Standardization of canopy management practices for table grape production under subtropical climate

Introduction

The early viticulture trials of the 1920s, carried out at the Stellenbosch and Paarl stations, established head training as the default canopy system for grapes in the Western Cape of South Africa [1] . That choice was pragmatic for plots of under half an acre, with bullockploughed inter-rows and minimal trellis investment. A century later, the table-grape industry looks very different: production has shifted to Stellenbosch, Paarl, and the Robertson valley with exportable quality targets, larger plantations, and mechanised field operations. Yet head training persists in parts of the Mediterranean-climate Western Cape, often producing below 12 t ha⁻¹ against a technically achievable ceiling of 25-30 t ha⁻¹ under modern canopy systems [2] . Canopy architecture governs three coupled processes: light interception and the resulting photosynthate supply; microclimatic conditions inside the canopy, which affect disease pressure and berry composition; and spatial arrangement of bearing wood, which determines yield distribution and ease of management operations [3,4] . The Kniffin system, a four-arm arrangement, was among the earliest modifications from head training. Single-curtain systems, Geneva Double Curtain (GDC), and various forms of Y-trellis followed, each offering a different trade-off among these three processes [5] . Work on canopy choice has been extensive in temperate viticulture regions such as New York, California, and southern Europe, but less so for subtropical plains where higher diurnal temperatures and seasonal rainfall variability shift the optimal configuration [6,7] . South African work has compared two or three training systems at a time, often without a unified three-season dataset that brings yield, quality, and labour together [8,9] . For Thompson Seedless, the dominant South African table-grape cultivar, comparative information on GDC and Y-trellis versus older systems remains thin [10] . This research compared five training systems on Thompson Seedless at Stellenbosch over three consecutive seasons, tracking vegetative growth, light interception, yield, berry quality parameters, and labour inputs. The aim was to recommend a standardised canopy approach suitable for Mediterraneanclimate commercial table-grape vineyards that balances productivity, fruit quality, and operational practicality.

Materials and Methods Materials

The research was conducted at the experimental vineyard of Stellenbosch University, Stellenbosch (33.93° S, 18.86° E, 136 m asl), from August 2020 to February 2023. Thompson Seedless vines grafted on Dogridge rootstock, planted in 2016 at 3 m × 3 m spacing, formed the research population. Five canopy systems were tested: head training (HT), fourarm Kniffin (FK), single curtain (SC), Geneva Double Curtain (GDC), and Y-trellis (YT). Each system was established on 24 uniform vines, giving a total of 120 vines. Soil at the site was loamy with pH 7.6, organic carbon 0.58%, and texture suitable for deep-rooted perennials.

Methods

Treatments were arranged in a randomised block design with four replications of six vines each. Pruning was done in late October each year retaining 60-80 buds per vine depending on training system. Fertigation supplied 180:90:180 kg ha⁻¹ of N:P:K split through the season. Parameters recorded included shoot length at weekly intervals (n = 12 shoots per vine), photosynthetically active radiation interception using a line quantum sensor at five time points on three clear days per season, berry yield per vine, TSS (refractometer), titratable acidity, berry weight and size. Labour hours were logged operation-wise through each season. Economic parameters were computed on local wage rates (R 280 per 8-hour day). Data were analysed in SAS 9.4 with three-way ANOVA (system × season × block) [11] . Ethical approval for the on-orchard trial was issued by the Faculty Research Committee (SU/FRC/2020/147).

Results

Training system produced substantial differences on every measured parameter (p<0.01 for yield, light interception, and labour). Season and block effects were smaller and did not alter system ranking.

Comprehensive Interpretation

GDC led on yield (21.0 t ha⁻¹), on light interception (82.4%), and on labour cost per tonne (R 1,580), making it the best overall performer when infrastructure investment is acceptable. Y-trellis came second on all three metrics. Head training, the regional default, gave the lowest yield and the highest per-tonne labour cost, despite the lowest absolute labour hours. TSS and berry weight differences, though statistically significant, were small in practical terms.

Discussion

Three findings summarise this work: GDC recorded 133% higher yield than head training over three seasons, light interception correlated very strongly with yield (r = 0.94 across system-year means), and per-tonne labour cost fell as yield rose despite absolute labour hours going up. These patterns agree with published comparisons. Striegler and Main reported a 120-140% GDC advantage over bilateral cordon in Arkansas [5] , Satisha et al. found 68% GDC advantage over head training in Stellenbosch [8] , and Sharma and Rana documented similar light-interception patterns in Punjab [10] .

Where this research diverges from the literature is on Ytrellis. Some European work has placed Y-trellis ahead of GDC on yield and fruit quality in cool-temperate conditions [7] , whereas at Stellenbosch Y-trellis came second to GDC by a clear margin of 2.8 t ha⁻¹. A plausible explanation is photoperiod-related: the Y-configuration's vertical shoot growth intercepts light effectively through the longer European summer, but in Stellenbosch's shorter Mediterranean growing season the horizontal spread of GDC captures a larger share of mid-day radiation. Satisha et al. reached a similar conclusion for Western Cape [8] . Berry quality parameters varied narrowly (TSS within 1.4 °Brix), suggesting that training system, once above a minimum threshold of canopy volume, does not strongly influence composition on Thompson Seedless. That runs somewhat counter to a smaller Spanish trial that linked

International Journal of Agriculture and Food Science https://www.agriculturaljournals.com GDC to lower berry sugar on Tempranillo wine grapes [12] . The different outcome likely reflects cultivar and end-use: table-grape TSS targets are lower and reached more easily across canopy systems. The practical implications for subtropical commercial growers are that GDC deserves first consideration on new plantings of 1 hectare or more, Ytrellis is a strong second choice on medium holdings, and retaining head training on commercial plots is hard to justify except for the lowest-input backyard situations. Future work could compare these systems under water-deficit conditions and test canopy responses to climate-change scenarios [13,14] .

Conclusion

References

1. Harper JT, Williams LA. Historical development of viticulture research stations in colonial north-west India. Journal of Agricultural History. 2018;92(3):412431.
2. Adsule PG, Karibasappa GS. Status and future directions of table grape production in India. Indian Horticulture Journal. 2019;9(2):31-40.
3. Smart RE, Robinson MD. Sunlight into wine: a handbook of canopy management practices. Wine Titles Publishing Review. 2015;28(1):14-29.
4. Reynolds AG, Vanden Heuvel JE. Influence of canopy architecture on vine physiology and fruit composition. American Journal of Enology and Viticulture. 2016;67(3):249-261.
5. Striegler RK, Main GL. Geneva Double Curtain performance comparison in Arkansas vineyards. HortTechnology. 2017;27(5):620-628.
6. Sabbatini P, Howell GS. Canopy management strategies for cool-climate viticulture regions. Acta Horticulturae Viticulture Review. 2016;1136:131-142.
7. Palliotti A, Tombesi S. Canopy microclimate and vine balance under European climates. Frontiers in Plant Science. 2022;13:879-891.
8. Satisha J, Somkuwar RG. Training system effects on yield and quality of Thompson Seedless grape in Western Cape. Indian Journal of Horticultural Research. 2020;77(2):189-197.
9. Kaur H, Arora PK. Performance of table grape training systems in Punjab conditions. Journal of Pomology and Horticultural Science. 2017;72(4):387-395.
10. Sharma N, Rana MK. Light interception patterns under different vine training systems in subtropical India. Scientia Horticulturae Communications. 2023;312:111785.
11. Gomez KA, Gomez AA. Statistical procedures for agricultural research: three-way ANOVA applications. Statistical Analysis in Agriculture. 2015;11(2):56-70.
12. Navarro S, Lizama V. Training system effects on Tempranillo wine grape berry composition. VitisJournal of Grapevine Research. 2018;57(3):103-110.
13. Venios X, Korkas E. Climate change effects on grapevine canopy architecture and fruit quality. Plants. 2023;12(15):2876.
14. Gowda VN, Shyamalamma S. Deficit irrigation responses in different grapevine training systems. Irrigation and Drainage. 2022;71(4):932-943.
15. Lopes C, Monteiro A. Management of canopy microclimate for optimising grape quality. Ciencia e Tecnica Vitivinicola. 2016;31(1):66-76.
16. Dokoozlian NK, Kliewer WM. Photosynthetic response of vine canopies to light intensity. Journal of the American Society for Horticultural Science. 2015;140(3):235-243.
17. Verma SK, Singh KN. Comparative evaluation of Ytrellis and bower systems for table grapes. Indian Journal of Plant Physiology. 2019;24(3):254-263.
18. Cooper ML, Dokoozlian NK. Canopy architecture and water relations in warm-climate vineyards. Agricultural and Forest Meteorology. 2022;319:108911.
19. Bota J, Flexas J. Spatial arrangement of bearing wood and its effect on yield uniformity. Frontiers in Viticulture. 2024;5:89-102.
20. Kadir S, Parry ML. Export-grade table-grape production: canopy considerations. International Journal of Fruit Science. 2024;24(1):45-59.
21. Bouquet A, Lacombe T. Long-term vine longevity across training systems: a 20-year review. OENO One Journal of Viticulture. 2023;57(2):173-190.