Next Article in Journal
Enhancement of Ecosystem Multifunctionality in Altay Natural Mowing Grasslands by Mixed Grass Species Overseeding
Previous Article in Journal
Grain Sorghum as a Climate-Resilient Alternative to Maize: Evapotranspiration, Water-Use Efficiency, and Yield Under Weed Competition and Reproductive-Stage Drought
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Agronomy
Volume 16
Issue 11
10.3390/agronomy16111111
agronomy-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Assessing the Effects of Large-Span Flexible Photovoltaic Arrays on Farmland Microclimate and Wheat Productivity: A Two-Year Field Experiment

by
Yanfei You
1,
Minli Yu
1,
Xiayun Geng
2,
Jiaxun Teng
3,
Zhonghao Qu
1,
Long Zhang
2 and
Encai Bao
2,4,*
1
School of Electrical Engineering, Hebei Vocational University of Technology and Engineering, Xingtai 054000, China
2
Institute of Agricultural Facilities and Equipment, Jiangsu Academy of Agricultural Sciences, Nanjing 210014, China
3
School of Electrical Engineering, Yanshan University, Qinhuangdao 066004, China
4
Key Laboratory of Protected Agricultural Engineering in the Middle and Lower Reaches of Yangtze River, Ministry of Agriculture and Rural Affairs, Nanjing 210014, China
*
Author to whom correspondence should be addressed.
Agronomy 2026, 16(11), 1111; https://doi.org/10.3390/agronomy16111111
Submission received: 1 May 2026 / Revised: 31 May 2026 / Accepted: 2 June 2026 / Published: 4 June 2026
(This article belongs to the Section Farming Sustainability)
Browse Figures
Versions Notes

Abstract

Agrivoltaics is an important pathway for promoting the coordinated development of clean energy production and agricultural utilization. However, the structural characteristics of flexible agrivoltaic (AV) systems may significantly alter field light and thermal conditions, while their effects on crop growth and yield formation remain unclear. To address this issue, a flexible AV system in Sihong County, Jiangsu Province, was selected as the study site, and continuous field monitoring combined with crop measurements was used to evaluate changes in microclimate, wheat physiological responses, and yield performance. The results showed that the flexible AV system significantly changed the field microclimate. During the wheat growing season, the monthly average solar radiation intensity under and between PV panels decreased by 62.0% and 56.9%, respectively, compared with that in the open field. The array also showed a certain thermal regulation effect, with heat preservation during the overwintering stage and cooling during the later growth stage. Shading reduced wheat net photosynthetic rate and stomatal conductance, but adaptive responses such as increased leaf area and chlorophyll content were observed. Wheat yield within the flexible AV system was significantly lower than that in the open field, with reductions of 43.4% and 47.2% in 2024 and 41.8% and 44.6% in 2025 for the areas under and between PV panels, respectively. Overall, light reduction under high coverage conditions remained the main factor limiting wheat yield. These results provide a theoretical basis for structural optimization and crop selection in flexible AV systems.

1. Introduction

Agrivoltaics is an integrated land-use model that combines photovoltaic (PV) power generation with agricultural production [1,2]. In recent years, it has shown promising application potential for open-field agricultural applications. By installing PV modules above farmland, such systems not only generate clean energy, but also expand agricultural production space by redistributing light and thermal resources [3]. On this basis, flexible agrivoltaic (AV) arrays have gradually emerged as a new structural form [4]. They are characterized by large spans and high ground clearance. Compared with conventional rigid support systems, they provide higher land-use efficiency and better adaptability to agricultural machinery [5]. However, while improving spatial utilization efficiency, these structures also modify field microclimatic conditions, such as light and temperature. This further affects the crop growth environment [6,7]. Therefore, it is important to investigate the changes in the light and thermal environment and their impacts on agriculture under large-span flexible AV systems.
In recent years, extensive research on agrivoltaic (AV) systems has been conducted by scholars in China and abroad, with particular attention to land-use co-benefits, microclimate regulation, and crop responses [8,9]. Existing studies have shown that PV panel coverage can significantly alter environmental factors in farmland, including solar radiation, air temperature, and soil temperature, thereby further affecting crop growth, physiological processes, and yield formation [10]. Santra et al. [11] investigated an AV system with three panel-density configurations and analyzed the effects of panel shading on photosynthetic photon flux density within the system. Zhang et al. [12] systematically evaluated the effects of fixed-tilt AV arrays with different installation heights on the light environment and fig tree growth. Jo et al. [13] assessed the effects of a vertical AV system in Korea on rice yield and its yield components through a two-year field experiment, and found that the system did not cause a significant reduction in rice yield overall, with yield variation being mainly driven by interannual environmental conditions. Hu et al. [14] found that semitransparent PV panels with 40% light transmittance could maintain soybean quality without significantly reducing yield. Overall, existing studies have mainly focused on fixed-mounted or vertical AV systems, whereas research on the dynamic changes in the photothermal environment of flexible AV systems and their effects on crop growth remains relatively limited.
This study selected wheat as the experimental crop primarily due to its crucial role in global food security, as well as its strong environmental adaptability and drought tolerance, making it one of the main staple crops widely cultivated within AV systems [15,16]. Wheat growth is highly dependent on light conditions, with photosynthetic efficiency and biomass accumulation varying significantly with changes in light intensity and quality [17]. Consequently, microclimatic factors under PV arrays, such as shading, temperature fluctuations, and redistribution of light and heat, directly influence wheat development and yield formation. Previous studies have systematically investigated wheat growth and yield within AV systems. For example, Prakash et al. [18] found that in fixed AV systems, wheat yield significantly decreased with increasing panel density, with reductions reaching up to 41.9% under full-density conditions compared to unshaded controls. Asa’a et al. [19] reported that reducing panel density using checkerboard or dashed layouts could increase winter wheat yield by up to 31% compared with standard layouts. Yalçın et al. [20] found that wheat could not obtain sufficient illumination under any of the layouts through simulations of three fixed AV configurations in Turkey, indicating that such designs are more suitable for sugar beet than for wheat. However, existing research has primarily focused on fixed-frame PV arrays, and systematic studies on the effects of flexible AV systems on field microclimate and wheat yield remain lacking.
Therefore, this study focused on the flexible AV system in Sihong County, Jiangsu Province, with wheat used as the test crop. A two-year field experiment was conducted to systematically evaluate the effects of the flexible AV system on farmland microclimate, wheat growth, and yield formation. Monitoring points were set up under the PV panels, between the panels, and in open-field control areas to continuously collect microclimate data, including solar radiation, air temperature, and soil temperature, throughout the entire wheat growth period. Based on the field monitoring data, a light-environment simulation model was constructed and validated to characterize the spatial distribution of solar radiation within the flexible AV system. Combined with measurements of chlorophyll content, photosynthetic characteristics, and yield components, the study analyzed the patterns of light and heat resource redistribution under PV shading conditions and their effects on wheat growth and yield formation. The results provide theoretical guidance and data support for structural optimization of flexible AV systems and for crop adaptation.

2. Materials and Methods

2.1. Large-Span Flexible AV System for the Experiment

The large-span flexible AV system used in this study is located at the AV complementary base of Three Gorges New Energy Sihong Co., Ltd. in Suqian City, Jiangsu Province, China (33°16′ N, 118°09′ E). The region has a subtropical monsoon climate, with an annual mean temperature of 14.8 °C, annual precipitation of approximately 900 mm, annual sunshine duration of about 2300 h, and total annual solar radiation reaching 4810.3 MJ/m2.
The AV system consists of 430 W monocrystalline silicon modules installed in a fixed south-facing orientation (Figure 1). The spacing between rows in the north–south direction is 3.25 m, the east–west span is 30 m, and the installation tilt angle is 16°. Each PV panel measures 2115 mm × 1052 mm × 35 mm. The clearance height from the lower edge of the modules to the ground is 4.0 m, and the vertical coverage ratio is 65.0%. This configuration allows the passage and operation of large agricultural machinery, such as combine harvesters. The supporting structure adopts a composite system consisting of Φ300 prefabricated reinforced concrete pipe piles and high-strength cables, thereby providing structural stability and adaptability.

2.2. Experimental Design

2.2.1. Microclimate Monitoring Experiment

This study analyzed the spatiotemporal variation in light and thermal environments inside and outside flexible AV systems during two complete winter wheat growing seasons, from November 2023 to June 2024 and from November 2024 to June 2025. The monitored variables included solar radiation intensity, air temperature, and soil temperature. Based on differences in light distribution within the AV system, the interior of the array was divided into the under-panel planting area (UP) and the between-panel planting area (BP), with three replicated monitoring points established in each area (Figure 2). The monitoring instruments used included HOBO S-LIB-M003 total solar radiation sensors (Onset Computer Corporation, Bourne, MA, USA; accuracy ±5%), HOBO UX100-011A temperature sensors (Onset Computer Corporation, USA; accuracy ±0.2 °C), and HOBO UX120-006M soil temperature sensors (Onset Computer Corporation, USA; temperature accuracy ±0.2 °C). Data were collected at 10 min intervals, covering the entire duration of the two winter wheat growing seasons.

2.2.2. Microclimate Assessment Method

In this study, the classification of weather types comprehensively refers to the composite criteria proposed by Zhang et al. [21]. Days with total cloud cover N ≤ 3 and sunshine duration ≥ 8 h are defined as typical sunny days, while days with total cloud cover N ≥ 8 and sunshine duration ≤ 3 h are classified as typical overcast days. Based on observational data from meteorological stations, and in accordance with the classification standards of the World Meteorological Organization (WMO), weather types during the experimental period were dynamically categorized.
For the definition of typical weather conditions, considering environmental characteristics and the physiological rhythms of wheat, and following agronomic conventions, a day is divided into daytime (06:00–18:00) and nighttime (18:00–06:00 of the following day) [12].

2.2.3. Wheat Planting Experiment

The wheat variety “Lianchuangmai 11” was selected as the experimental material. This experimental material was sown on 29 October 2023, and harvested on 8 June 2024. For the second growing season, wheat was sown on 27 October 2024, and harvested on 6 June 2025. The seeding rate was 300 kg·hm−2. Before sowing, basal fertilizers were applied during land preparation, including diammonium phosphate (containing 46% P2O5 and 8% N) at 190 kg·hm−2 and urea (containing 46% N) at 185 kg·hm−2. Urea (46% N) was top-dressed at 260 kg·hm−2 at the jointing stage. For the entire growth period, the total application rate was equivalent to 90 kg·hm−2 of P2O5 and 240 kg·hm−2 of pure nitrogen. The remaining cultivation practices were managed according to normal field standards.
Based on the variation in solar radiation intensity within the AV system, two treatments were established: the under-panel planting area (UP) and the between-panel planting area (BP). In addition, an open-field control group (CK) was set up outside the system. To ensure the accuracy and reliability of the experimental results, identical cultivation management practices were applied across all areas, including fertilization, irrigation, and pest and disease control. The temperature and precipitation conditions during the wheat growing seasons are shown in Figure 3. The total precipitation during the 2023–2024 growing season was 237.8 mm, and during the 2024–2025 growing season it was 378.6 mm. The average temperature over the two years was 12.9 °C, and both years were classified as drought years.

2.3. Measurement Items

2.3.1. Growth Indicators

At four key growth stages of wheat, namely jointing, heading, grain filling, and maturity, five representative plants were randomly selected from each plot for measurement, with three replicates. Plant height (from the ground surface to the top of the spike) was measured with a ruler, and stem diameter (the diameter of the second internode from the base) was measured using a vernier caliper. Leaf area was calculated using the correction coefficient method, according to the formula: Leaf area = leaf length × leaf width × 0.8. The correction coefficient of 0.8 was used to correct the overestimation of the area of wheat’s elongated leaves by the rectangular method [22].

2.3.2. Chlorophyll Content

At the jointing, heading, and grain-filling stages of wheat, chlorophyll content in the flag leaves was determined using the ethanol extraction method [23]. Five plants were randomly selected from each plot, and their flag leaves were collected for measurement, with three biological replicates.

2.3.3. Photosynthetic Characteristics

At the jointing, heading, and grain-filling stages of wheat, the photosynthetic parameters of flag leaves were measured using a LI-COR 6800 portable photosynthesis system (LI-COR Biosciences, Lincoln, NE, USA) [24]. Measurements were conducted under standard conditions: a photosynthetic photon flux density of 1000 μmol·m−2·s−1, a CO2 concentration of 400 μmol·mol−1, a leaf chamber temperature of 25 °C, and a relative humidity of 50%. The measured parameters included net photosynthetic rate (Pn), stomatal conductance (Gs), intercellular CO2 concentration (Ci), and transpiration rate (Tr). Five functional leaves were measured in each plot, with three replicates.

2.3.4. Crop Yield

After wheat reached maturity, a 2 m2 quadrat (2 m × 1 m) was harvested from an unsampled area in each plot. The samples were air-dried, weighed, and used to calculate biomass per unit area [25]. After threshing, grain yield was uniformly converted to a per-hectare basis at a standard moisture content of 14%. Twenty ears were randomly collected from each plot for indoor seed testing, and the number of grains per ear, the quality per thousand grains and the seed setting rate were investigated.

2.4. Statistical Analysis

Microsoft Excel was used for data organization, and statistical analyses were performed using SPSS 26.0 software. Differences among treatments were analyzed by one-way analysis of variance (ANOVA), and Duncan’s multiple range test was used for multiple comparisons at the significance level of p < 0.05. Figures were prepared using Origin 2023b software. In addition, model validation metrics, including root mean square error (RMSE), mean absolute error (MAE), and regression slope, were calculated to quantitatively assess the agreement between simulated and observed values.

3. Construction of the Light Environment Model for AV Systems

3.1. Model Construction Model Validation

To improve computational efficiency and avoid unnecessary model complexity, several secondary structural and environmental elements were not incorporated into the AV simulation model [26]. Specifically, the diagonal braces of the AV systems were not considered, and auxiliary facilities associated with PV electricity generation, such as combiner boxes, were also neglected. In addition, heterogeneity in field topography and the presence of surface vegetation were not included in the simulation [21]. A full-scale three-dimensional representation of the AV system was first developed using 3D modeling software and subsequently imported into ECOTECT for simulation analysis (Figure 4). Meteorological inputs for the simulation were obtained from the nearest available long-term weather datasets from Suqian, Jiangsu Province, which is the closest major city to Sihong, a county under the administration of Suqian. These datasets include more than 10 climatic variables based on records spanning the past 30 years [27]. To minimize edge effects and reduce the interference of lateral light and other unstable factors on the simulation outputs, the model configuration was optimized through repeated testing, and a seven-span PV array was ultimately adopted. The computational domain was defined within the central section of the array to ensure greater stability and representativeness of the simulated light environment. For numerical analysis, the target simulation plane was discretized into grids with an appropriate spatial extent and sufficient resolution so that the light environment parameters could be captured with adequate accuracy and detail.

3.2. Model Validation

To assess the applicability and reliability of the light environment simulation model for the AV system, the model outputs were validated against field measurements across the wheat growth period. Multiple monitoring points were set up beneath and between PV panels, as well as in the open field outside the AV site. Continuous observations with high-precision solar radiation sensors were conducted to obtain the monthly average daylighting rates during the wheat growth stages.
Simulated monthly average daylighting rates generated by the AV light environment model were compared with the measured data. The comparison showed that good agreement with measured values, with an RMSE of 1.03, an MAE of 0.84, a regression slope of 1.024, a relative error of 4.4%, and an R2 of 0.950, indicating that the light-environment model reasonably reproduced the radiation distribution within the flexible AV system (Figure 5). These results indicate that the model can accurately reproduce the average light conditions for each month of the wheat growth period, confirming that it is reliable and directly applicable for subsequent studies on the spatial heterogeneity of light environments in large-scale flexible AV systems.
Agronomy 16 01111 g004
Figure 4. Light environment model for AV systems.
Figure 4. Light environment model for AV systems.
Agronomy 16 01111 g004
Agronomy 16 01111 g005
Figure 5. Comparison of simulated and measured monthly average daylighting rates over two years of wheat growth.
Figure 5. Comparison of simulated and measured monthly average daylighting rates over two years of wheat growth.
Agronomy 16 01111 g005

4. Results

4.1. Variation in Solar Radiation Intensity Within the Flexible AV System

4.1.1. Monthly Average Solar Radiation Variation During the Production Period

Analysis of monthly average solar radiation monitoring data covering the full wheat growth cycle from November 2023 to June 2025 revealed distinct spatiotemporal variations in solar radiation intensity both inside and outside flexible AV systems (Figure 6). The inter-monthly fluctuations in radiation intensity are closely linked to the structural characteristics of the AV system. During winter at low solar elevation angles, the increased projected area of the panels reduces the disparity in radiation reception between the BP and UP areas. Conversely, as solar elevation angles rise in spring, the enhanced light-transmitting window effect between panels allows greater exposure to direct and diffuse radiation in the BP area.
The monthly average solar radiation intensity in the BP area was 56.9% lower than that in the CK treatment, while that in the UP area was 62.0% lower. The solar radiation intensity in the CK control consistently remains significantly higher than within the AV system, with monthly variations ranging from 152.0 to 463.1 W/m2 and peaking in May—a pattern consistent with the solar altitude angle variation patterns in the Northern Hemisphere.
The daylighting rate exhibited a distinct seasonal variation pattern during the observation period (Figure 7). Both the UP and BP positions reached their peak values in April, remaining at lower levels throughout the winter months. Overall, the light transmittance curve at the BP position consistently exceeded that at the UP position, indicating superior lighting conditions. Calculations revealed that the average total daylighting rate was 43.1% for the BP position and 38.0% for the UP position, with an absolute difference of 5.9%. This quantitatively demonstrates the heterogeneous characteristics of the light environment beneath the PV panel array.
The radiation intensity difference between the areas beneath and between panels within a PV array varies dynamically throughout the growth cycle. During the winter period, the difference is minimal, averaging 10.2%; from the regreening stage through the heading stage, it increases significantly, ranging from 11.3% to 26.8%, with the largest variation occurring in April; by the maturity stage in June, the difference narrows to 13.7%. This variation pattern is closely related to the structural characteristics of the AV system. Furthermore, the data reveal an important phenomenon: during the maturity stage, although the solar altitude angle continues to increase, the radiation intensity decreases across all regions due to increased cloud cover during the rainy season. This finding underscores the necessity to comprehensively consider the dual impacts of astronomical and meteorological factors when evaluating the performance of AV systems.

4.1.2. Typical Weather and Solar Radiation Variations

Based on the experimental data in Figure 8, this study further analyzes the spatiotemporal distribution characteristics of solar radiation intensity in the UP and BP areas of flexible AV systems, as well as in the CK control group, under typical cloudy and sunny conditions during winter (1 and 16 February 2024) and summer (1 and 12 June 2024).
Typical clear-day monitoring data reveal distinct alternating variations in solar radiation intensity between the UP and BP areas. This phenomenon stems from the specialized installation methods and structural design of flexible PV panels, which produce varying levels of shading and reflection effects on solar radiation at different times. The effect is particularly pronounced in winter, with solar radiation values in these areas reaching up to 501.9 W/m2 and 571.9 W/m2, respectively, equivalent to 84.0% and 92.0% of the CK reference values at that time. This indicates substantial solar radiation intensity in both UP and BP areas during overcast winter conditions. However, the shorter sunlight duration in the sub-area results in an average solar radiation level 76.2% lower than that in the BP area.
Typical overcast weather data indicate that cloud scattering alters the distribution pattern of solar radiation, resulting in more consistent numerical variations in BP and UP, with this effect being more pronounced in winter than in summer. Overall, solar radiation exhibits a gradient distribution: CK > BP > UP. In winter, the mean radiation values for BP and UP reach 44% and 40% of that for CK, respectively, while in spring they decrease to 28% and 24%. This demonstrates that under overcast winter conditions, the shading effect of flexible PV systems is reduced compared to summer. However, during this period, radiation values in BP and UP from 12:00 to 13:00 range only from 20.6~44.4 W/m2 and 31.9~49.4 W/m2, respectively, representing significant reductions of 79.6–89.2% and 47.7–86.5%, respectively, compared with CK during the same time frame in the harvest period.

4.2. Thermal Environment Changes Within Flexible AV Array Structures

The coverage of the flexible AV system alters the distribution of solar radiation and the balance of ground heat flux, creating a soil thermal environment within the system that differs markedly from that of the CK environment. The installation method of PV panels significantly influences the characteristics of soil water and heat transport. Soil temperatures in the areas beneath and between the panels exhibit regular diurnal and seasonal variations, which directly affect the intensity of wheat root activity and its nutrient absorption efficiency, thereby influencing plant growth and development.
During the observation period, the air accumulated temperature trends across the three regions exhibited high consistency, with inter-monthly fluctuations being the predominant characteristic. All regions entered a low-temperature accumulation phase starting in November, reaching peak values simultaneously in April of the following year, followed by a decline between May and June. Monthly variation patterns indicate that PV panel installation did not alter the fundamental seasonal patterns of accumulated temperature, with macroclimatic factors remaining dominant. Subtle spatial differences were observed: winter accumulated temperatures under panels were generally slightly higher than in other regions, suggesting a mild insulation effect; however, during the peak growing season in April, outdoor accumulated temperatures reached their highest levels. Data on monthly average air and soil temperatures throughout the wheat growth cycle (November 2023 to June 2025; Figure 9 and Figure 10) demonstrate that flexible AV systems exhibited significant insulation effects during the winter dormancy period. In November, the accumulated temperature in the BP area reached 350.3 °C, which was higher than the 341.7 °C recorded in UP but lower than the 367.6 °C recorded in CK. A similar pattern was observed in March, when CK showed the highest accumulated temperature at 333.6 °C, followed by BP at 315.7 °C and UP at 314.3 °C. The consistently higher accumulated temperature in CK may be attributed to the greater exposure of the open field to direct solar radiation, particularly as solar elevation angles increased in spring. From April to June, the accumulated temperature in the open-field area remained the highest. However, in the BP area, optimized light transmittance gradually brought the accumulated temperature close to that of the open field, while the UP treatment maintained consistently lower temperatures. This prevented high temperatures from inhibiting root respiration, providing a stable environment for grain filling.
As shown in Figure 10, accumulated soil temperature varied markedly among months during the wheat growth cycle. In the AV field, the accumulated soil temperature was generally higher than the accumulated air temperature, reflecting the combined effects of soil heat storage capacity and the altered microclimate under PV panels. Although PV panels reduced part of the incoming solar radiation, soil retained heat more effectively because of its greater heat capacity and slower heat release. In contrast, air temperature was more sensitive to short-term changes in radiation, shading, and ventilation within the AV field, resulting in greater fluctuations and lower accumulated temperature. Moreover, soil serves as a direct receiver and converter of solar radiation, whereas air is primarily heated indirectly through long-wave radiation and convection at the surface—a less efficient passive heating mechanism resulting in weaker heat accumulation capacity compared to soil. PV panel coverage altered this relationship by reducing the solar radiation received by the soil beneath the panels, thereby weakening its accumulated-temperature advantage and, in some months, resulting in lower accumulated soil temperature than that observed in the open field. This finding indirectly confirms that solar radiation remains the fundamental driving force behind soil heat accumulation. In June, the accumulated soil temperature was relatively low mainly because wheat was harvested in early June, which shortened the effective observation period for that month.

4.3. Light Environment Simulation Analysis

4.3.1. Shading Variation Characteristics of the Flexible AV System

The shadow variation characteristics of farmland within the large-span flexible AV system are shown in Figure 11. The light environment in farmland within flexible AV systems exhibits pronounced spatiotemporal heterogeneity. The light distribution clearly varies with both seasonal and daily patterns. At the winter solstice, the solar altitude angle is at its lowest, and the shadows cast by the PV panels are long and narrow, resulting in the largest rectangular shaded areas beneath the arrays and the shortest duration of sunlight. Consequently, farmland light resources are severely limited for most of the day. By the vernal equinox, as the solar altitude increases, the shadow range shortens and widens, the shaded area decreases, and the light distribution becomes more uniform. At the summer solstice, the solar altitude reaches its maximum, shadows are shortest, and nearly all areas beneath the arrays receive ample sunlight throughout the day, with only a very small region directly beneath the panels remaining shaded continuously.
Regarding daily variation, light distribution also differs significantly at different times within the same season. Taking the vernal equinox as an example, in the morning and afternoon, the sun’s azimuth is biased toward the east and west, respectively, causing the shaded area to shift westward or eastward, so the illuminated regions tend to be biased toward one side of the array. Around noon, when the sun is due south, the shaded area is most regular and symmetrical, and the solar altitude is at its highest, minimizing the shaded area and providing the optimal sunlight conditions of the day.
This spatiotemporal pattern, characterized by the strongest shading in winter, the most abundant light in summer, relatively favorable light conditions at noon, and suboptimal conditions in the morning and evening, is mainly determined by the geometric relationship between the apparent solar trajectory and the fixed-tilt PV array structure. Studies indicate that farmland under PV arrays is not in a continuously low-light environment; rather, it experiences a dynamic light-shadow field with regular gradients and patchiness due to the movement of the sun. This provides important spatial and temporal guidance for crop layout and light utilization in AV systems.

4.3.2. Variation Characteristics of Farmland Light Availability Within Large-Span Flexible AV Arrays

Figure 12 illustrates the spatial distribution of average daylighting rate in the flexible AV system. The comparison covers two representative periods, December–February and March–May. From a temporal perspective, the light availability within the system exhibits pronounced seasonal fluctuations. During the winter, when solar resources are relatively scarce, the light availability beneath most of the PV modules is generally below 30%, indicating that the shading effect of the PV panels is significant at low solar elevation angles, and the photosynthetically active radiation reaching the canopy is markedly reduced. As spring arrives and the solar elevation angle increases, the overall light environment within the system improves, with light availability rising to 40–60%. By early summer, although overall light availability remains relatively high, the central areas beneath some panels show a slight decline compared to spring, possibly due to secondary interception of light caused by the PV tilt angle and canopy development.
In terms of spatial distribution, the light resources within the system display a distinct gradient pattern. Directly beneath the PV panels, especially near the supporting structures, there is a perennial shadow core with the lowest light availability. In contrast, the pathway areas between PV panels and the system edges consistently maintain higher light levels, exceeding 60% during the growing season, resulting in a pronounced difference in the light environment between the areas under the panels and the pathways. This spatial differentiation arises from the interaction between the geometric structure of the PV array and solar radiation. Moving from the center of a single PV panel toward its edges, light availability generally shows a gradual increase. This analysis indicates that the flexible AV system does not create a uniform light environment but rather establishes a dynamically changing gradient light field across seasons and locations. This provides an important spatiotemporal basis for crop arrangement and efficient utilization of solar energy within the system.

4.4. Wheat Yield and Biomass

Figure 13 presents the dry and fresh weight as well as yield indicators for wheat from 2023 to 2025. Overall, wheat biomass accumulation increased throughout both years as growth progressed, reaching peak levels during the heading stage before slightly declining or stabilizing during the grain filling stage. For fresh weight, all treatments reached their maximum values in 2024, with the CK treatment significantly higher than others, while the UP and BP treatments showed sequential declines with statistically significant differences. A similar pattern emerged in 2025, though the difference between the UP and CK narrowed at certain stages, suggesting that climatic or management conditions may modulate treatment effects across years. Biomass accumulation peaked during the heading stage in both years, followed by a slight decline during grain filling. In 2024, the CK treatment significantly outperformed both the UP and BP treatments during the heading stage, whereas the differences between the UP and BP treatments were either insignificant or only significant at specific stages. In 2025, treatment differentiation became more pronounced: both the UP and BP treatments showed significantly lower values than the CK treatment during both the heading and grain-filling stages, although the gap narrowed, indicating converging treatment effects later in the growth cycle. Additionally, biomass accumulation during the jointing stage remained generally low in both years, with minor but statistically significant differences among treatments. Year-by-year comparisons revealed that fresh weight and biomass accumulation in 2025 were slightly lower than in 2024, particularly during heading. The consistent pattern across treatments remained evident: CK consistently exhibited the highest biomass accumulation capacity, while the UP and BP treatments partially inhibited wheat biomass formation.
The two-year yield results demonstrate that different treatments significantly influence wheat yield formation. In 2024, the CK treatment achieved the highest regional yield, reaching 6717.2 kg·hm−2 and markedly exceeding the yields of the BP and UP treatments. The yields of the BP and UP treatments were 3801.0 and 3547.0 kg·hm−2, respectively, with no statistically significant difference between them. However, these values represented decreases of 43.4% and 47.2%, respectively, compared with the CK treatment. In actual yields, the flexible AV system treatment yielded 3679.2 kg·hm−2, significantly lower than CK with a reduction of 45.2%. A similar pattern emerged in 2025: CK regional yield reached 6393.4 kg·hm−2, still significantly higher than BP and UP treatments; BP and UP yields were 3720.3 kg·hm−2 and 3551.7 kg·hm−2 respectively, showing decreases of 41.8% and 44.6% compared to CK. The flexible AV system treatment yielded 3646.6 kg·hm−2, a 43.0% reduction from CK with a statistically significant difference. Overall, the CK treatment maintained the highest yield levels, while the PV-related treatments significantly reduced wheat yields, although the difference between the BP and UP treatments was not significant. This suggests that PV shading or spatial configuration may alter yield formation by modifying crop light exposure conditions. The consistent yield trends across treatments over the two years indicate relative stability in these effects.
To further clarify the mechanism underlying wheat yield reduction under PV shading, three core yield components were analyzed, including the spike number per unit area, grain number per spike, and thousand-grain weight (Table 1). Across the two growing seasons, the CK treatment consistently showed the highest spike number, whereas both the BP and UP treatments showed significantly lower spike numbers. In 2024, the spike number decreased from 453.7 spikes m−2 in CK to 250.1 and 254.7 spikes m−2 in the BP and UP treatments, respectively, corresponding to reductions of 44.9% and 43.9%. In 2025, similar reductions were observed, with the spike number decreasing by 42.0% and 43.1% in the BP and UP treatments, respectively, compared with the CK treatment. These results indicate that the reduction in the spike number per unit area was the primary factor contributing to yield loss under PV shading.
Grain number per spike was also reduced in the AV-related treatments. Compared with the CK treatment, grain number per spike decreased by 9.9% and 16.9% in the BP and UP treatments in 2024, and by 11.9% and 15.0% in 2025, respectively. This suggests that shading not only affected population establishment but also limited spike fertility and grain formation. In contrast, the thousand-grain weight increased in both the BP and UP treatments over the two years. Compared with CK, the thousand-grain weight increased by 14.0% and 13.0% in BP and UP in 2024, and by 14.0% and 14.9% in 2025, respectively. This increase indicates a compensatory response in grain filling under lower spike density and reduced grain number. However, the increase in the thousand-grain weight was insufficient to offset the substantial reductions in the spike number and grain number per spike. Therefore, the yield reduction within the flexible AV system was mainly attributed to decreased the spike number per unit area, followed by reduced the grain number per spike, whereas increased the thousand-grain weight played only a limited compensatory role.

4.5. Wheat Growth

Figure 14 presents the growth indicators of wheat from 2023 to 2025. Overall, the two-year data show that wheat plant traits increased progressively with advancing growth stages under all treatments, though significant differences emerged between treatments and interannual variations were observed. In mid-2024, the UP treatment demonstrated clear advantages across most metrics, particularly during the grain filling stage. Leaf area increased by 24.2% compared to CK and 11.4% compared to BP, both differences reaching statistical significance. Stem diameter in the UP area was 6.5% higher than that in the BP area during the grain-filling stage, but it was not significantly different from that in CK. This indicates that the UP area provided a slight benefit in enhancing mechanical strength during the later growth stages, although this advantage was limited.
The 2025 results were generally consistent with the 2024 trends, but some indicators demonstrated stronger response to treatment. Regarding plant height, the CK treatment showed higher values during the grain-filling stage compared to UP and BP, with CK exceeding BP by 11.8%, indicating that UP and BP may have an inhibitory effect on plant height regulation that year. However, in functional growth indicators, UP still performed optimally. At the grain-filling stage, UP exhibited a 23.1% increase in leaf area relative to CK and a 43.2% increase relative to BP, both significantly outperforming the other treatments. Additionally, at the heading stage in 2025, UP increased stem diameter by 33.3% compared to CK, markedly enhancing early stem growth, suggesting that the UP treatment had a more pronounced promoting effect during the vegetative growth phase. A comparison of the two-year results shows that the UP treatment demonstrates consistent advantages in enhancing leaf area and spike traits, with these benefits being particularly pronounced during the grain filling stage, indicating its sustained promotion of photosynthetic product accumulation and grain formation. In contrast, the BP treatment exhibited performance intermediate between the CK and UP treatments, with some indicators even lower than those of CK, suggesting relatively unstable effects. Overall, the UP treatment exhibited more significant and stable promoting effects on wheat growth and development, showing distinct advantages in improving population photosynthetic capacity and reproductive growth metrics.

4.6. Photosynthetic Parameters of Wheat

4.6.1. Photosynthetic Characteristics

Figure 15 presents the photosynthetic indicators of wheat from 2023 to 2025. The results of two years of trials revealed that wheat photosynthetic parameters varied distinctly across different growth stages and showed significant differences among treatments. Overall, Pn, Tr, and Gs were highest during the heading stage under all treatments, while Ci remained relatively stable but exhibited notable differences between treatments.
In mid-2024, the CK treatment consistently showed higher Pn levels across all growth stages, especially during the heading stage, where it increased by 38.9% and 25.0% compared to UP and BP, respectively, demonstrating significant differences. During the grain filling stage, Pn in CK remained 27.5% and 18.6% higher than UP and BP. The trends for Tr and Gs closely followed that of Pn, with CK showing increases of 13.3% and 29.3% compared to UP during the heading stage, indicating a stronger gas exchange capacity under CK treatment. In contrast, Ci was significantly higher in the CK treatment than in the UP and BP treatments. For example, during the grain-filling stage, Ci in the CK treatment was 24.1% and 20.7% higher than that in the UP and BP areas, respectively. Given the simultaneous decreases in Pn and Gs in the UP and BP areas, the lower Ci should not be interpreted as an improvement in CO2 utilization efficiency. According to the photosynthetic biochemical model and the Pn–Ci interpretation framework, this pattern more likely indicates that reduced stomatal conductance limited CO2 diffusion into the leaf, thereby contributing to the decline in photosynthetic rate under PV shading [28]. In addition, prolonged low-light conditions may also have induced non-stomatal biochemical limitations, such as reduced carboxylation capacity or electron transport activity. Therefore, the reduced photosynthesis observed in the UP and BP areas was likely associated with both stomatal and potential non-stomatal limitations, with stomatal limitation playing an important role in restricting CO2 supply [29].
The 2025 results largely followed the pattern observed in 2024, with adjusted differences among treatments. For Pn, CK still showed the highest value at the heading stage, exceeding UP and BP by 33.3% and 30.4%, respectively. However, at the tillering stage, BP slightly outperformed CK, indicating BP’s positive effect on early photosynthetic capacity. For Tr, BP during the grain filling stage in 2025 was significantly higher than CK and UP by 21.0% and 43.2%, respectively, demonstrating its advantage in regulating transpiration later in the growth cycle. The trend for Gs mirrored that of Pn, with CK exhibiting a 29.3% increase over UP and a 10.2% increase over BP at the heading stage, suggesting that CK maintained higher stomatal opening. For Ci, CK consistently maintained the highest level, such as a 50.0% increase over BP at the heading stage, further supporting its lower CO2 utilization efficiency.

4.6.2. Chlorophyll Index

Figure 16 presents the changes in wheat chlorophyll content from 2023 to 2025. The two-year experimental results demonstrated that the total chlorophyll content in wheat exhibited a trend of initially increasing followed by a slight decrease during its growth cycle, reaching its peak at the heading stage with significant differences among treatments. In mid-2024, the UP and BP treatments showed significantly higher chlorophyll levels than the CK treatment: UP increased by 32.3% and BP by 30.3% compared to CK, although the difference between UP and BP was not statistically significant, indicating that both treatments markedly promoted chlorophyll accumulation. By the grain-filling stage, differences among treatments narrowed; the differences between UP and both CK and BP were no longer significant, yet overall chlorophyll levels remained high, suggesting stabilization during the later stages of reproductive growth.
The 2025 results were largely consistent with the 2024 trends, but the differences among treatments became more pronounced. During the heading stage, UP and BP still demonstrated the highest performance: UP increased by 20.6% compared to CK, and BP increased by 23.5% compared to CK, both significantly higher than CK. Notably, during the grain filling stage, the CK treatment outperformed UP and BP, with CK showing a 10.7% increase over UP and an 18.6% increase over BP—differences that reached statistical significance—indicating a decline in the chlorophyll retention capacity of UP and BP treatments during the later growth phase. Prior to the heading stage, UP improved by 12.5% compared to CK, while BP was slightly lower by 7.1%, suggesting that UP exerted a more stable promoting effect on early chlorophyll synthesis.
Agronomy 16 01111 g016
Figure 16. Changes in chlorophyll content during the main growth period of winter wheat. (a) From 2023 to 2024. (b) From 2024 to 2025. Different lowercase letters indicate significant differences in the indexes under different treatments (p < 0.05).
Figure 16. Changes in chlorophyll content during the main growth period of winter wheat. (a) From 2023 to 2024. (b) From 2024 to 2025. Different lowercase letters indicate significant differences in the indexes under different treatments (p < 0.05).
Agronomy 16 01111 g016
Based on two-year data, the UP treatment significantly increased chlorophyll content during the heading stage, demonstrating stable and substantial advantages, making this period critical for enhancing population photosynthetic capacity; meanwhile, the BP treatment performed well during heading but showed a marked decline during the 2025 grain-filling stage. Overall, treatment effects on chlorophyll regulation were primarily concentrated during the heading stage, with differences weakening or even reversing in later stages, indicating that different treatments exhibit distinct temporal patterns in regulating chlorophyll dynamics.

5. Discussion

5.1. Impact of Flexible AV System Shading on Microclimate

The core challenge currently faced in the field of agrivoltaics is how to optimize the distribution of solar thermal resources to achieve a win–win situation [30]. This study analyzes the thermal and light gradient characteristics within flexible AV systems, revealing the mechanisms by which shading regulates the light environment. Zhang et al. [21] also found in their research in East China that AV systems form significant light environment gradient distributions. The total monthly average solar radiation intensity beneath and between flexible PV panels is reduced by 62.0% and 56.9%, respectively, compared to that of open-field conditions. Under sunny conditions, the solar radiation beneath and between the panels alternates, a phenomenon that is more pronounced in winter. The solar radiation intensity in these areas can reach up to 84.0% and 92.0% of the open-field value, respectively. This alternating pattern leads to a more homogeneous light environment inside the array [31], reducing the radiation difference between the areas beneath and between the panels. This finding is in line with Zhang et al.’s [12] research on the impact of PV panel height on the light environment. Under sunny conditions, the overall trend of solar radiation intensity beneath and between the panels is similar to that of open-field conditions, with the difference being more evident in winter. This convergence in light resource utilization and its impact on crop growth provides an important basis for managing light resources in AV systems under varying weather conditions.
Flexible PV covers significantly alter the thermal environment of the soil, displaying notable spatial and temporal differentiation characteristics [8]. During the winter, the insulation effect of the AV system increases the accumulated temperature in the areas beneath the panels by 12.3% to 23.8% compared to that of open-field conditions, effectively mitigating low-temperature stress. However, during the growing season, the accumulated temperature beneath the panels is reduced by 5.2% to 8.7% compared to that of open-field conditions, preventing high temperatures from damaging the root system. This effect arises from the redistribution of solar radiation by the PV modules [32]: reducing ground heat loss in winter and lowering soil heat absorption through shading in summer. It is particularly noteworthy that the area between the panels, due to optimized light transmittance, maintains an accumulated temperature that lies between the values of the area beneath the panels and the open-field conditions, creating a buffered thermal environment. This may provide more stable growth conditions for crops [5]. However, the study also found that current thermal environment data lacks synchronized monitoring of air temperature and humidity, and future work should improve the analysis of multidimensional microclimate parameters.

5.2. Impact of Flexible AV System Shading on Wheat Yield

This study demonstrates that shading from flexible AV systems significantly reduces wheat yield. In 2024, the yield in the UP and BP areas decreased by 47.2% and 43.4%, respectively, compared to the CK control. In 2025, the reductions were 44.6% and 41.8%, with similar trends in both years, indicating a stable inhibitory effect of the flexible AV system on wheat yield. These results align with research on fixed AV systems. Prakash et al. [18] found that in fixed AV systems with different densities, the PV panels significantly reduced the supply of photosynthetically active radiation to the crop canopy, and the shading increased with higher panel density, thus limiting wheat productivity. This suggests that regardless of whether the system is fixed or flexible, the reduction in light resources remains the core factor limiting wheat yield. A systematic review also pointed out that the impact of AV systems on crop yield is jointly influenced by shading intensity, crop type, and regional climate. Insufficient light typically reduces grain crop yields, but proper system design can mitigate the risk of yield loss [33,34].
Compared with traditional fixed-frame AV systems, the flexible AV system used in this study has higher clearance and a larger span, which may improve ventilation and mechanical operation beneath the panels. However, the relatively high panel coverage of the flexible AV system still substantially reduced the light available to the wheat canopy. In the UP treatment, wheat plants showed increased leaf area and chlorophyll content under the flexible AV arrays, which should be interpreted as an acclimation response to reduced light rather than as direct evidence of improved productivity. Similar shade-induced adjustments in wheat morphology, chlorophyll content, and photosynthetic traits have been reported previously [35,36,37].
Although increased leaf area and chlorophyll content under the flexible AV system may enhance light interception and absorption at the leaf level, these responses were insufficient to offset the reduction in total radiation available to the canopy. Under the persistent and spatially heterogeneous shading created by the flexible AV arrays, canopy-scale photosynthetic carbon assimilation was likely constrained. Moreover, maintaining larger leaves under low-light conditions may consume part of the limited assimilates, increasing the carbon cost of vegetative growth and leaving less carbon for spike development and grain filling. This mechanism explains why increased leaf area and chlorophyll content in the UP treatment did not lead to higher yield. Wheat yield formation depends on both assimilate production by leaves and assimilate utilization by spikes and grains. When light availability is reduced during reproductive development, carbon supply to developing spikes and grains can become limited, leading to fewer effective spikes, fewer grains per spike, weaker grain filling, and lower grain yield [36,38]. Therefore, the increased leaf area and chlorophyll content under the flexible AV system represented a compensatory response to shading, whereas the yield reduction indicated that this compensation was insufficient to overcome canopy- and whole-plant-level limitations in carbon assimilation and allocation.
Notably, the yield differences between the areas under the panels and between the panels were not significant, indicating that the flexible array may have reduced the spatial heterogeneity commonly observed in fixed AV systems through a more even redistribution of light and heat. Yalçın et al. [20] also emphasized that module arrangement and light distribution optimization are key factors for maintaining crop yield in AV systems. Sturchio et al. [39] further pointed out that the sustainable development of AV systems requires coordinated optimization among energy capture, and crop production. Therefore, although wheat yield under the flexible AV system was lower than that of the CK, the similar yield reductions observed in the UP and BP areas suggest that the flexible AV system produced a relatively uniform crop response within the shaded field. Future efforts should focus on reducing panel coverage, optimizing panel spacing and light distribution, and selecting shade-tolerant wheat varieties to improve the synergy between PV power generation and wheat production.

5.3. Limitations and Future Research Directions

Although this study provides two-year field evidence on the effects of a large-span flexible AV system on farmland microclimate and wheat productivity, several limitations should be acknowledged. In modeling the light environment of the flexible AV system, necessary simplifications were adopted to improve computational efficiency, and some secondary structural components, such as diagonal braces, cables, and auxiliary photovoltaic facilities, were not explicitly represented. Field heterogeneity in topography, surface vegetation, and wheat canopy development beneath the flexible AV arrays was also simplified. Therefore, although the model was validated using field-measured solar radiation data, the simulated results mainly represent the average spatial pattern of light availability within the flexible AV system rather than all fine-scale variations in the crop canopy light environment. In addition, the field experiment was conducted at a single flexible AV site in Sihong County, Jiangsu Province, using one large-span flexible AV structure and one wheat variety. Therefore, the findings are more applicable to flexible AV systems with similar climatic conditions, soil backgrounds, crop management practices, and AV array structures. Moreover, the two-year observations under the flexible AV system are still insufficient to fully represent long-term climatic variability, especially because both experimental years were characterized by drought conditions. Therefore, caution is needed when extrapolating the present results to other ecological regions or long-term production scenarios.
Future research on flexible AV systems should include multi-site and multi-year field experiments to verify the applicability of the present findings under different climatic and agronomic conditions. Further studies should compare different flexible AV array designs, such as panel coverage ratio, row spacing, installation height, span, and tilt arrangement, to identify more suitable structural parameters for crop production beneath flexible AV systems. The light-environment model of flexible AV systems should also be improved by incorporating more detailed structural features, field-surface heterogeneity, and crop canopy development. In addition, long-term observations combined with crop physiological measurements and model optimization would help clarify the interaction between flexible PV shading and crop growth, thereby providing more practical guidance for the design and deployment of flexible AV systems.

6. Conclusions

This study focused on a flexible AV array in Sihong County, Jiangsu Province. Through continuous monitoring and analysis of changes in the internal light and thermal environment, wheat physiological and growth responses, and yield, the effects of the flexible AV system on field microclimate and wheat production were systematically evaluated. The main conclusions are as follows:
(1)
Throughout the wheat growing season, the monthly average solar radiation intensity in the BP and UP areas was 56.9% and 62.0% lower than that in the open field, respectively. Meanwhile, the array showed a certain regulatory effect on the thermal environment, characterized by heat preservation during the overwintering period and cooling during the later growth stage, thereby creating a buffered microclimatic environment distinct from the open field.
(2)
Compared with the open-field control, net photosynthetic rate, stomatal conductance, and related parameters of wheat in the BP and UP areas decreased overall, indicating that shading weakened leaf carbon assimilation capacity. However, wheat grown within the array showed adaptive responses during some growth stages, such as increased leaf area and higher chlorophyll content, suggesting that wheat could alleviate low-light stress through certain morphological and physiological adjustments.
(3)
Wheat yield under the flexible AV system was significantly lower than that of the open-field control, and the trends were consistent over the two years. In 2024, yields in the BP and UP areas were 43.4% and 47.2% lower than those in the open field, respectively; in 2025, the reductions were 41.8% and 44.6%, respectively. These results indicate that under conditions of relatively high panel coverage, light reduction remains the key factor limiting wheat yield formation. Overall, the flexible AV system has structural advantages in balancing power generation and agricultural machinery operation, but further improvements in the synergistic use of agriculture and photovoltaics will require optimization of coverage ratio and panel spacing, as well as the selection of shade-tolerant crop varieties.

Author Contributions

Conceptualization, E.B., and Y.Y.; methodology, L.Z. software, L.Z., J.T., and X.G.; validation, M.Y., and Z.Q.; formal analysis, X.G.; investigation, Y.Y.; resources, E.B.; data curation, M.Y., L.Z., and X.G.; writing—original draft preparation, Y.Y., and L.Z.; writing—review and editing, E.B.; visualization, X.G.; supervision, E.B.; project administration, E.B., and Y.Y.; funding acquisition, E.B., and Y.Y. All authors have read and agreed to the published version of the manuscript.

Funding

This work was supported by the S&T Program of Xingtai (Grant No. 2024ZZ045) and the Jiangsu Provincial Key Research and Development Program (Grant No. BE2023304).

Data Availability Statement

The data presented in this study are available on request from the corresponding author. The data are not publicly available due to also forming part of an ongoing study.

Acknowledgments

The authors are grateful to the handling editor and the anonymous reviewers for helping improve the manuscript.

Conflicts of Interest

The authors declare no competing interests. The authors have approved and agreed to the manuscript submission. All authors declare that they do not have any financial or personal relationships that might be perceived as influencing their work or this work.

References

  1. Stid, J.T.; Shukla, S.; Kendall, A.D.; Anctil, A.; Hyndman, D.W.; Rapp, J.; Anex, R.P. Impacts of agrisolar co-location on the food–energy–water nexus and economic security. Nat. Sustain. 2025, 8, 702–713. [Google Scholar] [CrossRef]
  2. Mazzeo, D.; Di Zio, A.; Pesenti, C.; Leva, S. Optimizing agrivoltaic systems: A comprehensive analysis of design, crop productivity and energy performance in open-field configurations. Appl. Energy 2025, 390, 125750. [Google Scholar] [CrossRef]
  3. Asa’a, S.; Reher, T.; Rongé, J.; Diels, J.; Poortmans, J.; Radhakrishnan, H.S.; van der Heide, A.; Van de Poel, B.; Daenen, M. A multidisciplinary view on agrivoltaics: Future of energy and agriculture. Renew. Sustain. Energy Rev. 2024, 200, 114515. [Google Scholar] [CrossRef]
  4. Kim, S.; Kim, S.; An, K. An integrated multi-modeling framework to estimate potential rice and energy production under an agrivoltaic system. Comput. Electron. Agric. 2023, 213, 108157. [Google Scholar] [CrossRef]
  5. Lee, H.J.; Park, H.H.; Kim, Y.O.; Kuk, Y.I. Crop cultivation underneath agro-photovoltaic systems and its effects on crop growth, yield, and photosynthetic efficiency. Agronomy 2022, 12, 1842. [Google Scholar] [CrossRef]
  6. Weselek, A.; Bauerle, A.; Hartung, J.; Zikeli, S.; Lewandowski, I.; Högy, P. Agrivoltaic system impacts on microclimate and yield of different crops within an organic crop rotation in a temperate climate. Agron. Sustain. Dev. 2021, 41, 59. [Google Scholar] [CrossRef]
  7. Willockx, B.; Reher, T.; Lavaert, C.; Herteleer, B.; Van de Poel, B.; Cappelle, J. Design and evaluation of an agrivoltaic system for a pear orchard. Appl. Energy 2024, 353, 122166. [Google Scholar] [CrossRef]
  8. Tan, Y.; Liu, J.; Li, W.; Yin, J.; Chen, H.; Peng, Y.; Tan, J.; Wei, M. Agrivoltaics development progresses: From the perspective of photovoltaic impact on crops, soil ecology and climate. Environ. Res. 2025, 266, 120540. [Google Scholar] [CrossRef] [PubMed]
  9. Ghosh, A. Nexus between agriculture and photovoltaics (agrivoltaics, agriphotovoltaics) for sustainable development goal: A review. Sol. Energy 2023, 266, 112146. [Google Scholar] [CrossRef]
  10. Zhang, L.; Gong, J.; Wu, C.; Murchie, E.H.; Gibbs, A.J.; Liu, B.; Yang, C.; Xu, G.; Zhang, J.; Guo, J.; et al. Research on time series prediction of microclimate in agrivoltaic systems based on the long short-term memory and attention mechanism. Front. Plant Sci. 2026, 17, 1755040. [Google Scholar] [CrossRef]
  11. Santra, P.; Meena, H.M.; Yadav, O.P. Spatial and temporal variation of photosynthetic photon flux density within agrivoltaic system in hot arid region of India. Biosyst. Eng. 2021, 209, 74–93. [Google Scholar] [CrossRef]
  12. Zhang, L.; Gong, J.; Yang, Z.; Wu, X.; Wang, W.; Yang, C.; Xu, G.; Wu, C.; Bao, E. Evaluating the contribution of decreasing heights of photovoltaic panels on light environment and agricultural production in agrivoltaic systems. J. Clean. Prod. 2025, 495, 145091. [Google Scholar] [CrossRef]
  13. Jo, H.; Song, J.T.; Cho, H.; Lee, S.; Choi, S.; Jung, H.-J.; Lee, H.-N.; Lee, J.-D. Evaluation of yield and yield components of rice in vertical agro-photovoltaic system in South Korea. Agriculture 2024, 14, 920. [Google Scholar] [CrossRef]
  14. Hu, Y.; Zhang, X.; Ma, X. Agrivoltaics with semitransparent panels can maintain yield and quality in soybean production. Sol. Energy 2024, 282, 112978. [Google Scholar] [CrossRef]
  15. Leroy, V.; Zarzoso-Lacoste, D.; Decocq, G.; Noirot-Cosson, P.-E.; Marrec, R. Impact of dual-axis elevated photovoltaic systems on bats’ activity and foraging behavior in wheat fields and hay meadows. Agric. Ecosyst. Environ. 2026, 404, 110373. [Google Scholar] [CrossRef]
  16. Pataczek, L.; Weselek, A.; Bauerle, A.; Högy, P.; Lewandowski, I.; Zikeli, S.; Schweiger, A. Agrivoltaics mitigate drought effects in winter wheat. Physiol. Plant. 2023, 175, 14081. [Google Scholar] [CrossRef]
  17. Clauw, H.; Van de Put, H.; Sghaier, A.; Kerkaert, T.; Debonne, E.; Eeckhout, M.; Steppe, K. The impact of a six-hour light–dark cycle on wheat ear emergence, grain yield, and flour quality in future plant-growing systems. Foods 2024, 13, 750. [Google Scholar] [CrossRef]
  18. Prakash, V.; Lunagaria, M.M.; Trivedi, A.P.; Upadhyaya, A.; Kumar, R.; Das, A.; Kumar Gupta, A.; Kumar, Y. Shading and PAR under different density agrivoltaic systems, their simulation and effect on wheat productivity. Eur. J. Agron. 2023, 149, 126922. [Google Scholar] [CrossRef]
  19. Asa’a, S.-N.; Ma Lu, S.; Kaaya, I.; Dupon, O.; de Jong, R.; van der Heide, A.; Bouguerra, S.; Radhakrishnan, H.S.; Poortmans, J.; Campana, P.E.; et al. Evaluating the influence of different agrivoltaic topologies on PV energy, crop yields and land productivity in a temperate climate. Renew. Energy 2025, 252, 123528. [Google Scholar] [CrossRef]
  20. Yalçın, Ö.; Kuzyaka, D.; Özden, T. Agrivoltaic system design for sugar beets and wheat in central Anatolia. Renew. Energy 2025, 245, 122800. [Google Scholar] [CrossRef]
  21. Zhang, L.; Yang, Z.; Wu, X.; Wang, W.; Yang, C.; Xu, G.; Wu, C.; Bao, E. Open-field agrivoltaic system impacts on photothermal environment and light environment simulation analysis in Eastern China. Agronomy 2023, 13, 1820. [Google Scholar] [CrossRef]
  22. Hussain, S.; Naseer, M.A.; Guo, R.; Han, F.; Ali, B.; Chen, X.; Ren, X.; Alamri, S. Nitrogen application enhances yield, yield-attributes, and physiological characteristics of dryland wheat/maize under strip intercropping. Front. Plant Sci. 2023, 14, 1150225. [Google Scholar] [CrossRef] [PubMed]
  23. Yang, X.; Wang, X.; Li, Y.; Yang, L.; Hu, L.; Han, Y.; Wang, B. Effects of drought stress at the booting stage on leaf physiological characteristics and yield of rice. Plants 2024, 13, 3464. [Google Scholar] [CrossRef]
  24. Liu, Y.; Cai, W.; Zhu, K.; Xu, Y.; Wang, W.; Zhang, H.; Gu, J.; Wang, Z.; Liu, L.; Zhang, J.; et al. Comparison of population photosynthesis characteristics and grain yield of wheat under various sowing dates and seeding rates. Crop Sci. 2024, 64, 2894–2907. [Google Scholar] [CrossRef]
  25. Prakash, V.; Lunagaria, M.M.; Upadhyaya, A.; Das, A.; Gupta, A.K.; Kumar, Y.; Subash, N.; Ghasal, P.C.; Singh, A.K.; Trivedi, A.P.; et al. Influence of agrivoltaic system-induced microclimate modifications on wheat growth and yield. Agric. For. Meteorol. 2025, 373, 110775. [Google Scholar] [CrossRef]
  26. Yang, X.; Ma, Y. Design and thermal-optical environment simulation of double-slope greenhouse roof structure based on Ecotect. Agriculture 2024, 14, 1410. [Google Scholar] [CrossRef]
  27. Ma, C.; Huang, Y.; Yu, K.; Liu, Z.; Lan, F.; Liu, B.; Ji, J.; Liang, Y.; Jia, C.; Sheng, Q.; et al. Shifting from large-scale greening to gray-space greening: Leveraging data to quantify the microclimate and energy benefits of viaduct under-spaces. Energy Build. 2026, 360, 117315. [Google Scholar] [CrossRef]
  28. Long, S.P.; Bernacchi, C.J. Gas exchange measurements, what can they tell us about the underlying limitations to photosynthesis? Procedures and sources of error. J. Exp. Bot. 2003, 54, 2393–2401. [Google Scholar] [CrossRef] [PubMed]
  29. Farquhar, G.D.; Von Caemmerer, S.; Berry, J.A. A biochemical model of photosynthetic CO2 assimilation in leaves of C3 species. Planta 1980, 149, 78–90. [Google Scholar] [CrossRef]
  30. Zainali, S.; Lu, S.M.; Fernández-Solas, Á.; Cruz-Escabias, A.; Fernández, E.F.; Zidane, T.E.K.; Honningdalsnes, E.H.; Nygård, M.M.; Leloux, J.; Berwind, M.; et al. Modelling, simulation, and optimisation of agrivoltaic systems: A comprehensive review. Appl. Energy 2025, 386, 125558. [Google Scholar] [CrossRef]
  31. Mundu, M.M.; Sempewo, J.I.; Mahoro, G.B.; Sankarapandian, V.; Uti, D.E. Environmental Impacts of Agri-Photovoltaics Systems on Local Ecosystems. Food Energy Secur. 2026, 15, e70186. [Google Scholar] [CrossRef]
  32. Jamil, U.; Hickey, T.; Pearce, J.M. Solar energy modelling and proposed crops for different types of agrivoltaics systems. Energy 2024, 304, 132074. [Google Scholar] [CrossRef]
  33. Ali Abaker Omer, A.; Li, M.; Zhang, F.; Hassaan, M.M.E.; El Kolaly, W.; Zhang, X.; Lan, H.; Liu, J.; Liu, W. Impacts of agrivoltaic systems on microclimate, water use efficiency, and crop yield: A systematic review. Renew. Sustain. Energy Rev. 2025, 221, 115930. [Google Scholar] [CrossRef]
  34. Mamun, M.A.A.; Dargusch, P.; Wadley, D.; Zulkarnain, N.A.; Aziz, A.A. A review of research on agrivoltaic systems. Renew. Sustain. Energy Rev. 2022, 161, 112351. [Google Scholar] [CrossRef]
  35. Fang, L.; Struik, P.C.; Girousse, C.; Yin, X.; Martre, P.; Kromdijk, J. Source–sink relationships during grain filling in wheat in response to various temperature, water deficit, and nitrogen deficit regimes. J. Exp. Bot. 2024, 75, 6563–6578. [Google Scholar] [CrossRef]
  36. Yang, H.; Dong, B.; Wang, Y.; Qiao, Y.; Shi, C.; Jin, L.; Liu, M. Photosynthetic base of reduced grain yield by shading stress during the early reproductive stage of two wheat cultivars. Sci. Rep. 2020, 10, 14353. [Google Scholar] [CrossRef] [PubMed]
  37. Li, H.; Jiang, D.; Wollenweber, B.; Dai, T.; Cao, W. Effects of shading on morphology, physiology and grain yield of winter wheat. Eur. J. Agron. 2010, 33, 267–275. [Google Scholar] [CrossRef]
  38. Li, Y.; Zhao, J.; Ma, H.; Pu, L.; Zhang, J.; Huang, X.; Yang, H.; Fan, G. Shade tolerance in wheat is related to photosynthetic limitation and morphological and physiological acclimations. Front. Plant Sci. 2024, 15, 1465925. [Google Scholar] [CrossRef]
  39. Sturchio, M.A.; Knapp, A.K. Ecovoltaic principles for a more sustainable, ecologically informed solar energy future. Nat. Ecol. Evol. 2023, 7, 1746–1749. [Google Scholar] [CrossRef]
Agronomy 16 01111 g001
Figure 1. Structural diagram of the large-span flexible AV system.
Figure 1. Structural diagram of the large-span flexible AV system.
Agronomy 16 01111 g001
Agronomy 16 01111 g002
Figure 2. Measuring points diagram of microclimate inside the flexible AV system.
Figure 2. Measuring points diagram of microclimate inside the flexible AV system.
Agronomy 16 01111 g002
Agronomy 16 01111 g003
Figure 3. Changes in temperature, precipitation and hours of sunshine during the wheat growth period from 2023 to 2024.
Figure 3. Changes in temperature, precipitation and hours of sunshine during the wheat growth period from 2023 to 2024.
Agronomy 16 01111 g003
Agronomy 16 01111 g006
Figure 6. Monthly mean solar radiation intensity during the two-year growth period.
Figure 6. Monthly mean solar radiation intensity during the two-year growth period.
Agronomy 16 01111 g006
Agronomy 16 01111 g007
Figure 7. Monthly mean daylighting rate during the two-year growth period.
Figure 7. Monthly mean daylighting rate during the two-year growth period.
Agronomy 16 01111 g007
Agronomy 16 01111 g008
Figure 8. Solar radiation curves within flexible AV systems on typical sunny and cloudy days.
Figure 8. Solar radiation curves within flexible AV systems on typical sunny and cloudy days.
Agronomy 16 01111 g008
Agronomy 16 01111 g009
Figure 9. Changes in accumulated air temperature during wheat growth cycle.
Figure 9. Changes in accumulated air temperature during wheat growth cycle.
Agronomy 16 01111 g009
Agronomy 16 01111 g010
Figure 10. Changes in accumulated soil temperature during wheat growth cycle.
Figure 10. Changes in accumulated soil temperature during wheat growth cycle.
Agronomy 16 01111 g010
Agronomy 16 01111 g011
Figure 11. Shadow variation characteristics of farmland within large-span flexible AV systems.
Figure 11. Shadow variation characteristics of farmland within large-span flexible AV systems.
Agronomy 16 01111 g011
Agronomy 16 01111 g012
Figure 12. Distribution of average daylighting rate of the flexible AV system in two periods (December–February and March–May).
Figure 12. Distribution of average daylighting rate of the flexible AV system in two periods (December–February and March–May).
Agronomy 16 01111 g012
Agronomy 16 01111 g013
Figure 13. Dry weight, fresh weight, and yield indicators of wheat. (a,b) Dry and fresh weight. (c,d) Wheat yield. Different lowercase letters indicate significant differences in the indexes under different treatments (p < 0.05).
Figure 13. Dry weight, fresh weight, and yield indicators of wheat. (a,b) Dry and fresh weight. (c,d) Wheat yield. Different lowercase letters indicate significant differences in the indexes under different treatments (p < 0.05).
Agronomy 16 01111 g013
Agronomy 16 01111 g014
Figure 14. Growth indicators of wheat. (a,b) Plant height. (c,d) Stem diameter. (e,f) Leaf area. Different lowercase letters indicate significant differences in the indexes under different treatments (p < 0.05).
Figure 14. Growth indicators of wheat. (a,b) Plant height. (c,d) Stem diameter. (e,f) Leaf area. Different lowercase letters indicate significant differences in the indexes under different treatments (p < 0.05).
Agronomy 16 01111 g014
Agronomy 16 01111 g015
Figure 15. Photosynthetic characteristics during the main growth period of winter wheat. (a) From 2023 to 2024. (b) From 2024 to 2025. Different lowercase letters indicate significant differences in the indexes under different treatments (p < 0.05).
Figure 15. Photosynthetic characteristics during the main growth period of winter wheat. (a) From 2023 to 2024. (b) From 2024 to 2025. Different lowercase letters indicate significant differences in the indexes under different treatments (p < 0.05).
Agronomy 16 01111 g015
Table 1. Yield components of wheat under different PV shading conditions.
Table 1. Yield components of wheat under different PV shading conditions.
YearTreatmentsSpike Number/
(104·hm−2)		Grain Number Per SpikeThousand
Grain Weight/g
2024CK453.67 ± 16.81 a37.08 ± 8.36 a39.92 ± 1.27 b
BP250.13 ± 11.53 b33.42 ± 4.19 ab45.49 ± 0.51 a
UP254.67 ± 21.78 b30.83 ± 6.10 b45.11 ± 1.18 a
2025CK448.13 ± 17.24 a36.33 ± 7.15 a39.27 ± 1.09 b
BP259.80 ± 12.06 b32.00 ± 3.85 b44.75 ± 0.73 a
UP254.88 ± 18.92 b30.87 ± 5.72 b45.14 ± 1.05 a
Different lowercase letters indicate significant differences in the indexes under different treatments (p < 0.05).
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

You, Y.; Yu, M.; Geng, X.; Teng, J.; Qu, Z.; Zhang, L.; Bao, E. Assessing the Effects of Large-Span Flexible Photovoltaic Arrays on Farmland Microclimate and Wheat Productivity: A Two-Year Field Experiment. Agronomy 2026, 16, 1111. https://doi.org/10.3390/agronomy16111111

AMA Style

You Y, Yu M, Geng X, Teng J, Qu Z, Zhang L, Bao E. Assessing the Effects of Large-Span Flexible Photovoltaic Arrays on Farmland Microclimate and Wheat Productivity: A Two-Year Field Experiment. Agronomy. 2026; 16(11):1111. https://doi.org/10.3390/agronomy16111111

Chicago/Turabian Style

You, Yanfei, Minli Yu, Xiayun Geng, Jiaxun Teng, Zhonghao Qu, Long Zhang, and Encai Bao. 2026. "Assessing the Effects of Large-Span Flexible Photovoltaic Arrays on Farmland Microclimate and Wheat Productivity: A Two-Year Field Experiment" Agronomy 16, no. 11: 1111. https://doi.org/10.3390/agronomy16111111

APA Style

You, Y., Yu, M., Geng, X., Teng, J., Qu, Z., Zhang, L., & Bao, E. (2026). Assessing the Effects of Large-Span Flexible Photovoltaic Arrays on Farmland Microclimate and Wheat Productivity: A Two-Year Field Experiment. Agronomy, 16(11), 1111. https://doi.org/10.3390/agronomy16111111

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop