Do Plants Recover from Daytime Heat at Night?—What Bt Cotton Reveals About Plant “Recovery Capacity”

Plant heat-stress experiments often use conditions such as “38°C for 24 hours” or “40°C for 48 hours.” These treatments are experimentally convenient, but they do not fully resemble what plants experience in the field. Even when daytime temperatures reach 38°C in midsummer, that temperature does not usually persist for 24 hours. At night, the air cools.

So how much does this cooler nighttime period matter to a plant?

A study published in Frontiers in Plant Science on August 27, 2026, exposed Bt cotton to 38°C during the day and 27°C at night and followed the amount of Cry1Ac protein in cotton squares for 10 days. The interesting result is that Cry1Ac did not decline immediately, despite exposure to a very high daytime temperature of 38°C.

It remained largely stable through day 4, declined clearly by day 7, and fell further by day 10.

The authors suggest that the relatively moderate nighttime temperature may temporarily compensate for physiological damage accumulated during the hot daytime period.

This makes the paper interesting not only as a study of “how much heat a plant can tolerate,” but from another perspective: how much of the damage received during the day can a plant recover from during the night?

Ten days of 38°C days and 27°C nights

The study used two Bt cotton cultivars, Sikang 1 and Sikang 3. Across two years, 2021 and 2022, plants at the pre-flowering peak square stage were transferred to controlled-environment chambers and compared under the following temperature regimes.

Treatment	Day, 06:00–18:00	Night, 18:00–06:00
Control	32°C	25°C
High temperature	38°C	27°C

The high-temperature treatment continued for up to 10 days. Fifteen-day-old squares were sampled at 11:00 a.m. on days 0, 4, 7, and 10. Daytime light intensity in the chamber was 200 μmol m⁻² s⁻¹, relative humidity was 70%, and 24 pots were used for each cultivar and treatment.

Cry1Ac concentration was measured by ELISA.

Even under the high-temperature treatment, Cry1Ac did not show a significant decline through day 4. By day 7, however, it had decreased, and by day 10 the reduction was larger.

In 2021, Sikang 1 showed reductions of 14.18% on day 7 and 30.88% on day 10 relative to the control. Sikang 3 showed reductions of 11.32% and 25.21%, respectively. The pattern was similar in 2022: Sikang 1 declined by 10.68% and 26.94%, and Sikang 3 by 11.87% and 25.24% on days 7 and 10. The same time-dependent pattern therefore appeared in two cultivars across two years.

However, this should not be interpreted as meaning that “seven days of heat are safe.” The sampling interval jumps from day 4 to day 7. What the experiment directly tells us is that a transition occurred somewhere between days 4 and 7. We cannot tell whether it happened on day 5 or day 6. The authors themselves describe the day 4–7 interval as a critical transition period.

Cry1Ac did not decline in isolation

As the high-temperature treatment continued, soluble protein content and GOT and GPT activities declined. In contrast, free amino acid content and protease and peptidase activities increased.

Cry1Ac concentration was positively correlated with soluble protein, GOT, and GPT, and negatively correlated with free amino acids, protease, and peptidase.

In simplified terms, the changes are consistent with weaker protein maintenance and synthesis on one side and stronger protein degradation on the other. Because Cry1Ac is itself a protein, the authors interpret its decline as part of this broader shift in protein metabolism.

But this point also requires caution. GOT, GPT, protease, and the other measured variables are not direct measurements of the synthesis rate or degradation rate of Cry1Ac itself. Correlation also does not establish causation.

A more precise interpretation is therefore that the decline in Cry1Ac occurred alongside metabolic changes consistent with reduced protein synthesis/maintenance and enhanced protein degradation.

Are the plants really recovering at night?

This is the most important question raised by the paper.

The authors propose that the 27°C nighttime period may compensate for physiological stress caused by 38°C daytime exposure, temporarily maintaining a balance between Cry1Ac synthesis and degradation.

For the first few days, daytime damage may be offset to some extent by nighttime recovery, so the net Cry1Ac level changes little. But if a small unrecovered fraction is carried forward each day, the burden can accumulate. Somewhere between days 4 and 7, cumulative damage may exceed the plant’s nighttime compensatory capacity.

That model—plants recovering at night—is, in my view, the most interesting way to read this study.

However, the study did not directly observe nighttime recovery itself.

Samples were collected only at 11:00 a.m. on days 0, 4, 7, and 10. The researchers did not measure how much the plant deteriorated by evening, how much it recovered during the night, or how close it was to its earlier state by the following morning.

The authors explicitly note that the 11:00 a.m. measurements represent a net outcome after repeated day/night cycles rather than real-time nighttime dynamics. In the conclusion, they also acknowledge that nighttime compensation is a hypothesis supported by the metabolic data and that direct molecular evidence—for example, measurements of protein turnover or gene expression—is still needed.

So the most accurate statement at this point is: under a fluctuating 38°C-day/27°C-night regime, the high-temperature-associated decline in Cry1Ac was not apparent during the first four days but was evident by day 7; nighttime physiological compensation is proposed as one explanation for that delay.

There is earlier evidence that recovery really occurs

This interpretation is not based on the 2026 study alone. The same research group has examined post-heat recovery in previous work.

In a 2013 study, Bt cotton was exposed to 37°C for either 24 or 48 hours and then returned to 27°C. Cry1Ac recovered relatively quickly after the 24-hour treatment, whereas recovery after 48 hours of heat exposure took longer.

A 2022 study used cotton squares and exposed them to 38°C for 72 or 96 hours before returning them to 27°C. Again, Cry1Ac concentration recovered, and longer heat exposure required a longer recovery period.

In other words, earlier experiments do support the idea that Bt cotton can restore Cry1Ac after heat stress, and that stronger or longer stress increases the time required for recovery.

One detail in the 2026 paper deserves mention. Its Discussion describes the recovery time in the 2013 study as being “inversely proportional” to the duration of heat stress. The original 2013 data show the opposite pattern: longer heat treatment required longer recovery. In context, this appears to be a wording error in the 2026 paper.

There is also a 2018 study from the same group that used Sikang 1 and Sikang 3 and examined boll shells under an alternating 38°C-day/27°C-night regime. This means the 2026 study is not the first demonstration that Cry1Ac decline can be delayed under a fluctuating high-temperature regime. Its contribution is better described as showing a similar time-dependent pattern and associated metabolic changes at the earlier reproductive square stage, across two cultivars and two years.

A little like sleep debt in humans

The pattern loosely brings human sleep to mind.

Plants and human sleep obviously do not share the same mechanism. The useful analogy is more general: biological condition may depend not only on how much stress is received, but also on how much recovery occurs before the next stress arrives.

The plant experiences high temperature during the day. It partially recovers at night. The next day brings heat again. If the recovery is incomplete, a residual burden may gradually accumulate.

Rather than a complete “reset,” it may be more accurate to imagine the plant compensating for part of the day’s damage every night.

What should be compared next?

The first experiment I would like to see next is a direct comparison with constant 38°C.

This study compared 32/25°C with 38/27°C. It did not include a 38/38°C treatment in the same experiment.

A future experiment could include conditions such as these:

Temperature regime	What it could test
32/25°C	Normal control
38/38°C	No nighttime cooling
38/27°C	Nighttime cooling
38/32°C, etc.	Temperature dependence of nighttime recovery
32.5/32.5°C	Constant temperature with the same simple daily mean as 38/27°C

If 38°C and 27°C are each applied for 12 hours, their simple 24-hour mean is 32.5°C. Comparing 38/27°C with 32.5/32.5°C would help distinguish between two possibilities: is Cry1Ac maintained simply because the average thermal load is lower, or does the presence of a cooler interval between hot periods itself provide an important recovery opportunity?

It would also be useful to sample not only at 11:00 a.m. but across the day—late afternoon, midnight, pre-dawn, morning, and midday.

If nighttime recovery is real, one might observe a daily cycle such as deterioration during the hot day → recovery at night → deterioration again the next day.

That would provide a much more direct test of the current nighttime-compensation hypothesis.

The experiment is more field-like, but not yet the field

The paper frames the day/night fluctuation as a heat-stress regime closer to the field than constant high temperature. In terms of temperature cycling, that is certainly a step toward realism.

But the controlled-environment light intensity was only 200 μmol m⁻² s⁻¹. The authors themselves note that summer field light can exceed 1,000 μmol m⁻² s⁻¹ and that it remains unclear whether the same compensatory relationship would hold under natural light.

Natural air temperature also does not jump instantly to 38°C at 06:00 and instantly fall to 27°C at 18:00.

So this experiment is best viewed as one step from constant heat toward a more realistic two-stage day/night temperature regime, rather than a complete simulation of field heat waves.

A decrease in Cry1Ac is not automatically the same as weaker pest control

The study measured Cry1Ac protein concentration. It did not feed insects to the plants and measure larval mortality, feeding, or growth.

Therefore, this experiment alone cannot tell us how much a 10–30% decrease in Cry1Ac would reduce practical insect-control efficacy.

Combining Cry1Ac measurements with insect bioassays would allow researchers to ask whether nighttime physiological compensation is sufficient to preserve agriculturally meaningful insect resistance.

One methodological point that caught my attention

The Methods state that 16 squares from different plants were collected for each treatment and time point, after which “16 samples were mixed together and then divided into three biological replicates for measurement.”

Read literally, this sounds as though material from 16 plants was pooled first and then split into three portions. If so, the three portions would resemble subsamples from a single pool more than three independent biological replicates.

The paper does not provide enough procedural detail for us to be certain that this is what was done, so this should be treated only as a point of concern about how statistical independence is described in the Methods. The reproducibility of the overall pattern across two years and two cultivars supports the consistency of the result, but follow-up work using multiple independently pooled biological replicates would make the inference stronger.

Plant strength is not only the ability to withstand stress

When we talk about heat tolerance in plants, we tend to ask, “How high a temperature can this plant withstand?”

But this study, especially when read together with the earlier recovery experiments, suggests another important axis:

How quickly can the plant return after being damaged?

A plant that suffers little damage at 38°C and a plant that is damaged but rapidly recovers during the night may both perform well during a real summer in the field.

In natural environments, the ability to avoid damage may be only part of the story. Plants repeatedly experience stress → recovery → stress → recovery, and their ability to maintain function across those cycles may be just as important.

In this Bt cotton study, Cry1Ac remained stable during the first four days under the 38°C-day/27°C-night treatment but declined by day 7.

For plants, nighttime may therefore be more than simply “the hours when it is not hot.”

It may also be a recovery window in which damage accumulated during the day is repaired and the plant prepares for the next round of stress.

To test that idea directly, future heat-stress studies should measure not only how much damage accumulates during the day, but also how much function returns during the night.

That is the broader question this paper raises for me.

---

Original paper

Liu Y, Ou R, Gu Z, Gao W, Chen Y, Zhang X, Chen D. (2026). Day/night temperature fluctuations delay the high-temperature-induced decline in Cry1Ac endotoxin concentration in cotton squares. Frontiers in Plant Science 17:1913082. DOI: https://doi.org/10.3389/fpls.2026.1913082

Related studies

Liu Z, Wang G, Zhang Z, et al. (2022). Recovery Characteristics of Cry1Ac Endotoxin Expression and Related Physiological Mechanisms in Bt Transgenic Cotton Squares after High-Temperature Stress Termination. Agronomy 12:668. https://doi.org/10.3390/agronomy12030668

Chen Y, Wen Y, Chen Y, et al. (2013). The recovery of Bt toxin content after temperature stress termination in transgenic cotton. Spanish Journal of Agricultural Research 11:438–446. https://doi.org/10.5424/sjar/2013112-2854

Zhang X, Rui Q, Liang P, et al. (2018). Dynamics of Bt cotton Cry1Ac protein content under an alternating high temperature regime and effects on nitrogen metabolism. Journal of Integrative Agriculture 17:1991–1998. https://doi.org/10.1016/S2095-3119(17)61878-1