Resilience of black-boned and Thai native chickens to heat stress: Implications of temperature–humidity index for sustainable egg production

Abstract

As climate change intensifies, improving heat tolerance has become an important objective for sustaining poultry productivity in tropical and subtropical environments. This study evaluated the effects of the temperature–humidity index (THI) on monthly egg production and genetic parameters in black-boned and Thai native chickens, with an emphasis on identifying heat stress thresholds and genotype-specific responses. A total of 136,816 monthly egg production records from 3,680 black-boned and 7,850 Thai native chickens were analyzed using a repeatability test-day model. Heat stress onset was observed at THI 72 in black-boned chickens and THI 74 in Thai native chickens. Thai native chickens maintained higher egg production rates and sustained positive responses over a wider THI range. Heritability declined with increasing THI (0.162–0.107 in black-boned; 0.175–0.132 in Thai native chickens), and genetic correlations between monthly egg production and heat stress shifted from positive to negative in both chicken genotypes (from 0.350 to −0.723 in black-boned chickens and from 0.500 to −0.524 in Thai native chickens). The rate of decline in egg production increased with THI, and was greater in black-boned chickens than in Thai native chickens (from 1.01 to −2.03 eggs/bird/THI level in black-boned chickens and from 1.77 to −1.22 eggs/bird/THI level in Thai native chickens across THI 70–80). The proportion of chickens with heat-tolerant breeding values decreased across THI levels, but remained higher in Thai native chickens. The results of this study demonstrated a nonlinear, genotype-dependent response to heat stress and provided a framework for integrating THI into genetic evaluations to enhance resilience and sustain egg production under tropical conditions.

Introduction

Heat stress is one of the most critical environmental constraints affecting poultry production in tropical and subtropical regions (He et al., 2018; Juiputta et al., 2023). Prolonged exposure to high ambient temperatures, in conjunction with elevated humidity, disrupts physiological homeostasis in laying hens, leading to reduced egg production, impaired egg quality, and increased production variability (Kim et al., 2024; Tesakul et al., 2025; Promket et al., 2025). These adverse effects threaten the sustainability of poultry systems under the ongoing effects of climate change, particularly in countries where temperature extremes have become more frequent and persistent (Thornton et al., 2021; Kennedy et al., 2022; Mogano et al., 2024).

Thai native and black-boned chickens are traditionally reared under low-input systems and are widely recognized for their adaptability to harsh environmental conditions (Budi et al., 2023; Loengbudnark et al., 2024; Mokoena et al., 2025). Compared with intensively selected commercial layers, these indigenous genotypes often exhibit greater tolerance to thermal stress, which has been attributed to their unique genetic background and long-term adaptation to tropical climates (Tirawattanawanich et al., 2011; Duangjinda et al., 2017; Rachman et al., 2024). However, even heat-adapted chickens experience declines in laying performance when environmental stress exceeds the physiological buffering capacity (Yan et al., 2022; Loengbudnark et al., 2023; Kim et al., 2024). Understanding the ability of these breeds to maintain egg production under heat stress is crucial for the development of breeding strategies that enhance heat tolerance and support sustainable poultry production under changing climatic conditions.

The temperature–humidity index (THI) has been extensively applied as an integrated climatic indicator to quantify heat stress and its impact on livestock performance. In poultry, THIs are particularly useful for evaluating changes in growth, fertility, and egg production traits (Özentürk et al., 2024; Boonkum et al., 2025; Kenchaiwong et al., 2025). THI-based approaches allow researchers to characterize not only phenotypic responses to heat stress but also genetic variation in resilience across different environmental gradients (Misztal, 2017; Lallo et al., 2018; Buranawit et al., 2025). Incorporating THI into genetic evaluations enables the identification of heat-tolerant individuals and supports the development of breeding programs that balance productivity with environmental robustness (Boonkum et al., 2025).

Despite the widespread application of THI in poultry research, most studies have focused on commercial layer strains (Lara and Rostagno, 2013; Kim et al., 2024), with limited attention given to indigenous chickens, such as native and black-boned breeds. Moreover, many evaluations assume linear responses to increasing heat loads, although biological responses to thermal stress often exhibit threshold or nonlinear patterns (Nienaber et al., 1999; Nawab et al., 2018). These limitations may obscure breed-specific resilience mechanisms and underestimate their genetic potential for sustained egg production under high THI conditions. A comprehensive assessment of egg production responses to heat stress requires THI models that accurately capture the interactions between environmental conditions and genetic backgrounds (Loengbudnark et al., 2023). Evaluating the THI thresholds associated with performance decline can provide critical insights into the adaptive capacity and resilience of a breed. Such information is particularly valuable for indigenous chickens, which have increasingly been recognized as genetic resources for climate-resilient poultry production.

This study aimed to evaluate the phenotypic and genetic responses of black-boned and Thai native chickens to heat stress through the assessment of temperature–humidity index (THI) effects on monthly egg production. By identifying critical THI thresholds and characterizing breed-specific responses, we aimed to improve understanding of the biological and genetic basis of heat-stress resilience and to support the development of breeding and management strategies for sustainable poultry production in tropical environments.

Materials and methods

All experimental procedures used in this study were approved by the Institutional Animal Care and Use Committee (IACUC) of Khon Kaen University (KKU) under the Ethical Guidelines for Animal Experimentation of the National Research Council of Thailand (Approval No. IACUC-KKU-108/68).

Animals and management

The experiment was conducted between March 2020 and April 2026 at the experimental farm of the Network Center for Animal Breeding and Omics Research, Faculty of Agriculture, Khon Kaen University, Thailand. The study population consisted of Hmong black-boned chickens and Thai native chickens (Pradu Hang Dum), which were managed under identical environmental and nutritional conditions throughout the study period. At hatching, all chicks were leg-tagged for initial identification, brooded for four weeks, and subsequently fitted with wing tags for permanent identification. Birds were vaccinated against infectious bronchitis, Newcastle disease, fowl pox, and fowl cholera following the vaccination program recommended by the Department of Livestock Development, Thailand. Newcastle disease and infectious bronchitis vaccines were administered at 7, 21, and 42 days of age and thereafter at 3-month intervals. The fowl pox vaccine was administered at 35 days of age, whereas the fowl cholera vaccine was administered at 42 days of age and subsequently every 3 months (Kammongkun, 2025). The chickens were reared in an open-sided housing system. During the brooding phase (0–4 weeks of age), supplemental heating was provided according to standard management practices, and birds received a commercial starter diet containing 19% crude protein and 2,900 kcal ME/kg. From 4 to 20 weeks of age, birds were fed a grower diet containing 15% crude protein and 2,900 kcal ME/kg. Throughout the growing and laying periods, birds were maintained under natural photoperiod conditions (approximately 12 h light/day) without supplemental lighting. Commercial feed and fresh water were provided ad libitum throughout the study. At 20 weeks of age, the hens were transferred to individual battery cages (20 × 45 × 40 cm) in compliance with the animal welfare guidelines established by the Department of Livestock Development, Thailand. Monthly egg production was recorded from the onset of lay for 365 days of production. During the laying period, hens were fed 110 g/bird/day of a layer diet containing 17% crude protein and 2,750 kcal ME/kg, with ad libitum access to fresh water.

Data collection

Egg production was recorded at the individual hen level throughout the laying cycle. Daily egg records were collected for each hen and aggregated into monthly egg production records, defined as the total number of eggs produced during each consecutive 30-day period (monthly egg production records). The final dataset comprised 43,220 and 93,596 monthly records from 3,680 Hmong black-boned chickens and 7,850 Thai native chickens (Pradu Hang Dum), respectively. Monthly records were used because they provide a more stable measure of laying performance than daily observations, which are often influenced by short-term biological and environmental variation. In addition, monthly records are more suitable for evaluating the cumulative effects of heat stress, as ovarian follicle recruitment, growth, and maturation occur over several weeks. Thus, the use of monthly egg production records was biologically consistent with the 30-day average THI used to characterize environmental exposure. Monthly egg production records were collected for 365 days following the onset of lay, representing the first year of production. This period encompasses the major stages of the laying cycle and provides repeated observations suitable for genetic analyses of heat stress responses and production persistency.

Environmental data, THI calculation, and onset of heat stress estimation

To investigate the effects of heat stress on monthly egg production, daily air temperature (T) and relative humidity (RH) were continuously monitored using three temperature–humidity data loggers (ELITECH GSP-6; Elitech Technology, Inc., San Jose, CA, USA), with measurement accuracies of ±0.5°C for temperature and ±3% for relative humidity. The devices were installed at bird level in three representative locations within the housing system: the front, center, and rear sections. These environmental records were used to calculate the temperature–humidity index (THI), which served as an indicator of heat stress for evaluating its effects on monthly egg production. The THI equation employed in this study was originally proposed by the National Research Council (NRC, 1971) as a composite measure integrating ambient temperature and relative humidity. Although initially developed for general livestock applications, this index has been widely adopted in poultry research to assess thermal stress and its impacts on productive and reproductive performance (Loengbudnark et al., 2023; ; Boonkum et al., 2025). THI was calculated using the following equation:

THI = $\left(1.8\times T_{avg}+32\right)-\left(0.55-0.0055\times R{H}_{Avg}\right)\times \left(1.8\times T_{avg}-26\right)$(NRC, 1971), where $T_{avg}$ is the average air temperature (°C), and $R{H}_{Avg}$ is the average relative humidity (%). THI values were evaluated at integer thresholds between 70 and 80. The THI range of 70–80 was evaluated because it encompassed more than 95% of observations recorded during the study period. Observations outside this range were infrequent and therefore insufficient for robust threshold estimation. As a result, the selected interval adequately represented the thermal conditions experienced by the study population under tropical production environments. The optimal THI threshold was determined based on the model fit criteria, including the highest coefficient of determination (R²) and lowest mean squared error (MSE) using SAS version 9.0 (SAS Institute Inc., Cary, NC, USA).

Genetic parameter estimation

A reaction norm repeatability test-day model was used to estimate the heat-stress threshold, variance components, genetic parameters (heritability, genetic correlations, and permanent environmental correlations), and the decline in monthly egg production associated with increasing THI. Variance components were estimated using the Average Information Restricted Maximum Likelihood (AI-REML) algorithm implemented in the AIREMLF90 program of the BLUPF90+ software suite (version 2.56; Misztal et al., 2024). Convergence was declared when the change in log-likelihood between successive iterations was less than 10⁻¹². A maximum of 5,000 iterations was permitted, although convergence was achieved before reaching this limit in all analyses. The model was specified as follows:

$$y_{ijklm}=H{G}_{ij}+MI{E}_k+AF{E}_l+\alpha \left(THI\right)+a_{0m}+a_{1m}\left[f\left(THI\right)\right]+p_{0m}+p_{1m}\left[f\left(THI\right)\right]+e_{ijklm}$$

where $y_{ijklm}$ is the individual monthly egg production record of hen (m), $H{G}_{ij}$ is the fixed effect of the i^th hatch and j^th generation class, $MI{E}_k$ is the fixed effect of the k^th months-in-egg class, $AF{E}_l$ is the linear covariate effect of age at first egg, $\alpha \left(THI\right)$ is the fixed regression coefficient describing the average change in monthly egg production per unit increase in THI, $a_{0m}$ is the random additive genetic intercept of hen (m), $a_{1m}$ is the random additive genetic slope representing genetic sensitivity to heat stress, $p_{0m}$ is the random permanent environmental intercept of hen (m), $p_{1m}$ is the random permanent environmental slope representing permanent environmental sensitivity to heat stress, $\left[f\left(THI\right)\right]$ is the heat-stress function, and $e_{ijklm}$ is the random residual error.

The heat-stress function was defined as:

$$f\left(THI\right)=\{\begin{array}{c}0;THI\le TH{I}_{threshold}\left(noheatstress\right)\\ THI-TH{I}_{threshold};THI>TH{I}_{threshold}\left(heatstress\right)\end{array}$$

The corresponding variance-covariance structure of the random effects was:

$$\text{Var}\left(\begin{array}{c}{\mathbf{a}}_0\\ {\mathbf{a}}_1\\ {\mathbf{p}}_0\\ {\mathbf{p}}_1\\ \mathit{\text{e}}\end{array}\right)=\left(\begin{array}{ccccc}\mathbf{A}{\sigma }_{a0}^2& \mathbf{A}{\sigma }_{a01}& 0& 0& 0\\ \mathbf{A}{\sigma }_{a01}& \mathbf{A}{\sigma }_{a1}^2& 0& 0& 0\\ 0& 0& \mathbf{I}{\sigma }_{p0}^2& \mathbf{I}{\sigma }_{p01}& 0\\ 0& 0& \mathbf{I}{\sigma }_{p01}& \mathbf{I}{\sigma }_{p1}^2& 0\\ 0& 0& 0& 0& \mathbf{I}{\sigma }_e^2\end{array}\right)$$

where ${\mathbf{a}}_0$ and ${\mathbf{a}}_1$ are vectors of random additive genetic (intercept) and heat-stress sensitivity (slope) effects, assumed to be ${\mathbf{a}}_0\sim \mathrm{N}\left(0,\mathbf{A}{\sigma }_{a0}^2\right),$ ${\mathbf{a}}_1\sim \mathrm{N}\left(0,\mathbf{A}{\sigma }_{a1}^2\right),$ where $\mathbf{A}$ is an additive relationship matrix describing additive genetic relationships among animals, ${\sigma }_{a0}^2$ is the additive genetic variance under thermoneutral conditions, and ${\sigma }_{a1}^2$ is the additive genetic variance in heat sensitivity, respectively; ${\mathbf{p}}_0$ and ${\mathbf{p}}_1$ are vectors of random permanent environmental intercept and heat-stress sensitivity (slope) effects, assumed to be ${\mathbf{p}}_0\sim \mathrm{N}\left(0,\mathbf{I}{\sigma }_{p0}^2\right),$ ${\mathbf{p}}_1\sim \mathrm{N}\left(0,\mathbf{I}{\sigma }_{p1}^2\right)$ where $\mathbf{I}$ is an identity matrix, ${\sigma }_{p0}^2$ is the permanent environmental variance under thermoneutral conditions, and ${\sigma }_{p1}^2$ is the permanent environmental variance in heat sensitivity, respectively; $\mathbf{e}$ is the vector of random residual effects, assumed to be $\mathbf{e}\sim \mathrm{N}\left(0,\mathbf{I}{\sigma }_e^2\right),$ where ${\sigma }_e^2$ is the residual variance.

Heritability ($h^2)$ for each THI value was estimated as follows:

$$h^2\left(TH{I}_i\right)=\frac{{\sigma }_a^2\left(TH{I}_i\right)}{{\sigma }_a^2\left(TH{I}_i\right)+{\sigma }_p^2\left(TH{I}_i\right)+{\sigma }_e^2}$$

where ${\sigma }_a^2\left(TH{I}_i\right)$= ${\sigma }_{a0}^2$+2$f\left(TH{I}_i\right){\sigma }_{a01}$+ $f{\left(TH{I}_i\right)}^2{\sigma }_{a1}^2$, and ${\sigma }_p^2\left(TH{I}_i\right)$= ${\sigma }_{p0}^2$+2$f\left(TH{I}_i\right){\sigma }_{p01}$+ $f{\left(TH{I}_i\right)}^2{\sigma }_{p1}^2,$ and $f\left(TH{I}_i\right)$ is the heat-stress function for each THI value.

Genetic correlations $\left(r_g\right)$ and permanent environmental effect correlations $\left(r_p\right)$ for each THI value were estimated between the intercept and slope components of the reaction norm repeatability test-day model. The intercept represented baseline egg production under thermoneutral conditions, whereas the slope represented sensitivity to heat stress above the THI threshold. These correlations therefore quantify the relationship between production potential and heat-stress sensitivity and were calculated as follows:

$$\begin{array}{c}\hfill r_g\left(\mathit{TH}{I}_i\right)=\frac{{\sigma }_{a01}f\left(\mathit{TH}{I}_i\right)}{\sqrt{{\sigma }_{a0}^2\times {\sigma }_{a1}^{2}f{\left(\mathit{TH}{I}_i\right)}^2}},\\ \hfill r_p\left(\mathit{TH}{I}_i\right)=\frac{{\sigma }_{p01}f\left(\mathit{TH}{I}_i\right)}{\sqrt{{\sigma }_{p0}^2\times {\sigma }_{p1}^2{\sigma }_{a1}^{2}f{\left(\mathit{TH}{I}_i\right)}^2}}\end{array}$$

Heat-tolerance breeding values were estimated from the random additive genetic effect associated with the heat-stress component of the reaction norm repeatability test-day model. These breeding values reflect the genetic ability of individual hens to maintain egg production as THI increases. Animals with positive heat-tolerance breeding values, indicating a smaller decline in egg production above the THI threshold, were considered more heat tolerant, whereas those with negative values were considered less heat tolerant. The proportion of animals with positive heat-tolerance breeding values was subsequently calculated for each breed across different THI levels.

Results

Descriptive statistics

The data structure used for the estimation of variance components and genetic parameters is presented in Table 1. In total, 11,530 animals with records were included, comprising 3,680 black-boned chickens and 7,850 Thai native chickens. The pedigree file consisted of 28,590 animals, including 8,450 black-boned and 20,140 Thai native chickens. The dataset included 136,816 monthly egg production records, with 43,220 records from black-boned chickens and 93,596 from Thai native chickens. The average monthly egg production was 14.08 ± 2.5 eggs/month/bird for black-boned chickens and 15.42 ± 3.0 eggs/month/bird for Thai native chickens. The average annual egg production was 169 ± 3.2 eggs/year/bird in black-boned chickens and 185 ± 4.0 eggs/year/bird in Thai native chickens. The average AFE was 165 ± 24 days for black-boned chickens and 180 ± 29 days for Thai native chickens. Environmental conditions during the study period were characterized by an average air temperature of 28.1 ± 2.4°C and RH of 77.5 ± 9.0%.

Table 1.

Data structure for estimation of variance components and genetic parameters in black-boned and Thai native chickens.

Categories	Total	Chicken breed
		Black-boned	Thai native
Animals with records (n)	11,530	3,680	7,850
Animals with pedigrees (n)	28,590	8,450	20,140
Number of records (n)	136,816	43,220	93,596
Average records per hen (n)	11.87	11.74	11.92
Average monthly egg production (eggs/month/bird ± SD)		14.08 ± 2.5	15.42 ± 3.0
Average annual egg production (eggs/year/bird ± SD)		169 ± 3.2	185 ± 4.0
Average age at first egg (days ± SD)		165 ± 24	180 ± 29
Average air temperature (°C ± SD)	28.1 ± 2.4	-	-
Average relative humidity (% ± SD)	77.5 ± 9.0	-	-

Testing of thermal stress onset

The variation in model fit statistics, including the coefficient of determination (R²) and MSE, across THI levels in black-boned and Thai native chickens is presented in Table 2. In black-boned chickens, the R² values increased slightly from 0.334 at THI 70 to a maximum of 0.340 at THI 72, followed by a gradual decline to 0.325 at THI 80. Correspondingly, the MSE values decreased from 7.430 at THI 70 to their lowest value of 7.424 at THI 72 and subsequently increased to 7.448 at THI 80. Based on these criteria, THI 72 was identified as the point at which the model fit was optimized, as indicated by the highest R² and lowest MSE. In Thai native chickens, the R² values increased from 0.405 at THI 70 to a peak of 0.421 at THI 74, followed by a consistent decline to 0.335 at THI 80. Similarly, the MSE values decreased from 9.489 at THI 70 to the lowest value of 9.470 at THI 74, and then increased to 9.497 at THI 80. Accordingly, THI 74 was identified as the optimal threshold based on the model fit statistics. The onset THI threshold differed between breeds, with a lower threshold observed in black-boned chickens (THI 72) than in Thai native chickens (THI 74).

Table 2.

Onset of heat stress in black-boned and Thai native chickens.

Temperature–humidity index	Chicken breed
	Black-boned chicken		Thai native chicken
	R2	MSE	R2	MSE
70	0.334	7.430	0.405	9.489
71	0.337	7.427	0.409	9.482
72	0.340	7.424	0.415	9.475
73	0.339	7.425	0.419	9.473
74	0.337	7.427	0.421	9.470
75	0.335	7.428	0.420	9.471
76	0.335	7.430	0.419	9.473
77	0.334	7.433	0.416	9.478
78	0.333	7.437	0.411	9.484
79	0.330	7.441	0.402	9.489
80	0.325	7.448	0.335	9.497

R²: coefficient of determination; MSE: mean squared error.

Genetic parameter estimates

The variance–covariance components and genetic parameters ±SE estimated from the reaction norm repeatability test-day model are presented in Table 3. Under thermoneutral conditions, the additive genetic variance (${\sigma }_{a0}^2$) was slightly higher in Thai native chickens (4.806 ± 0.163) than in black-boned chickens (4.399 ± 0.127). In contrast, the additive genetic variance associated with heat-stress sensitivity (${\sigma }_{a1}^2$) was greater in black-boned chickens (2.560 ± 0.577) than in Thai native chickens (1.070 ± 0.319), indicating greater genetic variability in response to increasing heat load in the black-boned population. The additive genetic covariance between intercept and slope (${\sigma }_{a01}$) was negative in both breeds, resulting in unfavorable genetic correlations between thermoneutral production potential and heat-stress sensitivity (r_g = −0.289 ± 0.054 and −0.172 ± 0.057 for black-boned and Thai native chickens, respectively). Similarly, permanent environmental covariances between intercept and slope were negative in both populations, with stronger permanent environmental correlations observed in Thai native chickens (r_p = −0.478 ± 0.086) than in black-boned chickens (r_p = −0.244 ± 0.071). Heritability estimates were low to moderate and were slightly higher in Thai native chickens (0.158 ± 0.004) than in black-boned chickens (0.135 ± 0.005), suggesting the presence of exploitable additive genetic variation for egg production under varying thermal environments.

Table 3.

Variance-covariance components and genetic parameters for the intercept (thermoneutral production potential) and slope (heat-stress sensitivity) effects estimated from the reaction norm repeatability test-day model in black-boned and Thai native chickens.

Parameters	Black-boned chicken	Thai native chicken
${\sigma }_{a0}^2$	4.399±0.127	4.806±0.163
${\sigma }_{a01}$	−0.970±0.452	−0.390±0.232
${\sigma }_{a1}^2$	2.560±0.577	1.070±0.319
$r_g$	−0.289±0.054	−0.172±0.057
${\sigma }_{p0}^2$	5.500±0.215	4.918±0.171
${\sigma }_{p01}$	−1.532±0.952	−3.101±1.241
${\sigma }_{p1}^2$	7.150±2.988	8.559±3.086
${\sigma }_e^2$	18.000±0.000	18.100±0.000
$r_p$	−0.244±0.071	−0.478±0.086
$h^2$	0.135±0.005	0.158±0.004

${\sigma }_{a0}^2$ = additive genetic variance under thermoneutral conditions (intercept); ${\sigma }_{a1}^2$ = additive genetic variance associated with heat-stress sensitivity (slope); ${\sigma }_{a01}$ = additive genetic covariance between intercept and slope; ${\sigma }_{p0}^2$ = permanent environmental variance under thermoneutral conditions (intercept); ${\sigma }_{p1}^2$ = permanent environmental variance associated with heat-stress sensitivity (slope); ${\sigma }_{p01}$ = permanent environmental covariance between intercept and slope; ${\sigma }_e^2$ = residual variance; r_g = genetic correlation between intercept and slope; r_p = permanent environmental correlation between intercept and slope; and h² = heritability estimated from the reaction norm model.

Estimated heritability, genetic and permanent environmental correlations across THI from 70 to 80 are presented in Fig. 1. The heritability estimates of monthly egg production across THI levels are presented in Fig. 1A. In both breeds, heritability decreased as THI increased. In black-boned chickens, estimates declined from 0.162 at THI 70 to 0.107 at THI 80, whereas in Thai native chickens, they decreased from 0.175 to 0.132 over the same range. Thai native chickens consistently showed slightly higher heritability estimates than black-boned chickens across THI levels. The genetic correlations between monthly egg production and heat stress across THI levels are shown in Fig. 1B. In black-boned chickens, correlations decreased from 0.350 at THI 70 to −0.723 at THI 80, with positive values at THI 70–72 and negative values from THI 73 onward. In Thai native chickens, correlations declined from 0.500 at THI 70 to −0.524 at THI 80, remained positive up to THI 74, and became negative thereafter. A more gradual decline was observed in Thai native chickens than in black-boned chickens. The permanent environmental correlations between monthly egg production and heat stress across THI levels are presented in Fig. 1C. In black-boned chickens, correlations decreased from 0.420 at THI 70 to −0.583 at THI 80, shifting from positive values at THI 70–72 to negative values from THI 73 onward. In Thai native chickens, correlations decreased from 0.620 at THI 70 to 0.250 at THI 74 and became negative from THI 75 onward, reaching −0.774 at THI 80.

Fig. 1.

(A) Heritability±SE estimates of monthly egg production; (B) genetic correlation±SE between monthly egg production and heat stress; and (C) permanent environmental correlations±SE between monthly egg production and heat stress in black-boned (blue lines) and Thai native (orange lines) chickens across THI levels between 70 and 80.

Changes in monthly egg production across THI values

Estimated change in monthly egg production (eggs/bird/month) in black-boned (blue bars) and Thai native (red bars) chickens derived from the reaction norm repeatability test-day model at THI levels between 70 and 80 is presented in Fig. 2. In black-boned chickens, the estimated change in monthly egg production decreased from +1.01 eggs/bird/month at THI 70 to −2.03 eggs/bird/month at THI 80. Positive estimates were observed at THI 70–71, indicating that egg production was maintained or slightly improved under these environmental conditions. In contrast, negative estimates were observed from THI 72 onward, indicating progressive reductions in monthly egg production as THI increased beyond the identified heat-stress threshold. In Thai native chickens, the estimated change in monthly egg production decreased from +1.77 eggs/bird/month at THI 70 to −1.22 eggs/bird/month at THI 80. Positive estimates were maintained from THI 70 to 73, whereas negative estimates occurred from THI 74 onward. In general, egg production follows a biological trajectory characterized by an increase after sexual maturity, a peak production period, and a gradual decline with advancing age. Consequently, the observed production patterns likely reflect the combined effects of age-related changes and environmental conditions. However, egg production consistently decreased once THI exceeded the breed-specific threshold, indicating that heat stress exerted an additional negative effect beyond the normal age-related decline in production.

Fig. 2.

Estimated change in monthly egg production (eggs/bird/month) across THI levels in black-boned (blue bars) and Thai native (red bars) chickens derived from the reaction norm repeatability test-day model at THI levels between 70 and 80. Positive values indicate maintenance or improvement of production, whereas negative values indicate reductions in monthly egg production associated with increasing THI.

Proportion of animals with positive heat-tolerance breeding values

The proportion of animals with positive breeding values for heat tolerance in monthly egg production across THI levels is presented in Fig. 3. In black-boned chickens, the proportion declined from 95.77% at THI 70 to 5.58% at THI 80. The most pronounced reduction occurred between THI 70 and 73, decreasing from 95.77% to 32.40%, followed by a more gradual decline at higher THI levels. In Thai native chickens, the proportion decreased from 99.22% at THI 70 to 15.16% at THI 80. Compared with black-boned chickens, Thai native chickens maintained a higher proportion of animals with positive breeding values for heat tolerance across all THI levels. More than 50% of animals remained heat tolerant up to THI 74, after which the proportion declined progressively.

Fig. 3.

Proportion of animals with positive breeding values for heat tolerance for monthly egg production in black-boned (blue bars) and Thai native (red bars) chickens at THI levels between 70 and 80.

Discussion

In livestock breeding, resilience is generally defined as an animal’s ability to maintain or recover performance following environmental or physiological disturbances, whereas robustness refers to the capacity to sustain acceptable levels of production, reproduction, health, and welfare across diverse environmental conditions. In the present study, resilience was assessed as the ability of hens to maintain egg production under increasing THI levels, whereas robustness was considered a broader adaptive attribute reflecting overall environmental tolerance. Our results revealed genotype-dependent differences in egg production and resilience to heat stress between black-boned and Thai native chickens reared under tropical conditions. Thai native chickens exhibited higher egg production than black-boned chickens; however, the performance of both breeds remained below that of commercial layers, highlighting the trade-off between productivity and environmental adaptability commonly observed in indigenous chicken populations (Wolc et al., 2012; Padhi, 2016; Mokoena et al., 2025; Loengbudnark et al., 2024). Notably, although black-boned chickens reached age at first egg (AFE) earlier than Thai native chickens, this advantage did not translate into greater cumulative egg production. This finding suggests that production persistence and tolerance to environmental stress may have a greater influence on long-term laying performance than early sexual maturity under field conditions.

Distinct THI thresholds (72 for black-boned chickens and 74 for Thai native chickens) indicate differences in heat stress sensitivity. Although the numerical difference is small, it may reflect biologically meaningful variation in thermal tolerance (Lara and Rostagno, 2013; Nawab et al., 2018; Oke et al., 2024). The earlier decline observed in black-boned chickens suggests a lower capacity to maintain physiological homeostasis under combined heat and humidity stress, whereas the delayed response in Thai native chickens may reflect long-term adaptation through more efficient thermoregulation and greater tolerance to oxidative stress (Aryal et al., 2025). Nevertheless, the identified THI thresholds should be interpreted with caution because the differences in model fit statistics (R² and MSE) among adjacent THI levels were relatively small. Although threshold selection was based on objective model-fitting criteria, the chosen values were further supported by concurrent changes in genetic and phenotypic responses to heat stress, including reductions in heritability, alterations in genetic correlations, and declines in egg production performance. Therefore, these thresholds should be regarded as operational indicators of heat stress rather than definitive biological breakpoints. Future studies employing alternative analytical approaches, such as segmented regression, breakpoint analysis, or generalized additive models, may provide a more detailed characterization of the nonlinear nature of heat stress responses and further improve threshold estimation.

Heritability declines progressively with increasing THI, indicating a reduced contribution of additive genetic variance to phenotypic expression under heat stress (Loengbudnark et al., 2023; Boonkum et al., 2024). This pattern is most plausibly explained by a disproportionate inflation of environmental variance under thermal challenge, rather than an absolute reduction in additive genetic variance. An elevated THI imposes multifactorial physiological stress on laying hens, including reduced feed intake, altered nutrient partitioning, endocrine disruption of the hypothalamic–pituitary–gonadal axis, oxidative stress, acid–base imbalance due to panting, and electrolyte disturbances. These responses collectively impair metabolic homeostasis, reproductive function, and eggshell formation, ultimately leading to reduced egg production and quality (Kim et al., 2024; Olayiwola and Odedokum, 2025; Tesakul et al., 2025; Abd El-Hack et al., 2026), all of which contribute to increased residual and permanent environmental variation (Boonkum et al., 2024). In parallel, heat-induced oxidative stress and mitochondrial dysfunction can impair cellular energy metabolism, whereas endocrine dysregulation within the hypothalamic–pituitary–gonadal axis reduces follicular growth and ovulatory efficiency (Nawab et al., 2018; Habashy and Adomako, 2023; Yan et al., 2022). These processes collectively increase phenotypic variability independent of the genotype, which diminishes the relative expression of additive genetic effects and reduces heritability estimates. Similar THI-dependent reductions in heritability have been reported in poultry and other livestock species under heat stress conditions (Misztal, 2017; Buranawit et al., 2025).

The observed shift in genetic correlations from positive to strongly negative values across increasing THI levels provides further evidence of genotype-by-environment interactions manifested as re-ranking of genetic merit (Hammami et al., 2009; Reis et al., 2025). At a low THI, favorable alleles associated with higher egg production are expressed effectively; however, under heat stress, these same genotypes may exhibit reduced performance owing to limited physiological robustness (Lara and Rostagno, 2013). This antagonistic relationship suggests that selection decisions based solely on performance under thermoneutral conditions may not translate into genetic improvement under heat stress, and may even be counterproductive in tropical production systems (Misztal, 2017; Boonkum et al., 2024). The steeper decline and more extreme negative correlations observed in black-boned chickens indicate a greater sensitivity of their genetic architecture to environmental perturbation. This suggests a narrower reaction norm and reduced capacity to maintain genetic expression under stress. In contrast, the more gradual transition observed in Thai native chickens reflects the greater stability of genetic effects across environmental gradients, likely due to adaptive mechanisms, such as lower metabolic heat production, improved thermoregulatory efficiency, and enhanced antioxidant defense systems (Wasti et al., 2020; Boonkum et al., 2021; Loengbudnark et al., 2023).

These findings demonstrate that heat stress not only reduces phenotypic performance but also fundamentally alters the structure and reliability of genetic parameters, with direct implications for breeding program design. The decline in heritability combined with antagonistic genetic correlations suggests that conventional selection schemes may yield biased or suboptimal genetic gains when applied across heterogeneous thermal environments (Boettcher et al., 2015; Reis et al., 2025). Incorporating environmental descriptors such as THI into genetic evaluation frameworks—through reaction norm models or environment-specific breeding values—represents a critical step toward improving selection accuracy and achieving sustainable genetic progress under climate variability (Misztal, 2017; Boonkum et al., 2024).

The nonlinear decline in egg production across THI levels further supports threshold-dependent responses rather than simple linear relationships. The transition from positive to negative production rates marks a critical point beyond which compensatory mechanisms are insufficient to maintain egg production performance (Lara and Rostagno, 2013; He et al., 2018). The steeper decline observed in black-boned chickens, particularly at higher THI levels, suggests a lower capacity to buffer environmental stress. This may be associated with the reduced heat dissipation efficiency and greater susceptibility to oxidative stress and metabolic imbalance under elevated temperature and humidity conditions. Heat stress is known to increase reactive oxygen species production, impair mitochondrial function, and disrupt endocrine regulation, ultimately reducing reproductive efficiency (Nawab et al., 2018; Habashy et al., 2023). In black-boned chickens, these constraints may occur at lower THI thresholds or progress more rapidly, resulting in a sharp decline in egg production.

In contrast, Thai native chickens maintained more stable production across a broader THI range, which is consistent with their long-term adaptation to tropical environments. Indigenous chickens are generally characterized by lower metabolic heat production, improved heat dissipation, and enhanced tolerance to oxidative stress, which contribute to greater resilience under thermal challenge (Wasti et al., 2020; Boonkum et al., 2021). However, the present results indicate that adaptive capacity varies among indigenous genotypes. Although black-boned chickens are also considered indigenous, their greater sensitivity to increasing THI suggests differences in their genetic background or physiological regulation (Duangjinda et al., 2017; Rachman et al., 2024). This highlights the fact that resilience is not uniform across local breeds and requires genotype-specific evaluations. Although increasing THI was associated with reduced egg production, laying performance is also influenced by age-related physiological changes and seasonal variation. Egg production typically increases after sexual maturity, reaches a peak during early lay, and gradually declines with advancing age. Therefore, some of the variation in monthly egg production may reflect normal biological processes rather than heat stress alone. However, the consistent decline in production beyond the identified THI thresholds, together with concurrent changes in genetic parameters, suggests that heat stress was a major factor contributing to the observed reduction in performance. Future studies incorporating age-specific production curves and seasonal covariates would help disentangle the relative contributions of physiological aging and environmental stress.

The proportion of animals classified as heat tolerant was consistently higher in Thai native chickens than in black-boned chickens. These findings indicate that Thai native chickens possess greater population-level resilience to heat stress. In contrast, the decline observed in black-boned chickens suggests greater susceptibility to thermal stress, potentially reflecting less effective thermoregulatory capacity and a reduced ability to cope with heat-induced physiological challenges. Consequently, a smaller proportion of black-boned chickens maintained favorable breeding values for heat tolerance as THI increased, resulting in greater variability in performance and a more pronounced decline in productivity.

From a management perspective, this genotype may benefit from targeted heat-mitigation strategies, including improved ventilation, shade provision, and evaporative cooling systems (Anarbaev et al., 2020; Saleeva et al., 2020; Souza et al., 2025). Nutritional interventions, such as antioxidant and electrolyte supplementation, may further support physiological stability under heat stress (Surai, 2020; Olayiwola and Adedokun, 2025). From a breeding perspective, greater emphasis should be placed on selecting heat-tolerant individuals within this population. Incorporating THI-based or reaction norm approaches into genetic evaluation programs may further enhance selection for thermal resilience and improve adaptation to increasingly challenging environmental conditions.

Because egg production follows an age-dependent trajectory throughout the laying cycle, periods of elevated THI may coincide with specific stages of production. However, the present dataset encompassed multiple generations and production periods, resulting in heat exposure across a wide range of laying stages rather than being confined to a particular phase. In addition, production stage was included in the statistical analyses, reducing potential confounding between age-related changes in egg production and environmental stress. Nevertheless, some overlap between seasonal conditions and the natural progression of egg production cannot be completely excluded and should be considered when interpreting the results.

Although THI was the primary environmental indicator evaluated in this study, other seasonal factors may also affect egg production. Birds were maintained under natural photoperiod conditions; therefore, seasonal variation in day length may have influenced reproductive activity and laying performance. However, because the study was conducted in a tropical region where annual fluctuations in day length are minimal, the influence of photoperiod was likely less pronounced than that of thermal stress. Future studies incorporating both climatic and photoperiodic variables would provide a more comprehensive understanding of environmental influences on egg production and heat-stress responses in indigenous chicken populations.

From a breeding perspective, incorporating THI into genetic evaluation enables the estimation of environment-specific breeding values and the development of selection indices prioritizing performance under heat stress. This approach facilitates the identification of individuals with stable genetic merit across thermal gradients and supports improved resilience in tropical poultry production systems. Methodologically, combining a test-day model with continuous THI gradients provides a more realistic representation of environmental variation than discrete classifications, and allows the evaluation of dynamic genetic responses. However, the THI range examined in this study (70–80) reflects the predominant conditions in the dataset and may not capture extreme heat scenarios. In addition, genotype-by-environment interaction was inferred rather than explicitly modeled, representing a limitation to be tackled by future research.

In conclusion, our results demonstrate that heat stress substantially alters both the phenotypic performance and genetic architecture of monthly egg production trait. The observed genotype-specific responses highlight the importance of incorporating environmental descriptors, such as THI, into genetic evaluation models. Selection strategies that account for environmental sensitivity, rather than relying solely on performance under optimal conditions, are essential for improving resilience and ensuring sustainable egg production under tropical and climate-variable conditions.

CRediT authorship contribution statement

Doungnapa Promket: Writing – review & editing, Writing – original draft, Validation, Project administration, Methodology, Funding acquisition, Formal analysis, Conceptualization. Khanitta Pengmeesri: Writing – review & editing, Writing – original draft, Validation, Conceptualization. Vibuntita Chankitisakul: Writing – review & editing, Writing – original draft, Validation, Conceptualization. Wuttigrai Boonkum: Writing – review & editing, Writing – original draft, Validation, Supervision, Software, Resources, Methodology, Formal analysis, Data curation, Conceptualization.

Disclosures

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.