Impact of wind intensity on tree hydraulic and photosynthetic dysfunction following an extreme hurricane

High winds produced by tropical cyclones are a major concern for the management of most plant-based natural and agricultural systems in the humid tropics and subtropics. This is especially true for trees because storms can impact water relations for decades (Smith-Martin et al., 2022). Landfall of such storms is expected to become more frequent in the coming decades (González-Trujillo et al., 2023). Thus, predicting the impact of various high wind speeds on trees is an important aspect of understanding the effects of climate change on trees in natural and agricultural systems. High winds can cause different types of damage depending both on wind speed and on the physical characteristics of the exposed trees. Following high wind-speed events, immediate damage includes abrasion to and abscission of leaves and fruit and the breaking of some branches (Gardiner, 2021). Depending on the resistance of the roots and soil, the tree may become uprooted during an intense weather event (Gardiner, 2021). Uriarte et al. (2019) found that trees with denser wood underwent less limb breakage in lower category storms, but not after higher windspeeds, suggesting risk factors can vary depending on the severity of the individual weather event. Recent work has highlighted the interaction between plant hydraulic traits and wind speed at the global scale for the first time (Smith-Martin et al., 2022; He et al., 2025), yet the direct relationship between windspeeds and xylem function remains unknown, particularly whether there is a relationship between maximum windspeed and xylem dysfunction. Tropical storms can negatively impact plant hydraulics. Walker (1995) studied the timing of posthurricane tree mortality in Puerto Rico for > 3 yr following Hurricane Hugo. Tree mortality increased due to defoliation caused by wind damage and a higher number of uprooted trees. Low precipitation during the storm likely contributed to this, with drought and hydraulic dysfunction playing key roles in posthurricane mortality (Walker, 1995). Smith-Martin et al. (2022) found that community weighted means of hydraulic safety margins and leaf water potential at turgor loss point decreased for up to 26 yr after Hurricanes Hugo and Georges. Additionally, species with higher hydraulic safety margins were less vulnerable to hurricane-induced mortality (Smith-Martin et al., 2022). Dynamic bending, in which branches were bent back and forth, as is expected in a gusty windstorm, dramatically affected tree stem hydraulic conductivity, suggesting hurricanes are likely to impact subsequent water relations by reducing hydraulic conductivity (Mayr et al., 2014). Ziemińska et al. (2023) studied anatomical traits of 65 trees across 10 species before and after a hurricane and found an increase in theoretical hydraulic conductivity based on area after the hurricane but not in diameter growth. They found differences in anatomy-growth effects that could impact the response of the rest of the trees in the forest following the hurricane. Although windspeed has been hypothesized to impact hydraulic conductivity of trees (Smith-Martin et al., 2022), no study has reported on the impacts of different windspeeds on this important trait for trees nor tested for hydraulic failure differences among windspeeds. In September 2022, Hurricane Ian made landfall on peninsular Florida, USA as a Category 4 storm, passing over citrus production areas with intensities ranging from tropical storm (63–118 km h−1) to Category 3 (178–208 km h−1). The presence of clonal trees distributed across a region larger than the windfield of a Category 4 hurricane provided the opportunity for a natural experiment. We hypothesized that increasing windspeeds would result in greater losses of conductivity in the stems, and that losses in hydraulic conductivity would delay recovery and result in chronic effects on tree health. To assess how different windspeeds during Hurricane Ian affected hydraulic conductivity, we selected three locations with ‘Valencia’ sweet orange (Citrus × sinensis) to measure precent loss of hydraulic conductivity (PLC). These locations had undergone different maximum windspeeds depending on the path of the storm (Fig. 1) and had different observable degrees of damage to the canopies (Fig. 2; Supporting Information Table S1). There was an interaction between hurricane category and wind exposure side (P = 0.024; Table S2) affecting PLC, signifying hydraulic conductivity was affected by the hurricane. Of the three exposure categories, tropical storm had the lowest PLC (15.1% ± 7.49%), followed by Category 1 (27.6% ± 20.2%; Fig. 3), then Category 3 with the highest PLC (52.2% ± 23.6%). In Category 3, the windward side of the tree had a higher PLC than the leeward side of the tree. However, this was the inverse of Category 1 in which the leeward side of the canopy had a higher PLC than the windward side of the tree (Fig. 3). This difference could be due to the fact that the eye of the hurricane passed over the Category 3 site, resulting in the trees receiving wind from different directions. The dynamic of FV/FM over time was different for trees exposed to each category (interaction P < 0.0001; Table S3). FV/FM decreased in Category 3 until it reached its nadir in February 2023, 5 months after Hurricane Ian (Fig. 4). Trees at the Category 1 site experienced a sharp decline during the dry months after February and recovered more slowly than those exposed to the Category 3 storm. By July, the health of Category 1 and tropical storm trees converged while the FV/FM of Category 3 trees continued to increase. High winds caused by tropical cyclones present a major challenge to the growth of trees in natural and managed systems throughout the tropics and subtropics. This challenge is expected to be exacerbated by a warming climate. Extreme wind events produced by tropical cyclones, tornadoes, and derechos are expected to increase in frequency and severity in the coming decades (González-Trujillo et al., 2023). The present study suggests that loss of hydraulic conductivity drives poststorm impairment of photochemical function on trees and is impacted differentially by different categories of storms. Notably, we observed high variability in PLC among trees exposed to Categories 1 and 3 hurricanes, likely reflecting complex interactions between individual tree architecture, branch position, and wind direction under natural field conditions. Windspeeds affected stem hydraulic dysfunction after Hurricane Ian. The trees that underwent Category 3 (129.7 km h−1) suffered the greatest loss of hydraulic conductivity, followed by Category 1 (122.3 km h−1) and tropical storm (77.2 km h−1) with the smallest loss of hydraulic conductivity. This confirmed our hypothesis that PLC would increase with maximum windspeeds. Photosynthetic efficiency as measured via FV/FM was reduced the most in trees that experienced Category 1 windspeeds (Fig. 4), confirming effect of high wind speeds on photochemical function. However, Category 1 winds reduced subsequent FV/FM more than Category 3 winds, which had the greatest decline in PLC. This surprising result may be caused by the relative loss of canopy and of stem conductivity. Our observations indicate that the Category 3 trees were almost completely defoliated, with frequent branch breakage. The relatively small transpirational surface area resulting from subsequent regrowth is not likely to exceed the conductance capacity of the remaining limbs. On the other hand, Category 1 trees experienced only moderate defoliation and minimal branch breakage. Thus, the relatively greater transpirational surface area from greater leaf retention was more likely to create a scenario of transpiration in excess of stem conductivity resulting in photooxidation during the dry season. Growth phenology may have impacted the specific results obtained in the present work. The typical phenology of woody growth for this species in Florida is April to October, and as 2 months of potential growth occurred between sampling at the Category 1 site and the tropical storm and Category 3 sites, it is possible that our measured PLC underestimates the maximum PLC after surviving Hurricane Ian at the Category 3 site. All trees assessed in the present work were infected with a bacterial disease known as huanglongbing, which reduces the tree relative growth rate (Pierre et al., 2021). Nevertheless, the PLC measurements at the Category 3 site occurred before the vast majority of the typical new growth, which is the primary mechanism by which PLC is recovered in tree species (Hammond et al., 2019). The effects of weather conditions after the storm in the present study reinforce the importance of high wind event impacts on tree water relations. Following Hurricane Ian, there was a period of drought. The monthly average precipitation continued to decrease until March 2023, after which the precipitation increased up to July 2023 (Fig. 4). March, the end of the dry period, showed the greatest differences in FV/FM among locations. Given that each Chl fluorescence sampling selected the youngest mature leaves, the FV/FM results likely represent the leaves with the highest function available on the canopy at each time. Thus, those older leaves that may have undergone the actual hurricane-force winds that still remained intact would be expected to be more severely impacted than the leaves that were sampled. By July 2023, after 4 months of consistent rainfall, most of these differences had been erased (Fig. 4), suggesting that reduced hydraulic conductivity due to storm exposure exacerbated the effect of water deficits, a process that may drive the hydraulic responses observed by Smith-Martin et al. (2022) for multiple years after hurricanes. Mechanical strain can lead to embolisms by increasing alternating tension and pressure on either side of the stem, which compresses and stretches the water columns (Mayr et al., 2014). Mayr et al. investigated the effects of branch bending on xylem hydraulic conductivity in conifer branches (Picea abies, Pinus sylvestris), finding a decrease in hydraulic conductivity from static bending (bent once and held in position) and a greater decrease from dynamic bending (bent back and forth multiple times). Wind-driven branch bending may be the primary driver of windspeed effects on hydraulic conductivity, which is also supported by the greater ‘leeward’ side PLC in the Category 3 site, which underwent high windspeeds from multiple directions. An image of wind-whipped branches can be found in Fig. 2(d). Branches blown in the wind could experience dynamic bending depending on the windspeed they experience. In this work, we report a novel and significant set of observations providing preliminary evidence of a relationship between maximum windspeed and subsequent loss of hydraulic conductivity. The breadth of our observations is limited to a small number of sites and only a single species. The wind response could differ among species with different hydraulic traits and wood tensile characteristics (Uriarte et al., 2019; Smith-Martin et al., 2022). Although recent work by He et al. (2025) addresses impacts of annual mean windspeeds on plant hydraulics (the maximum mean windspeed was c. 25 km h−1), the nature of the responses may be quite different from those of short periods of extreme wind as are reported in the present work. For example, the mean windspeed in peninsular Florida is 11.9 km h−1, despite occasional storm-induced speeds that are greater by an order of magnitude. It is also important to recognize that the effects of the same weather event can vary significantly between individual trees. Differences between individual tree architecture may affect drag and thus mitigate or exacerbate windspeed impact (Gardiner, 2021), while characteristics across different species and varieties could leave the tree to be more susceptible to branch bending and breaking (Uriarte et al., 2019). Measuring PLC before and after storms and over drought periods would be especially revealing. Future work could address this temporal aspect and would more fully test the hypothesis. Importantly, this study demonstrates that field studies deployed following large storms can reveal relationships between wind speed and plant dysfunctions that are necessary to quantify current and predict future ecosystem-scale processes. Specific responses of various growth forms, lineages, root health, and environmental temperatures to different windspeeds would enable better modeling of ecosystem-level recovery and facilitate better prediction of persistent agricultural losses after high wind events. The present work shows that increasing maximum windspeed drives increasing PLC. Chronic loss of hydraulic conductivity drives the tree recovery pattern from Category 1 winds, while it plays a lesser role in Category 3 recovery due to defoliation, despite the greater PLC. Hydraulic impacts of storms on tree branches impose a structural limitation on subsequent growth and help explain the subsequent responses to precipitation. Important avenues of future investigation include (1) expanding to species whose structural and physical characteristics differ from Citrus, (2) assessing different windspeed exposures, such as Category 2, and (3) studying the relative impact of subsequent weather patterns on storm effects on subsequent growth. These results can help inform global models of the effects of high-wind effects on plant growth. Thank you to Rebecca Ebert, Myrtho Pierre, Dorcas Alade, Sharpton Toussaint, Mary Sutton, Jean Francois, Sherline Estaing, Maryam Khalilzadeh, and Bilal Syed Hussain for all the help and support during the data collection process. A very large thank you to the growers that participated in the study. This work was performed with funding from the USDA-NIFA A1712 Rapid Response to Extreme Weather Events Grant no. 2023-68016-39040 ‘Recovering from Ian and enhancing future hurricane readiness’. None declared. CV, WMH and GZM conceived of and designed the study. GZM implemented the study. WMH, RC-A and YS measured hydraulic conductivity. GZM analyzed the data and wrote the manuscript, with guidance from CV. WMH and CV revised the manuscript. All data are included in the Supporting Information. Dataset S1 Precipitation data from a study of citrus trees in locations undergoing different windspeeds during Hurricane Ian. Dataset S2 Percent loss of hydraulic conductivity results from a study of citrus trees in locations undergoing different windspeeds during Hurricane Ian. Dataset S3 Chlorophyll fluorescence results from a study of citrus trees in locations undergoing different windspeeds during Hurricane Ian. Methods S1 Methodology, including sampling design, timing and measurement of chlorophyll fluorescence and hydraulic conductivity responses of citrus trees to after Hurricane Ian. Table S1 Locations of growers with the varieties, soil order, and ages of trees included in the study, as well as the maximum windspeeds (km h−1) in each location. Table S2 Analysis of variance of percent loss of conductivity of leeward and windward branches of Valencia trees from three different locations: Fort Meade, Zolfo Springs, and Fort Pierce that underwent different categories of windspeed from Hurricane Ian. Table S3 Analysis of variance of linear model for FV/FM of leaves of citrus located throughout the path of Hurricane Ian during c. 10 yr after the storm, as affected by expected factors relating to wind exposure. Varieties ‘Valencia’ sweet orange and ‘Valquarius’ sweet orange (Citrus × aurantium). Hurricane Ian impacted the state on September 28, 2022. Survey period included monthly or bimonthly sampling from December 2022 to September 2023. Please note: Wiley is not responsible for the content or functionality of any Supporting Information supplied by the authors. Any queries (other than missing material) should be directed to the New Phytologist Central Office. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article. The New Phytologist Foundation remains neutral with regard to jurisdictional claims in maps and in any institutional affiliations.