Skip to content
What sugarcane, coffee and citrus breeding can teach us about performance, adaptation and resilience
It is a familiar experience in breeding: a genotype that looks outstanding in one trial can lose its advantage in the next. Even when plants come from the same population and trials are conducted within the same region, rankings can shift as disease pressure increases, rainfall declines, temperatures fluctuate, or stress occurs at a different stage of crop development.
The harder question is what a change in ranking actually tells us.
Phenotypic performance is conditional. It reflects genetic potential expressed through a particular combination of environment, season, management and stress. Genotype × environment (G×E) interaction is therefore not simply a statistical effect to be accommodated in the analysis; it can be part of the biological signal behind the phenotype.
When the Environment Changes the Signal
Sugarcane smut provides a clear example. Multi-environment trials in South Africa showed significant effects of genotype, location and season on disease expression, including genotype × location, genotype × season and genotype × location × season interactions. Some environments also provided greater discrimination among genotypes for smut resistance than others [1].
Resistance, therefore, needs to be interpreted within its environmental and epidemiological context. Limited disease pressure can mask genetic differences, whereas conditions that favour infection can expose them.
When performance changes across trials, the shift itself may become part of the evidence.
I experienced this directly in sugarcane during the 2023/24 season, when an unusual wilt syndrome affected both materials under selection and commercial materials. Its causes were not fully understood at the time; subsequent investigations associated the syndrome with Phaeocytostroma sacchari, Colletotrichum falcatum and Fusarium spp. [2,3]. Under such uncertainty, a single season was not sufficient to justify an immediate selection decision.
An unusual phenotype is information before it is a verdict. It deserves investigation before it becomes a selection criterion.
When Does Performance Travel?
Coffee reinforces the same principle from another angle. An international evaluation of 29 Coffea arabica varieties across 23 sites in Africa, Asia and Latin America found strong genotype × site interaction for resistance to coffee leaf rust. Some varieties maintained their performance across locations, whereas others expressed a much stronger advantage in particular environments. Notably, the most resistant varieties were not necessarily the most stable [4].
Mean performance and stability therefore capture different aspects of adaptation. A genotype can have a high average level of resistance while responding strongly to environmental conditions, whereas another can maintain a more consistent response across sites. The relevant profile depends on the target population of environments and the breeding objective.
My experience working with coffee reinforced this broader view of selection. When evaluating Coffea canephora, we were not selecting for disease resistance or productivity in isolation. We considered yield alongside disease response, fruit characteristics, maturation and plant architecture, while also looking for tolerance to drought. The value of a genotype depended on how these traits performed together under the conditions in which the material would ultimately be grown.
When Survival Is Not Enough
Huanglongbing (HLB), one of the most important diseases affecting citrus production, highlights the difference between disease response and productive performance under disease pressure. Infection status or symptom severity alone provides an incomplete picture of breeding value. A tree that maintains growth, yield and fruit quality despite infection represents a different breeding outcome from one that simply develops fewer symptoms.
Recent work with early-season sweet orange under endemic HLB pressure identified surviving selections that maintained canopy vigour, fruit retention, yield and juice quality despite infection. The study combined field performance with genomic and physiological analyses, providing evidence that HLB tolerance involves complex mechanisms beyond simply limiting pathogen infection [5].
For citrus breeding, the relevant phenotype is therefore broader than disease response itself. The key question is whether commercially important performance can be maintained despite disease.
What Makes a Genotype Good?
The harder question is what happens when selection criteria and production value do not point in exactly the same direction.
I have seen this distinction clearly in sugarcane. In advanced trials conducted on customers’ farms, some materials were discontinued within the breeding programme, yet growers continued propagating them because of their field performance. In some cases, this continued use contributed to their eventual release as regional varieties.
For me, these experiences shifted the focus from identifying the genotype that performs best to understanding which strengths, weaknesses and trade-offs matter in the target production system.
This is how I think about resilience in breeding: not as a stand-alone trait, but as the capacity to sustain agronomic value when environmental conditions change.
Beyond the Trial
No breeding programme can reproduce every environment a cultivar will encounter. Multi-environment testing therefore helps reveal where genetic differences are consistent, where they are environment-specific, and where further evidence is needed.
Ultimately, the value of a genotype is tested beyond the trial, when it enters the production system for which it was selected.
Perhaps, then, the more useful question in breeding is not “Which genotype is the best?” but “Best for what?”
The answer depends on the target environment, the stresses that matter and the traits that create value in the production system.
A good genotype is not necessarily the one that wins the most trials. It is the one whose strengths remain relevant when it moves from experimental conditions into the production environments for which it was selected.
Breeding is not only about selecting for the environments we have today. It is also about selecting for the environments we expect to face.
References
1. Zhou, M. M. (2022). Location and seasonal effects on sugarcane smut and implications for resistance breeding in South Africa. South African Journal of Plant and Soil, 39(1), 1–9. https://doi.org/10.1080/02571862.2021.1969692.
2. Santos, A. K. V. dos. (2025). Fungi associated with rot and wilting syndrome in sugarcane. FAO AGRIS.
3. Santos, A. K. V. dos. (2025). Fungos associados à podridão e à síndrome da murcha em cana-de-açúcar (Fungi associated with rot and wilting syndrome in sugarcane). [Dissertação de mestrado, Universidade Federal de Lavras]. Repositório Institucional da UFLA. https://repositorio.ufla.br/handle/1/60261
4. Berny Mier y Teran, J. C., Pruvot-Woehl, S., Maina, C., et al. (2025). Global Coffea arabica variety trials reveal genotype-by-environment interactions in resistance to coffee leaf rust (Hemileia vastatrix). Frontiers in Plant Science, 16, 1583595. https://doi.org/10.3389/fpls.2025.1583595.
5. Ramekar, et al. (2026). Genomic assessment of early-season sweet oranges provides insights into the mechanisms behind their enhanced tolerance to ‘Candidatus Liberibacter asiaticus’ and huanglongbing disease. Physiological and Molecular Plant Pathology, 145, 103369. https://doi.org/10.1016/j.pmpp.2026.103369.
