📄 agriculture

Genotype × environment interaction, adaptability, and stability of soybean genotypes in a new production region of Brazil

This study utilized REML/BLUP and HMRPGV methods to analyze genotype × environment interactions in soybeans across five environments in northern Rio de Janeiro, Brazil, identifying genotypes 95R95IPRO, BMX COMPACTA IPRO, and TMG7061IPRO as superior candidates for cultivation in this new production region due to their high yield, adaptability, and stability.

Original authors: Josimar Nogueira Batista, Derivaldo Pureza Cruz, Yure Pequeno de Souza, Letícia Pastore Mendonça, Pedro Henrique Silva Ferreira, Roni Fernandes Guareschi, Elderson Pereira Silva, Benedito Fernandes So

Published 2026-07-08

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Josimar Nogueira Batista, Derivaldo Pureza Cruz, Yure Pequeno de Souza, Letícia Pastore Mendonça, Pedro Henrique Silva Ferreira, Roni Fernandes Guareschi, Elderson Pereira Silva, Benedito Fernandes Souza Filho, Claudia Pozzi Jantalia, Jerri Édson Zilli, Roberto Kazuhiko Zito, Geraldo Amaral Gravina

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you are a chef trying to find the perfect recipe for a cake. You have eight different cake mixes (the soybean genotypes) and you want to see which one tastes best. But here's the catch: you aren't baking in just one kitchen. You are baking in five different kitchens (the environments), each with its own oven temperature, humidity, and altitude.

This study is essentially a massive taste test to figure out which cake mix works best in these specific Brazilian kitchens, and which ones are "fussy" (only good in one kitchen) versus "versatile" (good in all of them).

Here is the breakdown of what the researchers found, using simple analogies:

1. The Big Problem: The "Kitchen" Matters More Than the "Mix"

The researchers discovered something surprising: The kitchen mattered way more than the cake mix.

In scientific terms, they found a huge "Genotype × Environment" interaction. Think of it like this: If you take a great runner and put them on a muddy track, they might lose. If you put them on a smooth track, they win. The runner (the soybean) didn't change, but the track (the environment) changed their performance completely.

In this study, the differences in weather and soil between the five locations caused about 80% of the variation in how much soy the plants produced. This means that a soybean variety that produces a huge harvest in one town might produce a tiny harvest in a town just 50 miles away.

2. The Challenge: Predicting the Future is Hard

Because the "kitchens" were so different, it was very hard to predict which cake mix would win just by looking at the average. The researchers' "accuracy score" for predicting the winners was quite low (0.18).

Imagine trying to guess who will win a race where the track changes every time you run. It's a gamble. This tells us that you can't just pick the "best" soybean based on one test; you have to test it in many different places to be sure.

3. The Winners: The "All-Weather" Champions

Despite the chaos of the different environments, the researchers used a special statistical tool (think of it as a super-smart referee) to find the soybeans that were the most consistent.

They looked for three things:

High Yield: Did it produce a lot of beans?
Stability: Did it perform well even when the weather was weird?
Adaptability: Could it handle different conditions without crashing?

The Top Three Champions:

95R95IPRO
BMX COMPACTA IPRO
TMG7061IPRO

These three were the "all-weather" athletes. They didn't just win in one specific kitchen; they consistently produced high yields across the different farms and seasons. They were the most reliable performers.

4. The Losers: The "Fussy" Varieties

Some varieties, like BRS 544RR, struggled. They were like a cake mix that only works if the oven is exactly 350 degrees. If the temperature dropped or the humidity rose, the result was poor. The study noted that this variety produced significantly less than the others, making it a risky choice for farmers in this new region.

5. The "Magic" Math Tool

The researchers used a method called REML/BLUP. You can think of this as a magic filter.

Regular math might just say, "This plant produced 100 beans on average."
The "Magic Filter" says, "Wait, that plant produced 100 beans, but the soil was terrible that day. If the soil had been normal, it probably would have produced 120. Let's adjust the score to reflect the plant's true talent, not just the bad luck of the weather."

This tool helped them separate the plant's actual genetic potential from the luck of the weather.

The Bottom Line

The study concludes that in this new soybean region of Brazil (Northern Rio de Janeiro), you cannot just pick the "best" soybean seed blindly. Because the weather and soil change so much from farm to farm, you need seeds that are stable and adaptable.

The varieties 95R95IPRO, BMX COMPACTA IPRO, and TMG7061IPRO are the recommended "all-weather" champions for this area. They are the ones farmers should plant if they want a reliable harvest, regardless of whether the season is wet, dry, hot, or cool.

Technical Summary: Genotype × Environment Interaction, Adaptability, and Stability of Soybean Genotypes in a New Production Region of Brazil

Problem Statement
Soybean (Glycine max) production in Brazil is expanding into new agricultural frontiers, specifically the northern and northwestern regions of Rio de Janeiro State, which offer over 300,000 hectares suitable for mechanized cropping. However, the performance of soybean genotypes is heavily influenced by edaphoclimatic conditions, leading to significant genotype × environment (G×E) interactions. While statistical methods like mixed models (REML/BLUP) are established for assessing stability and adaptability, there is a scarcity of research addressing these specific parameters for soybean genotypes under the unique conditions of Rio de Janeiro's emerging production regions. This gap hinders the accurate recommendation of cultivars and the optimization of production strategies in this new frontier.

Methodology
The study evaluated eight commercial soybean genotypes (including cultivars such as 95R95IPRO, BMX COMPACTA IPRO, and BRS 7780IPRO) across five distinct environments. These environments were defined by combinations of two locations (Campos dos Goytacazes and Macaé) and two growing seasons (2019/2020 and 2020/2021). The experiments utilized a randomized complete block design with four replications.

Data analysis was conducted using the Selegen REML/BLUP software. The study employed a linear mixed model ($y = Xr + Zg + Wi + e$) to estimate variance components and genotypic values. Key statistical parameters calculated included:

Variance Components: Genotypic ( $\sigma^2_g$ ), G×E interaction ( $\sigma^2_{ge}$ ), and residual ( $\sigma^2_e$ ) variances.
Genetic Parameters: Heritability based on genotype means ( $\hat{h}^2$ ), selection accuracy (SA), and genotypic correlation between environments ( $\hat{r}_{ge}$ ).
Selection Indices: To assess adaptability and stability, the study utilized the Harmonic Mean of Genotypic Values (HMGV) for yield and stability, the Relative Performance of Genotypic Values (RPGV) for adaptability, and the Harmonic Mean of the Relative Performance of Genotypic Values (HMRPGV) for simultaneous selection of yield, stability, and adaptability.

Key Results

G×E Interaction: A significant and predominant G×E interaction was detected, accounting for approximately 80% of the phenotypic variance in grain yield. The genotypic correlation between environments was extremely low ( $\hat{r}_{ge} = 0.01$ ), indicating that genotype rankings shifted significantly across different environments.
Genetic Parameters: The study observed low heritability for grain yield ( $\hat{h}^2 = 0.03$ ) and low selection accuracy (SA = 0.18), attributed to the high influence of environmental factors and the strong G×E interaction. The experimental coefficient of variation (CVe) was 10.76%, falling within acceptable limits for soybean trials.
Genotype Performance:
- Overall Yield: The overall mean grain yield was 3,281.15 kg ha⁻¹.
- Top Performers: The genotypes 95R95IPRO, BMX COMPACTA IPRO, and TMG7061IPRO consistently demonstrated superior performance.
- Stability and Adaptability: The HMRPGV method identified 95R95IPRO, BMX COMPACTA IPRO, and TMG7061IPRO as the top three genotypes for simultaneous yield, stability, and adaptability.
- Specific Gains: Genotype 95R95IPRO showed the highest genetic gain (13.82 kg ha⁻¹) over the general mean when considering interaction-free values.
- Environmental Specificity: While some genotypes (e.g., BRS 7780IPRO) excelled in specific environments (such as E2, E3, and E4), the top three identified by HMRPGV maintained high performance across the diverse conditions tested.
- Yield Comparison: The top-performing genotypes in the Macaé region exceeded the national average yield for the 2023/2024 harvest (3,202 kg ha⁻¹), with some reaching over 6,000 kg ha⁻¹ in specific favorable environments.

Significance and Claims
The paper claims that the significant G×E interaction observed necessitates the use of mixed-model approaches (REML/BLUP) combined with stability and adaptability indices (HMGV, RPGV, HMRPGV) rather than relying solely on average performance for cultivar recommendation in this new region.

The study highlights that genotypes 95R95IPRO, BMX COMPACTA IPRO, and TMG7061IPRO possess the necessary genetic potential for broad adaptability and stability in the northern region of Rio de Janeiro. These genotypes are presented as viable candidates for recommendation in breeding programs and for cultivation in the target region. The authors emphasize that while average performance is important, the consistency of genotypic expression across environments—validated by the HMRPGV method—is critical for successful genetic improvement and cultivar deployment in areas characterized by complex environmental interactions. The results reinforce the utility of these statistical tools in supporting genotype selection strategies for emerging soybean frontiers in Brazil.