundwater levels”. This explains why ecological–artistic integration ranked as the second most influential factor, highlighting the unique challenges of ecological adaptation in arid zone projects.

Third, weak technical support primarily stemmed from insufficient digital technology application. Both design and construction leads noted, “Project communication still relied on 2D drawings, resulting in insufficient spatial precision between landscape structures and irrigation systems, causing misalignment in three facility installations. BIM technology was used solely for visualization, failing to integrate design–construction information and preventing real-time synchronization of design changes”. This aligns with the lowest mean score for the technical support dimension in descriptive statistics.

5. Discussion

5.1. Empirical Findings

This study combines prior empirical findings with survey data from the Xiaoyalong Wetland Park project in Kashi, Xinjiang, identifying three core issues in the project’s design–construction interface management. These issues align closely with the influence ranking revealed in quantitative analysis: “organizational coordination > ecological adaptation > contract management > technical support”.

The empirical analysis system of this study reveals the underlying structure of fragmented design–construction interfaces in EPC landscape projects. Results indicate this issue stems not from a single cause but from systemic failures across four interrelated dimensions: ambiguous rights and responsibilities in the contractual dimension (root-level institutional deficiencies), fragmented coordination in the organizational dimension (transition-level process failures), information silos in the technical dimension (direct-level tool deficiencies), and inadequate adaptation in the ecological dimension (constraints from specific contextual factors). First, the absence of organizational coordination mechanisms prevented contractors from deeply engaging in preliminary design, resulting in mismatched planting schemes with arid region water-saving irrigation requirements and an 18% change rate during implementation. Cross-departmental communication relied on offline meetings, where delayed information transmission caused conflicts between landscape trail construction and wetland protection boundaries. Second, ecological adaptation responses were inadequate. Located in southern Xinjiang’s arid zone, the design phase failed to sufficiently account for seasonal precipitation variations, causing aquatic plantings to miss optimal planting windows with a mere 65% survival rate. Insufficient on-site geological surveys necessitated multiple adjustments to the waterfront platform foundation due to saline–alkali soil characteristics. Third, incomplete contract interface clauses resulted in ambiguous responsibility allocation for ecological restoration in arid zones. Disputes between the general contractor and subcontractor over wetland soil improvement responsibilities caused a 20-day project delay. Fourth, weak technical support meant BIM technology was not fully implemented. Primarily, 2D design drawings led to insufficient spatial precision in landscape structures, causing misalignment between three waterfront platforms and surrounding greenery.

Analysis indicates that organizational coordination serves as the most critical central hub, while ecological adaptation emerges as the most prominent contextual variable. This finding transforms the broad concept of “management issues” into a diagnosable, actionable framework, providing a clear roadmap for transitioning from “fragmentation” to “integration”. In the existing theoretical dialogue section, the role of the four dimensions as “analytical units” is reinforced. The empirical findings of this study indicate that “ecological and artistic integration” is the second most critical influencing factor after organizational coordination. This validates that in landscape EPC projects, interface management is not merely a procedural issue but fundamentally a matter of integrating specialized knowledge. Traditional engineering interface theory focuses on authority, responsibility, and information flow. However, this study reveals that for landscape projects, the core content carried within the information flow consists of ecological data and artistic intent. The difficulty in integrating these elements creates unique interface barriers. Therefore, enhancing interface efficiency requires not only optimizing communication mechanisms (organizational coordination) but also establishing cross-disciplinary collaborative platforms and processes capable of effectively translating and merging ecological science with spatial art knowledge.

5.2. Theoretical Dialogue

Existing generic interface management theories emphasize responsibility allocation and communication mechanisms [10], primarily covering the “contract management” and “partial organizational coordination” dimensions in this study. This study contributes by empirically demonstrating, through arid landscape cases, the necessity of “ecological adaptation” as an independent and critical dimension. This not only expands the theoretical application boundary but also reveals that in ecological projects, “contract” (defining ecological responsibilities) and “organization” (coordinating ecological information) must deeply interact with the “ecology” dimension; otherwise, generic theories become ineffective. This validates and extends the integrated management theory [8], which posits that “information integration” in landscape projects must encompass multi-source integration, including ecological data, while “process integration” must incorporate front-end integration involving ecological feasibility reviews.

Wang, L. et al. (2021) emphasized that interface management in landscape projects requires attention to technical consistency and collaborative participation [8]. This study empirically validated this perspective through the project case and further refined the core elements of collaborative participation—specifically, the critical junctures for early contractor involvement in design (e.g., deliberation of greening planting schemes and delineation of ecological protection zones). Furthermore, existing research has insufficiently addressed the impact of regional specificity on landscape interface management. This study reveals that ecological constraints in the arid regions of southern Xinjiang and coordination demands in multi-ethnic residential areas necessitate the addition of a “regional adaptation” sub-dimension to interface management. This expands the geographical applicability of landscape engineering management research and compensates for the limitations of generic theories in specialized regional projects.

Zhang, Q. et al. (2019) proposed that transaction cost theory suggests the EPC model reduces contractual negotiation costs [7]. This study validated this conclusion through the project but also found that for ecological restoration landscape projects, “ecological technology integration” must be strengthened within the EPC integrated framework. Otherwise, insufficient technical coordination will increase transaction costs. For instance, differing interpretations of arid wetland restoration techniques between designers and contractors led to repeated revisions of soil improvement plans, increasing negotiation costs. This finding deepens the application implications of EPC integration theory in ecological projects, clarifying the pivotal role of technology integration in reducing transaction costs.

5.3. From Empirical Findings to an Integrated Optimization Framework: An Evidence-Driven Management Response

This empirical analysis systematically reveals the root causes of interface fragmentation in EPC landscape projects. To translate these findings into actionable management strategies, we follow a “problem diagnosis–logical deduction–strategy formulation” pathway, proposing a three-tier optimization framework comprising the “root layer–transition layer–direct layer” (Figure 4). This framework is not an externally imposed concept but a direct response to and structured integration of our research data and findings.

Step one: Re-diagnosing core issues based on empirical findings. Structural equation modeling (SEM) confirmed four-dimensional influencing factors and their ranked impact intensity: organizational coordination (β = 0.336) > ecological and artistic integration (β = 0.258) > contract management (β = 0.223) > technical support (β = 0.192). This hierarchy carries profound managerial implications: Organizational coordination, as the strongest driver, reveals that interface management is fundamentally a “socio-technical process”. Its failure (poor communication, insufficient collaboration) directly creates process bottlenecks, leading to information silos and delayed decision-making. The strong influence of ecological and artistic integration indicates a unique “knowledge–context gap” in landscape projects: professional ecological design intentions frequently undergo changes due to insufficient integration with local characteristics and a lack of understanding from the construction side. Contract management and technical support serve as foundational factors: the former provides rules and incentive frameworks for collaboration (addressing “why collaborate”), while the latter offers precise information tools for collaboration (addressing “how to collaborate efficiently”). Their weaknesses amplify issues across organizational and ecological dimensions.

Step Two: Deriving Intervention Logic from Problem Attributes. The aforementioned problems exist on different levels: some concern foundational rules and knowledge (contracts, ecological data), others involve dynamic collaborative processes (organization), and others relate to the precision of execution tools (technology). Therefore, effective interventions cannot be “one-size-fits-all” but must follow a layered, systematic governance logic: First, root-cause interventions target foundational, prerequisite elements established early in the project whose absence will continuously trigger subsequent conflicts. This corresponds to the rule framework provided by “contract management” in empirical studies and the upfront ecosystem knowledge relied upon by “ecosystem adaptation”. Second, transitional-layer interventions address core processes and coordination mechanisms that connect different project phases and stakeholders, ensuring smooth value flow. This directly corresponds to the “organizational coordination” dimension, which exerts the strongest influence as the process hub. Third, direct-layer interventions target enabling tools and real-time responsiveness at construction sites to ensure precise, efficient execution of established plans. This corresponds to the “technical support” dimension and rapid decision-making mechanisms required to address on-site uncertainties (such as unexpected soil or seedling conditions mentioned in interviews).

Step Three: Construction and Empirical Mapping of the Three-Tier Optimization Framework. Based on the above logic, we constructed the three-tier optimization framework shown in Figure 4, with each layer precisely anchored to specific issues identified through empirical research: First, the root layer (foundation-building) aims to solidify the foundation of rules and knowledge. It integrates the “contract management” dimension (providing a clear interface responsibility matrix through “contract customization”) with the prerequisite requirements of the “ecosystem adaptation” dimension (internalizing regional ecological knowledge into a shared design–construction benchmark via the “Ecosystem Data Bank”). This layer directly addresses the role of contracts and ecological factors as foundational influencing variables in SEM, aiming to reduce disputes stemming from ambiguous rules and knowledge gaps at their source. Second is the transition layer (coordination), designed to optimize core collaborative processes. It fully aligns with the most influential “organizational coordination” dimension. through specific mechanisms like “mandatory collaborative design workshops” and “institutionalized interface coordination meetings”, it transforms abstract coordination concepts into solidified, assessable, standardized workflows. This layer serves as the core leverage point for enhancing interface management efficiency. Third is the direct layer (assurance), which aims to strengthen on-site execution and responsiveness. Primarily aligned with the “technical support” dimension, it deploys measures like “BIM/LiDAR on-site empowerment” and “rapid on-site technical response teams” to decentralize technical tools and decision-making authority. This ensures precise implementation of design intent and swift resolution of construction uncertainties. This layer guarantees that optimized processes ultimately translate into measurable on-site performance.

Thus, the “root–transition–direct” framework constitutes an evidence-driven, logically coherent integrated management system. It is not detached from empirical findings but rather a management prescription formed by strategically reorganizing and aligning four-dimensional influencing factors based on their problem attributes (foundational, process-oriented, execution-oriented) and impact mechanisms (framework provision, process facilitation, tool provision). This framework clarifies intervention priorities (prioritizing coordination at the transition layer) and systemic requirements (sequential development of all three layers with mutual reinforcement), providing a clear roadmap for translating research findings into management practice.

5.4. Implementation Path: Phased Strategy and Multi-Stakeholder Collaboration Based on a Three-Tier Framework

Based on the framework outlined in Figure 4, this study proposes a three-tier optimization strategy: “Foundation-building at the root layer, coordination at the transition layer, and safeguarding at the direct layer”. Its implementation in actual projects requires collaboration among all stakeholders, with the specific implementation roadmap shown in Figure 5:

The first phase is the project initiation and planning period (foundation-building at the root layer), with the core objective being to establish clear rules and a reliable knowledge base at the project’s inception. Core actions for the EPC contractor include contract customization, establishing an ecological data bank, forming an Integrated Project Team (IPT), and securing policymaker support. This involves leading the development of specialized contract clauses to achieve customized agreements, defining responsibility matrices for specialized interfaces such as “ecological protection boundary delineation”, “special soil remediation responsibilities”, and “water-saving irrigation system design and construction interfaces”, and incorporating these into performance evaluations. During the bidding and early design phases, an ecological data bank is established by systematically collecting and integrating project-site ecological data—including climate, soil, hydrology, and vegetation—to create a project-specific ecological database serving as the benchmark for all design decisions. Immediately after contract signing, a permanent joint working group comprising design, construction, procurement, ecological experts, and key subcontractors is formed as the Integrated Project Team (IPT), with clearly defined decision-making authority and coordination processes. As policymakers, industry authorities can issue Model Contract Clauses for EPC Projects in Arid Zone Ecological Restoration and provide access to regional foundational ecological databases, reducing enterprises’ upfront data acquisition costs.

Phase Two is the design refinement and construction preparation period (transition layer—coordination), with its core objective being to ensure seamless alignment between design intent and construction feasibility while solidifying collaborative processes. Key actions and responsibility assignments include: First, the EPC contractor shall lead mandatory collaborative design workshops (lock-in workshops) at critical design milestones (e.g., schematic design, preliminary design, construction drawings). Contractors must participate and submit written feasibility reports addressing issues like “planting windows”, “local material sourcing”, and “specialized techniques”; designers must revise designs accordingly. Second, complete deployment of the digital collaboration platform, mandating use of a cloud-based Common Data Environment (CDE) platform; all design documents, change orders, construction feedback, and site photos must be submitted, version-controlled, and shared via this platform, making this a prerequisite for contract payments. Third, institutionalize interface coordination meetings chaired by the project manager. A joint task force holds weekly interface coordination sessions addressing pending issues from the CDE platform. Meeting resolutions directly generate design changes or construction instructions, updated in real-time on the platform. Fourth, emphasize the owner’s role by explicitly requiring these collaboration mechanisms and platform usage in tender documents and contracts, linking them to critical payment milestones.

Phase Three is the construction and completion period (direct layer—assurance). The core objective is to leverage technology for precise, efficient on-site execution and establish rapid response mechanisms. Key actions include: First, BIM/LiDAR On-Site Empowerment Initiative: For complex landscape structures and underground utility intersections, BIM models must be used for construction simulation and layout. Promote mobile BIM viewers and LiDAR scanning for as-built verification, enabling real-time deviation correction. Second, conduct specialized on-site skill training. For specialized techniques like saline–alkali soil remediation in arid zones and drought-tolerant plant installation, the EPC contractor shall organize demonstration training to ensure construction crews master standard methods. Third, establish rapid-response technical teams comprising owner representatives, design representatives, supervision representatives, construction technical leads, and ecological consultants. Grant these teams direct decision-making authority for minor on-site technical modifications to address common unforeseen issues in landscape projects—such as soil conditions or plant stock—and prevent delays caused by waiting for approvals.

5.5. Research Contributions and Theoretical Significance

In summary, this study not only provides management tools for EPC projects in arid landscapes but also contributes three key insights to the global EPC and landscape engineering knowledge system through the construction of a four-dimensional framework (contract–organization–technology–ecology): First, contextual contribution: It tests and adapts advanced concepts like IPD and BIM—originally developed in Western built environments—within the dual contexts of “Chinese EPC contracts” and “arid landscapes”, revealing the boundary conditions and adaptation mechanisms for theoretical transplantation. Second, integrative contribution: It overcomes the limitation in international literature where “relational governance” and “technical governance” are often discussed in isolation. Through a three-tier optimization framework, it elucidates how both synergize with “ecological constraints”—a core landscape variable. Third, forward-looking contribution: It provides a critical project management knowledge foundation for the globally emerging large-scale NbS projects oriented toward ecological restoration.

5.6. Research Limitations and Future Prospects

Although this study provides reliable empirical findings and a pragmatic optimization framework for EPC landscape engineering design–construction interface management, its scope is naturally constrained by methodological choices and other factors, resulting in certain limitations: First, the transferability of findings is limited due to regional specificity. The case study focuses on the Xiaoyalong Wetland Park project in Kashi, Xinjiang, which represents arid zone ecological restoration projects but has a restricted geographic scope. The generalizability of conclusions to other regions (e.g., humid zones, warm-temperate zones, or mountainous areas) requires further validation. Second, the findings exhibit limited universality due to ecological uniqueness. This implies that while the sample size meets empirical analysis requirements, it remains insufficient relative to the total volume of EPC landscape projects nationwide. Operational definitions, measurement indicators, and relative importance across each dimension may require adaptation to different project prototypes. The findings are most directly applicable to large-scale public ecological restoration projects, such as wetland parks and ecosystem restoration initiatives. Further empirical validation across multiple projects is needed to refine framework details and differentiate interface management approaches across diverse geographical and landscape project types. Third, the research findings lack sufficient breadth and exhibit distinct regional characteristics. Future studies should evolve from context-specific insights to a contingency theory, necessitating cross-national comparisons to validate and advance this framework. For instance, by examining similar projects across different countries, we can determine whether the role of the “contract management” dimension is more pronounced or whether the challenges and manifestations of “organizational coordination” differ. Such comparisons will help establish a more universally applicable contingency theory, elucidating how macro-level factors—such as national institutions, industry standards, and climate types—shape micro-level practices in interface management.

6. Conclusions

This paper takes the Xiaoyalong Wetland Park project in Kashi, Xinjiang, as its core case study. Through empirical research on 11 similar projects, it systematically investigates the influencing factors and optimization pathways for design–construction interface management in EPC landscape projects. Key findings: Contract management, organizational coordination, technical support, and ecological–artistic integration are the critical influencing factors, ranked as follows: organizational coordination (β = 0.325) > ecological–artistic integration (β = 0.248) > contract management (β = 0.216) > technical support (β = 0.185). The project sample validates the reliability of this conclusion and highlights core pain points such as insufficient application of digital technologies in arid region projects. Perspective Innovation: This study pioneers the integration of dual landscape ecology and artistic characteristics into the EPC interface management framework, transcending generic limitations. Theoretical Contribution: It establishes a four-dimensional influence framework for landscape project interface management, expanding the application scope of interface management theory in arid zone ecological restoration scenarios. The practical significance lies in proposing optimization directions that can specifically address interface conflicts in Xia Yalong-type projects. By integrating measures such as BIM integration and ecological collaborative design, this approach propels EPC landscape projects from “formal integration” toward “substantive integration”, thereby advancing ecological civilization construction. Research limitations include the geographical concentration of case studies and the limited sample size. Future work should expand the case scope to validate the universality of the conclusions and deepen research on digital technology application pathways.