Introduction

Beef production plays a critical role in the global economy, food systems, and human nutrition. The industry supports more than 24 million jobs worldwide in various sectors, including farming, processing, and distribution1. In 2019, global demand for beef reached 70 million tons, and it is expected to increase further, since beef is a key source of high-quality protein2, which is necessary for muscle growth, tissue repair, and overall bodily function. The United States (U.S.) produces more than 12 million tons of beef per year, which represents 17% of global production, followed by Europe (15%), Brazil (13%), China (9%), Argentina, India (4%), and Australia (4%)3. In the U.S., the beef industry contributes $167.0 billion annually to the economy, which represents ca. 17% of total agricultural gross income1.

In 2023, the U.S. had ca. 95 million head of cattle and calves (NASS, 2024), with more than one-third of the continental U.S. land area used for beef production4. Beef production is particularly prevalent in the Southern Plains, including the states of Kansas, Oklahoma, and Texas, where it is a major part of the regional culture and economy, contributing $26.6 billion annually5. However, this region is water-limited, with mean evaporation rates that exceed mean precipitation5. Due to its high climate variability, the Southern Plains experiences large fluctuations in beef production4. Climate variability there is projected to increase, with increasing temperatures and droughts likely to decrease beef productivity and profitability6,7,8.

One of the key challenges associated with these climatic changes is heat stress, which occurs when high air temperatures (T) and relative humidity (RH) make it difficult for animals to stay cool. This stress can worsen their health and decrease productivity in several ways9, such as by worsening the metabolic and health status, making breathing more difficult, increasing mortality, and decreasing reproductive performance10,11,12. T is a crucial factor, since T that exceeds certain thresholds can decrease production greatly13. Ideal T for livestock ranges from 17 to 24 °C, but higher RH can decrease these thresholds14. Consequently, RH strongly influences an animal’s ability to dissipate heat through evaporative cooling via the skin and lungs15. In contrast, animals can tolerate higher T when RH is low because they can efficiently release excess heat through sweating. Other weather factors such as wind speed (WS) and solar radiation (RAD) also influence heat stress in animals16,17. Therefore, understanding and managing these weather extremes through long-term historical datasets is crucial for optimizing agricultural systems.

The state of Oklahoma, a part of the Southern Plains region of the U.S., is an ideal microcosm for studying patterns of and trends in heat stress that influence cattle production. Rangeland cow-calf production systems and the grazing of stocker cattle on winter wheat are crucial components of Oklahoma’s economy. Beef production contributes ca. $2.7 billion annually (ODAFF, 2024; Panyi et al., 2023), and the state ranks second in the U.S. in the number of beef cattle (ODAFF, 2024). However, climate change may lead to more frequent extreme conditions that exceed critical physiological thresholds for cattle productivity in Oklahoma. For instance, climate projections predict a significant increase in T by 1.7-2.1oC over the next few decades (Lee et al. 2025). In this regard, the state’s unique network of Mesonet stations19 allows for high-resolution spatio-temporal analyses, which can provide valuable insights into regional impacts.

The main objective of this study was to determine patterns of and trends in critical weather conditions that influence beef production systems in Oklahoma. The temperature humidity index (THI)20 and comprehensive climate index (CCI)21 were used to determine critical thresholds for cattle production. THI provides a heat-focused index by considering T and RH, while CCI indicates both heat and cold stress of livestock by considering T, RH, WS, and RAD. We calculated the indices using high-quality weather data from Mesonet stations across Oklahoma. As an assessment endpoint, we analyzed production statistics at both state and county levels to assess impacts of heat stress on cattle production.

Methods

Study area

Oklahoma has a wide range of climates: a humid subtropical climate predominates in the east, while a semi-arid climate predominates in the west (Lee et al. 2024). Its main land use is grassland/pasture (42%), followed by forest (24%), cropland (17%), shrubland (6%), and developed areas (5%)23 (Fig. 1a). Cattle and calves are the most important agricultural commodities, contributing 48% of the state’s gross income from agricultural commodities, and winter wheat, the dominant crop often used for grazing, occupies 11% of the total land use23. The present study used daily weather data (i.e., T, RH, WS, and RAD) collected at 121 Mesonet stations across the state (Fig. 1b). The Mesonet stations provide consistent spatial coverage of relevant data, with at least one station in every county. Based on these data, mean daily T ranged from 13oC in the north to 17oC in the south (Fig. 1b), while mean annual precipitation ranged from 400 mm in the west to 1400 mm in the east (Lee et al. 2024).

Fig. 1
figure 1

Maps of (a) land use and land cover and (b) Mesonet stations (black dots) and mean daily temperature from 1998–2022 across Oklahoma, United States.

Full size image

Temperature humidity index

THI aggregates T and RH to indicate the degree of discomfort that animals experience in warm weather. It has been widely used due to its minimal input requirements, since T and RH are easily available for most regions24. According to Thom20, THI is calculated as follows:

$$\:THI=0.8\times\:T+\left(\frac{RH}{100}\right)\times\:\left(T-14.3\right)+46.4$$
(1)

where T is the dry-bulb T (oC) and RH is expressed in percentage.

Several studies identified THI thresholds at which heat stress begins, and they vary among livestock species8. We used the THI thresholds defined for cattle by Valente et al.25 (Table 1), since cattle are the main livestock species in Oklahoma. Daily THI was calculated for the 121 Mesonet stations from 1998 to 2022 using the maximum daily temperature for T, and then divided into three categories: moderate, high, or extreme (Table 1). The number of days per year that exceeded each threshold was counted for each of the 121 stations to provide a spatial pattern of heat stress based on THI across Oklahoma.

Table 1 Ranges of heat stress categories in cattle based on the temperature humidity index (THI)25.
Full size table

Comprehensive climate index

CCI aggregates T, RH, WS, and RAD to assess an animal’s comfort level by considering both heat and cold stress throughout the year, which makes it particularly useful in areas with four distinct seasonal climates24. CCI is valid for T from ca. -30 to 45 °C and adjusts T using correction factors for RH, WS, and RAD (RHC, WSC, and RADC, respectively). RHC is derived from an exponential relationship between T and RH, using 30% RH as a baseline. At high T (e.g., 45 °C), increasing RH from 30 to 100% raises T by 16 °C, while at low T (e.g., -30 °C), the same increase in RH lowers T by 3 °C. This effect highlights how RH enhances both heat stress and cold stress. RAD enhances heat stress and mitigates cold stress, while WS mitigates heat stress. The Oklahoma Mesonet uses CCI to estimate the degree of cattle comfort. RHC, WSC, and RADC are calculated as follows:

$$RHC = \:{e}^{(0.00182\times\:RH+1.8\times\:{10}^{-5}\times\:T\times\:RH)}\times\:(0.000054\times\:{T}^{2}+0.00192\times\:T-0.0246)\times\:(RH-30)$$
(2)
$$WSC = \:\left[ {\frac{{ - 6.56}}{{e^{{\left\{ {\frac{1}{{\left( {2.26 \times \:WS + 0.23} \right)^{{0.45}} }} \times \:(2.9 + 1.14 \times \:10^{{ - 6}} \times \:WS^{{2.5}} - {\text{log}}_{{0.3}} \left( {2.26 \times \:WS + 0.33} \right)^{{ - 2}} } \right\}}} }}} \right] - 0.00566 \times \:WS^{2} + 3.33\:$$
(3)
$$RADC = \:0.0076\times\:RAD-0.00002\times\:RAD\times\:T+0.00005\times\:{T}^{2}\times\:\sqrt{RAD}+0.1\times\:T-2$$
(4)

with WS expressed in m/s and RAD expressed in W/m2.

Using these correction factors, CCI equals T + RHC + WSC + RADC. See Mader et al.21 for details of CCI. Since the Oklahoma Mesonet has provided CCI estimates since 2008, daily maximum and minimum CCI estimates were extracted for the period 2008–2022 and divided into five categories (Table 2), and then the number of days per year in the Heat Danger (HD) and Cold Danger (CD) categories were counted.

Table 2 Heat and cold stress categories of the mesonet comprehensive climate index (CCI)21.
Full size table

Trend analysis

The number of days per year that exceeded the heat stress threshold, and for CCI, cold stress threshold, were used to analyze long-term (i.e., annual) trends. To detect these trends, the Sen’s slope non-parametric test26 was used to quantify linear trends, which were considered significant at p < 0.05 based on the Modified Mann-Kendall test27. Trends were analyzed using the pyMannKendall package in Python28.

Comparison of trends in indices, the cattle and calf inventory, and gross income from cattle production

Annual inventory data for cattle and calves, available for Oklahoma’s 77 counties since 1975 (NASS, 2024), were used to examine relations between heat stress based on CCI and the inventory at the state and county levels. Since the inventory data are recorded in January each year, inventory data for the year after a given year’s heat stress were examined. In addition, annual gross income data for cattle and calves were also available at the state level and used to assess economic impacts of trends in the inventory.

Results and discussion

Spatial and temporal patterns based on THI

Based on the THI thresholds, the mean number of days of “extreme” heat stress (THI > 94) was negligible, with ca. 95% of the stations in Oklahoma having less than 1 day per year. Therefore, the “extreme” and “high” heat stress categories (THI > 82) were used to assess heat stress patterns across Oklahoma. Note that the THI thresholds defined (Table 1), while appropriate for a particular production system, are not universally applied, since other studies have used different thresholds. Despite these differences, THI has been used widely as a standard for decades, providing a valuable basis for comparative analysis across studies. The mean number of “high” heat stress days per year was 96 (range: 51–119) (Fig. 2), of which ca. 80% occurred in summer (23% in June, 30% in July, and 27% in August), followed by September and May. The southern part of the state had more heat stress days than the northern part, which closely aligned with the state’s spatial pattern of T (Fig. 1b).

Fig. 2
figure 2

Spatial pattern of the number of days per year in the “extreme” and “high” heat stress categories based on the temperature humidity index (THI) from 1998–2022 in Oklahoma.

Full size image

Spatial and temporal patterns based on CCI

Eliminating stations that lacked CCI estimates from 2008 to 2022 decreased the number of stations to 102. Because the mean number of days per year of CD during the study period was negligible (range: 0.1–1.5), we focused on HD, which had a mean of 54 days per year (range: 29–81) (Fig. 3). Compared to the “extreme” and “high” heat stress based on THI, the HD category for CCI had fewer days of heat stress since it is the highest category. The spatial pattern of HD days was similar to that of the heat stress based on THI, since the southern part of the state had the most heat stress days, while the Panhandle region had the fewest. Based on CCI, the summer contained 87% of HD days (16% in June, 37% in July, and 31% in August), which was a relatively higher percentage than that for heat stress based on THI. These results confirmed that T strongly influences estimated heat stress, since the spatial patterns of THI and CCI were similar to that of T. However, their spatial patterns differed in central Oklahoma, in part because CCI also considers WS and RAD.

Fig. 3
figure 3

Spatial patterns of the number of days per year in the Heat Danger category based on the comprehensive climate index (CCI) from 2008–2022 across Oklahoma.

Full size image

Trends in heat stress based on THI and CCI

Based on THI, all stations showed positive increasing trends the number of days per year in the “extreme” and “high” heat stress categories from 1998 to 2022, but only 14 of the trends were significant, most of them in the southern part of the state (Fig. 4). Since T is the main factor used to calculate THI, we examined whether T showed any significant trends from 1998 to 2022 by calculating the mean daily maximum T (\(\:{T}_{max}\)) in summer (i.e., June, July, and August) each year and analyzing annual trends in them for each station. The analysis showed that only one station showed a significant increasing trend in mean \(\:{T}_{max}\) in summer, which can explain the relatively small percentage of significant increasing trends in heat stress days based on THI.

Fig. 4
figure 4

Sen’s slope annual trends in the number of days per year in the “extreme” and “high” heat danger categories based on the temperature humidity index (THI) from 1998–2022 across Oklahoma. Trends were considered significant at p < 0.05.

Full size image

In contrast, based on CCI, more than 60% of stations (62 of 102) showed significant increasing trends in the number of days per year in the HD category from 1998 to 2022 (Fig. 5). Not only did more stations show significant trends when heat stress was based on the CCI rather than the THI, but Sen’s slope was also higher. According to Sen’s slope, the number of days per year in the HD category across the state increased from 1 to 4, which can worsen cattle health and decrease productivity greatly. Since maximum \(\:{T}_{max}\) showed nearly no trend in summer, the factors not considered in THI, such as WS and RAD, can influence HD greatly24. Thus, we expanded the analysis to examine whether RH, WS, and RAD showed any trends from 1998 to 2022.

Fig. 5
figure 5

Sen’s slope annual trends in the number of days per year in the Heat Danger category based on the comprehensive climate index (CCI) from 2008–2022 across Oklahoma. Trends were considered significant at p < 0.05.

Full size image

Like \(\:{T}_{max}\), mean RH, WS, and RAD were analyzed for summer each year, when most heat stress occurred. For RH, all but 1 of the 121 stations showed an increasing trend, of which 16 were significant, which can increase heat stress. Notably, for WS, 50 stations showed significant decreasing trends across most of the state (Fig. 6), which suggested that they strongly increased heat stress based on the CCI. This argument was supported by the observation that only 5 stations showed significant trends in RAD, which indicated that the significant decreases in WS and slightly increasing trends in RH in summer led to more heat stress days based on CCI across the state.

Fig. 6
figure 6

Sen’s slope annual trends in wind speed (WS) (m/s) from 1998–2022 across Oklahoma. Trends were considered significant at p < 0.05.

Full size image

Relationships between heat stress and the cattle and calf inventory and gross income

The cattle and calf inventory in Oklahoma decreased from 6.5 million head in 1975 to 5.2 million head in 2022. At the state level, 2022 and 2011 had the largest mean number of HD days per year based on CCI (83 and 76, respectively) (Table 3). In 2023 and 2012, the number of cattle and calves decreased greatly, from 5.2 to 4.5 million and from 5.2 to 4.6 million head, respectively, which were the two largest changes in the inventory during the study period (Table 3). Except for these two peak years, however, there was no general pattern between the number of HD days and the inventory, which suggests that there may be a critical threshold (e.g., effects combined with those of droughts and/or consecutive days of heat stress) that influences cattle production greatly. With increasing trends in HD and impacts of climate change that can create unfavorable conditions for animals, this threshold may be crossed more frequently, which may decrease cattle production even further in the future.

In 2023 and 2012, like for the inventory data, the largest mean number of HD days per year was accompanied by a decrease in annual gross income from cattle production, even though the gross income showed an increasing trend (Table 3). Although several factors besides heat stress can influence the gross income from cattle production, these results highlight the potential for strong economic impacts of heat stress on cattle production and highlight the importance of implementing strategies to mitigate its effects.

Table 3 Mean annual number of heat danger (HD) days based on the comprehensive climate index (CCI), the cattle and calf inventory, and gross income from cattle and calves in Oklahoma from 2008–2023.
Full size table

At the county level, Texas County, in the Panhandle region, had the highest annual mean inventory of cattle and calves (277,000 head), which represented ca. 5% of the state’s inventory (Fig. 7a). Trend analysis showed that 26 counties showed significant decreasing trends in the inventory (Fig. 7b). Notably, 70% of these counties (18 of 26) showed significant increasing trends in HD based on CCI (Fig. 7b). This result suggests a potential relationship between an increasing number of HD days and decreasing cattle and calf inventory at the county level, which indicates that increased heat stress based on CCI may contribute to fewer cattle. However, since HD days are point-based observations, but inventory data are aggregated at the county level, these two datasets cannot be compared directly due to their difference in scale. In this study, we considered that a county had a significant increasing trend in HD days if it contained a Mesonet station that had the same trend. Using the same criteria, the significant trends in heat stress based on THI (Fig. 4) were not correlated with the trends in the inventory, which highlights the importance of using a comprehensive index to accurately detect heat stress for cattle production.

Fig. 7
figure 7

(a) Annual mean inventory of cattle and calves by county across Oklahoma from 1975–2022, and (b) counties with a significant decreasing trend in inventory of cattle and calves (outlined in red). Counties in gray contained a Mesonet station that showed a significant increasing trend in Heat Danger based on CCI. Counties in red showed a significant increasing trend in Heat Danger and a decreasing trend in the inventory (p < 0.05).

Full size image

Implications

The results based on THI showed fewer stations with significant trends, and those that were significant had a lower Sen’s slope, since THI does not consider changes in WS or RAD, as CCI does. The comparisons with historical data at the county level also indicated that trends in heat stress based on THI did not correlate well with trends in the cattle inventory. In contrast, CCI showed more significant increasing trends in the number of HD days, which showed strong correlations with the cattle inventory and gross income from cattle production. The increasing trends were driven mainly by significant decreasing trends in WS, with slightly increasing trends in RH. The trends in WS and their influence on heat stress are important for the cattle industry. Feedlots are located on the Southern Plains in part because of their breezy conditions in summer. Changes in the magnitude of WS, along with the depletion of the Ogallala aquifer29, could influence cattle production in the region strongly.

Climate change changes the frequency and duration of heatwaves and droughts, which can decrease agricultural productivity30. Given the expected increase in droughts31 and interannual variability in precipitation32 in this region, the increasing trend in heat stress will further challenge beef cattle production. Indeed, comparing heat stress to the inventory and gross income showed that heat stress strongly influences cattle production in this region. The large decreases in the inventory and gross income in 2011 and 2022 indicated that there may be critical threshold conditions that worsen animal health and productivity greatly. When these thresholds are crossed, cattle may experience severe physiological stress, which decreases growth rates and reproductive success, and in extreme cases, increases mortality rates. However, despite the economic importance of cattle production in this region, its vulnerability to these changes has received little attention33.

Heat stress can also decrease the quantity and quality of forage, which can influence animal feeding34. Since winter wheat is often used for both grain production and grazing in Oklahoma35, decreasing winter wheat production due to climate change could further decrease cattle production. Thus, cattle producers and other stakeholders need to develop and implement adaptive strategies to minimize effects of heat stress and other climate factors in this region.

While this study provides insights into impacts of heat stress on cattle production in Oklahoma, its results have broader implications for the U.S. and global cattle industry. Since climate change continues to drive increases in heatwaves that influence agriculture36,37,38, these trends are likely to influence cattle production at a larger scale. Proactive adaptation and mitigation strategies, such as increasing the resilience of forage crops and investing in heat-resistant structures (e.g., shade, airflow in cattle production systems) and cattle breeds (e.g., Bos indicus or Criollo heritage breeds), will be essential for mitigating impacts of heat stress globally. Continued research is needed at both regional and global scales to ensure the sustainability of cattle production in the face of a changing climate.

Conclusion

Both the THI and CCI showed similar spatial patterns of heat stress, with the highest frequencies observed in the southern part of the state. These heat stress patterns followed T patterns across Oklahoma. In addition, most heat stress occurred in summer. Therefore, we conclude that T strongly influences overall temporal and spatial patterns of heat stress. However, for long-term trends in the frequency of heat stress, THI and CCI differed in several key ways. While most stations showed no trends based on THI, more than 60% of stations showed significant increasing trends in heat stress based on CCI. This difference was driven mainly by a significant decrease in WS in summer, a factor that could dramatically increase the region’s vulnerability to heat stress for cattle production.

At the state level, there was a significant decreasing trend in the cattle and calf inventory. Notably, the years with the most HD days (2022 and 2011) also showed significant decreases in the inventory the following year, which suggests that extreme heat stress may worsen cattle health and productivity and that a critical threshold for heat stress may exist at which cattle production decreases greatly. At the county level, ca. 70% of Oklahoma counties showed decreasing trends in the inventory, with significant decreases in 26 of them. Moreover, counties with significant increases in HD based on CCI often experienced inventory decreases, which indicated a potential relationship between heat stress and the cattle inventory. However, trends in heat stress based on THI did not correlate with inventory trends, which highlights the need to use comprehensive heat stress indices to accurately assess impacts of heat stress on cattle production. With an expected increase in T, increase in interannual variability in precipitation, and decrease in WS, these significant increasing trends in heat stress are expected to worsen cattle farming in Oklahoma and other parts of the world. Therefore, incorporating mitigation strategies, such as developing policies that promote the adoption of heat-tolerant infrastructure and breeds, will not only decrease the immediate impacts of heat stress but also contribute to the long-term resilience of cattle production systems under future climate scenarios.