Climate stress priming of juvenile Southern California giant kelp (Macrocystis pyrifera) to marine heatwaves

Seaweed aquaculture production is expanding globally but threatened by increasing climate stressors, including extreme events. Ocean-based aquaculture is especially vulnerable to marine heatwaves, which can rapidly raise temperatures above the physiological limits of organisms, including macroalgae. While several interventions to increase aquaculture's resilience to climate impacts are being explored across different scales, ‘priming'-exposing an earlier stage of an organism to a mild stressor to prepare the individual for future stress-has emerged as a possible species adaptation for seaweeds that maintains genetic diversity but hardens individuals to stressors later in life. California has a developing seaweed sector while also experiencing some of the most extreme marine heatwave conditions on record. Using laboratory thermal experiments, we explored temperature impacts and potential priming effects on giant kelp, Macrocystis pyrifera, an important foundation species along the West Coast of the United States. Our trials focused on the juvenile sporophyte stage grown on miniaturized aquaculture spools. First, we determined the baseline reaction norms of the local M. pyrifera (5 to 30 °C) at the outplanting stage. Next, we tested if priming (with heat and/or nutrients) in a hatchery setting prepares M. pyrifera for outplanting to a marine heatwave. We measured growth, survival, photosynthetic function (Fv/Fm), and carbon and nitrogen assimilation via isotopes. We found temperatures above 20 °C had significant negative impacts on all metrics of performance of unprimed juvenile sporophytes, with conditions closer to 10 °C more favorable. Further, we determined heat priming in conjunction with hatchery level (i.e., elevated) nutrients resulted in overall increased performance when exposed to a simulated marine heatwave (18 °C) at outplanting. These findings support the continued exploration of priming as a tool for climate resilience and can inform current hatchery practices for aquaculture practitioners looking to improve crop outcomes for this species.

Global food production and security are increasingly threatened by climate change and the associated extreme variability (Duchenne-Moutien and Neetoo, 2021; Pörtner et al., 2022), with farmed seaweeds among the growing sectors at risk (Pörtner et al., 2022; Froehlich et al., 2022). Beyond shifts in mean conditions, extreme events like atmospheric rivers and heatwaves (Atalah et al., 2024) are creating unprecedented acute challenges for wild and farmed systems alike (Owino et al., 2022; Motha, 2011; Ruthrof et al., 2018). Marine heatwaves (MHWs), a minimum of five consecutive days where temperatures remain above the 90th percentile established using a 30-year historical baseline (Hobday et al., 2018), can alter thermal conditions so rapidly they surpass the physiological capacity of organisms, resulting in negative shocks to populations and ecosystems (Free et al., 2023; Oliver et al., 2021). While impacts and possible adaptations to extremes are increasingly studied, aquaculture has been comparatively overlooked (Pörtner et al., 2022; Froehlich et al., 2022). This knowledge gap is especially stark for farmed seaweeds, a sector accounting for 36 million tons (wet weight) and $17 billion (USD) annually (United Nations Geneva Report, 2024). Of the estimated 32 algal species farmed worldwide (FAO, 2024), 40% belong to brown algae (Phaeophyceae) (Ruiz Martínez et al., 2024), kelps are a particularly important subset used as fertilizers, feed, and biofuels, and increasingly viewed as a potential low-carbon food source for humans (Ruiz Martínez et al., 2024). Given the growing threat of MHWs, determining how to increase cultured kelp resilience to these shock events is important for the growing sector.

There are several on-farm, organismal interventions to help adapt to climate change and potential extremes (e.g., genetic selection) (Froehlich et al., 2022), but one emerging adaptation for species lacking a nervous system is “ priming ” (Veenhof et al., 2024). In scientific literature, priming is also referred to as “conditioning”, “hardening”, and even “acclimation” (Jueterbock et al., 2021). Priming is already used in land-based crop agriculture (Ibrahim, 2016; Wojtyla et al., 2020; Liu et al., 2022) and is being tested on invertebrates (Lewis-Smith et al., 2025). Priming in aquaculture occurs when an organism is exposed to a mild stressor at a critical point (typically early) in their life stage in a hatchery, resulting in a “stress memory”, which prepares (i.e., primes) the individual for future stressful conditions (Hilker et al., 2016; Jueterbock et al., 2021). The stress memory is defined as a plastic change at the phenotypic level, induced by a non-lethal stressor that does not change the underlying genetics for the organism (Hilker et al., 2016; Jueterbock et al., 2021). Although currently unknown for most seaweeds, mechanisms of priming in other organisms include epigenetic modifications, such as microRNA, histone modifications, and DNA methylation, and resulting changes to gene expression, physiology, and metabolism that can be retained over short- (i.e., days to weeks) or possibly long-term timeframes (i.e., lifetime, generations if bred) (Hilker et al., 2016; Jueterbock et al., 2021). It is also hypothesized that priming needs to occur on earlier stages of seaweed and marine plants to be most effective due to cellular changes in growth and development (Jueterbock et al., 2021). Relatedly, the level of nutrients added at the stage kelp is growing in a hatchery can have a priming influence. Nutrient concentrations can have complex interacting effects with high temperatures because temperature influences metabolic rate and thus nutrient demand (Fernández et al., 2020; Umanzor et al., 2021). However, the “nutrient rescue hypothesis” posits that supplemental nutrients can increase tolerance to higher temperatures through differing mechanisms, including maintaining lipid integrity of cell membranes, if available at the right moment in time (Ahanger et al., 2019; Fernández et al., 2020; Kumari et al., 2022). Importantly, in some regions (e.g., USA), adaptative interventions for new or non-established species, such as farmed seaweeds, are required to protect surrounding genetic diversity, thus limiting more typical selective practices. For example, Alaska's “50–50 rule”—being adopted by other states, including California—where spore collection must be from at least 50 individuals within 50 km of the outplanting site (Gruenthal and Habicht, 2022). While MHW intensity, frequency, and extent increase globally (Dong et al., 2025), considerable impacts and research across the California current provide an ideal region to explore such impacts and interventions.

The state of California has not only seen some of the most extreme MHWs, impacting socio-ecological ecosystems across the coast (Sanford et al., 2019; Lonhart et al., 2019; Free et al., 2023), there is also an emerging seaweed sector. The 2014–2016 MHW off the west coast of North America, known as “the blob”, had devastating, although varied, impacts on wild populations of giant kelp, Macrocystis pyrifera (Cavanaugh et al., 2019). M. pyrifera is a foundation species (Miller et al., 2015, 2018), playing a large role in nearshore community structure and function (Bruno and Bertness, 2001), supporting over 800 other species along the California coast (Schiel and Foster, 2015). In 2020, Ocean Rainforest broke ground on the first offshore pilot farm (81 km) in California, growing M. pyrifera. Yet, recent climate studies project a potential decrease in suitability for seaweed culture across the West Coast of the United States, with southern California being particularly vulnerable (Fong et al., 2024). Certain life stages, such as adult sporophytes and gametophytes, of these species have received a lot of research attention (Leal et al., 2017; Mabin et al., 2019; Hollarsmith et al., 2020; Fernández et al., 2020; Kurman and terHorst, 2023). In fact, a recent study found heat priming gametophytes for 4 weeks led to better outcomes for their sporophytes when later exposed to high temperatures (Gauci et al., 2024). However, juvenile sporophytes—the stage farms outplant into the wild—are understudied. A recent study did find MHW intensity and duration negatively affect growth in young sporophytes (Bunting et al., 2024), but the stage (older juveniles) and context (non-spooled) of the study were not aquaculture specific. This provides an opportune moment to partner with farms to explore interventions at the hatchery stage that could improve growth and survival during the outplant stage at higher temperatures.

The objective of our research was to 1) establish the baseline of how temperature impacts growth at the early sporophyte stage of M. pyrifera from a local source in California and 2) explore how priming in the hatchery stage may trigger shifts in morphology and/or physiology that lead to an increased ability to withstand marine heatwaves in the future at the individual (single blade) and population scale (spool). These included measures of growth, survival, photosynthetic function, and carbon and nitrogen assimilation (via stable isotopes). We expected juvenile sporophytes, like their adult counterparts (Provost et al., 2017; Fernández et al., 2020), to experience decreased growth outcomes at or above 25 °C, but may be more sensitive to less extreme thermal conditions (e.g., 20 °C) due to the younger stage (e.g., Muth et al., 2019). We hypothesized that the stress caused by increased heat will be mitigated by a combination of heat and nutrient priming during the early stage. Understanding the viability of heat priming sporophytes at this early stage will inform the active measures kelp farms can take to climate proof their crop as human-induced climate change continues.

Overview

We conducted two experiments which we will refer to as the “ kelp performance experiment ” and the “ kelp priming experiment ” (Figure 1). In the kelp performance experiment, we exposed early juvenile M. pyrifera at the outplanting size (>1 mm) to a spectrum of thermal conditions to determine the baseline of how temperatures between 5 and 30 °C will impact the locally sourced kelp growth at this stage (i.e., performance curve/reaction norm). The baseline kelp performance experiment helped determine the expected thermal performance of unprimed sporophytes, providing the reference against which any priming-induced shift in thermal tolerance could be evaluated across a fuller range of temperatures. For the kelp priming experiment, we manipulated nutrients, temperature, or both for the early sporophytic stage to simulate short-term priming in the hatchery. Following a recovery period, the sporophytes underwent a simulated outplanting into a MHW to determine what combination of priming improves the performance of young sporophytes to more immediate challenging outplanting conditions.

Spore collection and culture

M. pyrifera spore collection and culture for both experiments were provided by Ocean Rainforest. All sorus material was collected via SCUBA between 7 and 12 meters of water within 50 km of the farm, located just north of Santa Barbara and 8.1 km off the coast. Their protocols, which are current and similar to standard existing protocols (Le et al., 2022), were followed for sporulation.

To simulate growing conditions of a typical kelp hatchery, “mini-spools” were co-designed (N = 72 total number of spools) with Ocean Rainforest to emulate the structure and function of a seeded spool (a full-sized spool is 15” x 2” sterilized PVC pipe). Specifically, six centimeters of twine was tightly wound around 4” x 1” PVC to create the mini-spools for M. pyrifera seeding (Figure 2). At Ocean Rainforest, the mini-spools were submerged in a gametophyte solution (40,000 spores/mL) for 24 h and then grown in their facility until transitioning to the sporophyte life stage. This size was selected as it is hypothesized that priming needs to occur in the earlier stages of seaweeds to be most effective due to cellular changes in growth and development (i.e., “reset”) (Jueterbock et al., 2021) (Figure 1). For the kelp performance experiment, all mini-spools were reared together in a single, large aquarium tank. Based on logistical challenges and labor capacity at Ocean Rainforest, there were some slight differences in mini-spool creation for the kelp priming experiment. Furthermore, due to the laboratory capacity, the kelp priming experiment took place in two phases, each with 24 mini-spools (Figure 1B). For phase 1, two spools were raised in the same 3L mesocosm jars (n = 12 jars) at Ocean Rainforest. In phase 2, all 24 mini-spools were reared together as in the kelp performance experiment. Regardless or experiment or phase, once sporophytes were visible (< 1 mm; approximately 1-week before outplanting size), the mini-spools were transported in a single cooler on ice to UC Santa Barbara. During an approximate 1-h window of time, mini-spools were carefully and individually wrapped in plastic wrap to maintain moisture, placed on ice in a single cooler while transported, then kept refrigerated at 10 °C until placed into their designated tanks for initial acclimation that same day.

Experimental conditions

The kelp performance experiment and kelp priming experiment took place in the Froehlich Lab Aqualogic tank system at UC Santa Barbara. The tank system was equipped with six headers to control temperatures and 12 experimental tanks (Figure 3). Header tanks were connected to chilled seawater (10 °C) that had gone through six gravity sand filters that remove particles above 20 microns. The chilled seawater is gravity fed into the experimental tanks, acting as a water bath for the 3 L mesocosm jars (two jars per tank). After transport from Ocean Rainforest, mini-spools were hung from the top of each mesocosm jar [kelp performance experiment N = 24 and kelp priming experiment N = 48 split evenly across two phases (n phase1 = 24, n phase2 = 24)]. Inside each jar, we placed UV-treated aquarium rocks to maintain negative buoyancy of the jar. Above each tank was a 12-inch grow light (Barriana T5 LED full spectrum grow light; apogee PAR meter registers 50 μmol m −2 s −1 in tanks with no water) with a light-cycle of 14:10 day to night, replicating Ocean Rainforest's protocol.

Both experiments were split into a hatchery and outplanting stage (Figure 1), described in more detail below. Before being added to the mesocosms, the filtered seawater was exposed to a 4-h UV treatment (20 W, 254 nm wavelength) to reduce bacteria and diatom contamination. We elected not to use geranium dioxide to inhibit diatom growth on advice from the hatchery. During the hatchery stage (7-days), nutrients were added using a concentrated Guillard f/2 formula produced by the Mercer of Montana and used inhouse by Ocean Rainforest. For the outplanting stage (10-days) of the experiment, all mini-spools were rotated to hang from the opposite side, allowing for more equal exposure to grow lights on both sides of the spool, and the water in each mesocosm was exchanged for non-UV treated seawater, with no additional nutrients added. The latter, non-treated water was meant to simulate “real world” ocean conditions.

Temperatures across both experiments were tracked using a mix of Onset HOBO pendant temperature and temperature/light loggers. A total of 24 loggers were deployed, two in each experimental tank, that logged temperature (°C) every 30 min. For each treatment, these temperatures were averaged during each discrete temperature window.

Kelp performance experiment

To help inform and interpret the kelp priming experiment, we conducted a baseline performance curve experiment on M. pyrifera juvenile sporophytes across a thermal gradient (N = 24 mini-spools; n = 4 per thermal treatment; Figure 1A) from October 7th through 24th, 2024. In order to test the comparable outplant stage of the priming experiment, all mini-spools were grown at 10 °C with hatchery levels of nutrients (1 mL of f/2 medium per L of seawater) for 7 days in the lab (i.e., hatchery stage). Once the hatchery stage was complete, headers were set to temperatures between 5 and 30 °C, in five-degree increments (i.e., outplanting stage). Mini-spools were exposed to these temperatures for 7 days before all headers were returned to 10 °C for 3 days of recovery and subsequent data collection.

Kelp priming experiment

To test the performance of primed M. pyrifera, we conducted a 2-factor fully crossed experiment during the hatchery stage, split across two phases (N = 48; 24 per phase; Figure 1B). Given the known importance and potential interactions between temperature and nutrients (Fernández et al., 2020; Umanzor et al., 2021), we explicitly accounted for the potential interactive effects of the “nutrient rescue hypothesis” in the experiment design. The treatments included a control (unprimed; n = 12), and three different priming treatments. All treatments occurred during the 7-day hatchery stage. The first priming factor was “nutrient primed”, where half the mini-spools were given “high” nutrient concentrations (1 mL of f /2 medium L −1), while the others were given half (0.5 mL of f /2 medium L −1). Note, the high nutrient treatment is the amount typically administered in the hatchery. The second priming factor was “heat primed”, a short-term, moderate thermal heat exposure. The heat priming consisted of 48 h at higher temperatures, beginning after a 2-day acclimation period. Tanks were raised to 12 °C for 24 h and raised again to 15 °C for 24 h before being returned to 10 °C for 2 days of recovery. Tanks took approximately 3.5 h to reach the higher temperature. This short exposure regiment aligns with the current standing priming science for macroalgae (Jueterbock et al., 2021). Temperatures were based on thermal dynamics near the farm (NSF LTER, NOAA buoy, and Ocean Rainforest hobo logger) and known thermal bounds of adult sporophytes (Fernández et al., 2020). Phase 1 was conducted from March 22nd to April 8th, 2024 and included the fully primed (nutrients and heat combined; n = 12) and control treatments. Phase 2 was conducted from July 12th to 29th, 2024 and included the nutrient primed (n = 12) and heat primed only (n = 12) treatments.

After 7 days in the hatchery stage, all mini-spools experienced a simulated outplanting—although the twine remained spooled—into a MHW (i.e., no added f/2 nutrients, and no-UV treated seawater). This began by raising all header temperatures to 12 °C for 24 h, then to 15 °C for another 24 h. Finally, the temperature was raised to 18 °C for 5 days before a 2-day recovery period (Figure 3). Each change in temperature took approximately 3.5 h. Again, the MHW level was informed by past thermal measures in the region of interest, specifically during the months outplanting typically occurs (November/December). The MHW was based on the common and minimum criteria of a defined heatwave event; at least 5 days in a row of temperatures exceeding 90% of the past observations at that location (Communications, 2023).

Kelp measurements

Wet weights and photographs of the spools were taken throughout the kelp performance experiment and kelp priming experiment to determine differences in percent cover, coloration, and total weight across treatment and over time. Percent cover was taken to capture evenness of visible growth over the course of the experiments, while total weight of the spool allowed us to calculate relative growth rate. These combined measures provide insights into overall performance (growth and survival). Changes in coloration or loss of pigment can signal blade response to environmental stress(ors) (North, 1987; Sánchez-Barredo et al., 2020).

At the conclusion of the experiment, all spools were unwound, a subset of blades were counted, and all sporophytes were removed. At this time chlorophyll- a fluorescence (Fv/Fm) was measured and individual blade photographs taken for color analysis (performance experiment only). All sporophytes were dried, weighed, and later sent for stable isotope analysis. Blade counts can be used to compare blade density and overall success of each spool. Differences in chlorophyll- a fluorescence can indicate stress impacts on photosystem II (PSII), while an indirect effect of high temperature in kelp is loss of pigmentation, or bleaching (North, 1987). Blade measurements, such as surface area, show differences in growth and development between treatments. Dry weights, or dry content matter, indicate the total amount of growth on each spool. Any difference in measurements between the performance and priming experiment are noted below (see Table 1).

Proportion of cover

To determine sporophyte percent cover of the mini-spools during both experiments, photographs (on the same side) were taken upon receipt of the spools (day 0), at the conclusion of the hatchery stage (day 7), and at the conclusion of the respective experiments (day 17). All photos were uploaded to and analyzed in Fiji Is Just ImageJ (FIJI) (Schindelin et al., 2012). Images were cropped to the growth area, converted to 8-bit in the Image menu and made binary in the Process menu (Figure 4). Black and white images were examined by an expert to ensure glares and shadows were removed, minor alterations were made when necessary, using the black and white paint tool. Finally, the Histogram tool from the Analyze menu was used to determine the number of black and white pixels (List function within Histogram), 0 indicating black and 255 white.

Weight

We calculated the relative growth rates (RGR) of the mini-spools using a logarithmic formula for weight:

Spools were subsequently unwound and all sporophytic material collected, rinsed in fresh water, and dried to a consistent dry weight (g) at 60 °C for at least 24 hrs in a VWR drying oven for dry weight measures and isotopic analysis (see below).

Blade count

To compare the density of blades between treatments at the end of the experiments, the unspooled twine was randomly cut into two 15 cm sections from the middle of the spool. Each 15 cm section was photographed and the blades were enumerated using the count tool in FIJI. There is no prescribed optimum density in current M. pyrifera aquaculture reports, thus our measurements potentially provide a comparative baseline.

Chlorophyll-a fluorescence

We used a junior-PAM (Walz) chlorophyll- a fluorometer to measure the Fv/Fm of the juvenile sporophytes. Fv/Fm —a ratio that indicates the efficiency of photosynthesis based on maximum quantum yield of PSII and is commonly used as an indicator of plant stress—is the variable fluorescence divided by the maximum fluorescence. After at least 15 min of dark acclimation, Fv/Fm measurements were taken in the middle of the blade across ~10 blades for each spool. If there were not 10 blades present, we recorded zero values. We calculated the average of these Fv/Fm readings, resulting in a single value per spool.

Coloration in performance curve

Given the importance of photosynthesis to the success of kelp growth and survival but limited capacity to measure in actual farm application, we set out to assess the relationship between coloration and photosynthetic efficiency across temperature treatments; an easier metric to potentially collect and assess (i.e., a photo). Kelps and other seaweeds are known to bleach when under stress (North, 1987), which can indicate a negative impact to photosynthesis (Irving et al., 2004). Measurements of mini-spool coloration were performed using FIJI on photographs at the end of the hatchery stage (day 7) and the end of the experiment (day 17) for the kelp performance experiment. The area of growth was outlined manually with the polygon tool and the histogram tool was used to determine the mean value for spool coloration (McMahon et al., 2024) (Figure 5). Smaller values of coloration indicate darker pixels in the image.

Blade surface area, length, and width in performance curve

For the kelp performance experiment, we collected blade surface area, length, and width in FIJI. These measurements provide individual, trait-based assessments of growth to compare to the overall performance of the spool. When possible, five blades per spool were collected from the string tails (these tails were discarded and not used in the dry weight, coloration, or percent cover analysis). Blades were placed on a micrometer and images were taken using a dissecting scope. In FIJI, the scale was set in the Analyze menu using the micrometer. Surface area was calculated using the polygon tool to manually outline the blade, while length and width were determined with the straight tool (Figure 6). Blade surface area, length, and width were averaged across mini-spool (N = 24).

Nutrient content and stable isotope values

Tissue samples from each spool were collected and dried for stable isotope analysis. All samples were ground, weighted to the nearest microgram, and packaged in aluminum foil following the protocols for the UC Davis Stable Isotope Facility, where the analyses were performed (Brandeberry, 2020). These results help inform how thermal stress may be impacting nutrient uptake and assimilation. More specifically, isotopic composition can reflect shifts in carbon fixation pathways, metabolic rates, nutrient use, and internal allocation of resources under different environmental conditions because biochemical reactions discriminate between isotopes in predictable ways (i.e., fractionation) (Snyder et al., 2022; Bunting et al., 2024). For instance, changes in δ 13 C in kelp tissue can indicate changes in photosynthetic carbon assimilation or shifts in the source of carbon (Drobnitch et al., 2018; Velázquez-Ochoa et al., 2022). Our samples were analyzed for total carbon content (%), total nitrogen content (%), δ 13 C, and δ 15 N using an EA-IRMS system at the UC Davis Stable Isotope Facility.

Statistical analysis

Kelp performance experiment

To analyze the reaction norms of the proportion of mini-spool coverage, relative growth rate, final dry weight, blade count and Fv/Fm across the temperature gradient (day 17), we compared several generalized linear models—specifically an intercept, linear, and second order polynomial regression—using a beta distribution with a logit link function and calculated the associated Akaike Information Criterion Corrected (AICc) values. The beta distribution was selected because proportion is bounded between 0 and 1. AICc allows a likelihood comparison of model fit in relation to model complexity (Akaike, 1998; Burnham and Anderson, 2002). Best fit was determined by the model with the lowest AICc value. The R package glmmTMB was used to build the comparative models, while the aictab function in the package AICcmodagv was used to compare model performance (Brooks et al., 2017; Mazerolle, 2023). We further tested the effect of time (via day 0, 7, 17) across the gradient on coverage to discern when significant changes were detectable, if any.

To assess how color related to photosynthetic performance, we visualized the difference in color intensity between day 7 and day 17 relative to the corresponding Fv/Fm. We followed this with a Kruskal Wallis test to compare Fv/Fm levels between replicates that clustered together.

To discern how temperature impacts kelp growth on an individual level, we performed one-way analysis of variance (ANOVA) for blade surface area, height, and width as residuals were normal and variances homogeneous. We followed those with Tukey's post hocs.

We also assessed the relationship between % C by % N, fit with a linear model. Such a comparison provides insights into the range and relationship for these earlier life stages. This was followed by a post hoc cluster analysis to determine if the relationship between carbon and nitrogen grouped by temperature exposure treatment.

Percent carbon, carbon to nitrogen ratio, δ 13 C, and δ 15 N met normality assumptions for parametric statistics and measures across the six treatments were analyzed using one-factor ANOVAs followed by Tukey's post-hoc test. A Kruskal-Wallis followed by a Dunn's post hoc was used to analyze %N due to normality assumptions not being met.

Kelp priming experiment

In addition to the differences between phase 1 and phase 2 described above, there was a bacterial outbreak during phase 2 at Ocean Rainforest before the spools were transported to UC Santa Barbara. Due to this outbreak, water was changed more frequently at Ocean Rainforest and the sporophytes were a few days older than phase 1 sporophytes at the start of the experiment. “Phase” was too even to be added as a random effect in our models, so instead we compared the two phases with appropriate post-hoc analyses.

To assess the comparative impact of our four treatments [unprimed (control), heat primed, nutrient primed, and fully (heat + nutrient) primed] on the proportion of mini-spool coverage over time (day 0, 7, and 17), we followed the same steps described for the kelp performance experiment. We also conducted a two-factor ANOVA on day 0 proportions to ensure no significant differences at the start of the experiment, given the phased approach we had to employ due to tank-space limitations.

The other “end of experiment” metrics (day 17)—blade count, relative growth rates, Fv/Fm, and dry weight—were visualized and regressed with appropriate distributions relative to the data type. Since time was not a factor, the various models only compared the differences of the four treatments. Final dry weight met normality assumptions and was analyzed using a linear model. For the other metrics, the glmmTMB package was used to assess the significance of any differences. For blade count, we used a gamma distribution with a log link function. Relative growth rates were compared using a gaussian distribution and an identity link function, while we used a beta distribution with a logit link function for Fv/Fm. All models were followed by post-hoc tests calculating the least-squares means using the function emmeans to determine pair-wise differences between treatments.

Percent carbon, percent nitrogen, and carbon to nitrogen ratio met normality assumptions for parametric statistics and primed treatment effects were analyzed using linear models. To assess differences in δ 13 C we used a glmmTMB with a gaussian distribution and an identity link function. For all nutrient analyses, we once again conducted appropriate post-hoc tests to compare phase 1 and phase 2 trials. Due to obvious phase differences, for δ 15 N we only conducted Kruskal Wallis comparing phase 1 and phase 2.

All statistical analyses across the two experiments were performed in R version 4.2.3 (R Core Team, 2021).

At the conclusion of the kelp performance experiment, 22 HOBO pendants successfully logged temperatures across the 17 days. Most experimental tanks had two replicate loggers, however two loggers from different tanks failed to record. The average temperature across all tanks before the heatwave treatment began was 10.2 °C (SD = 0.16). During the 7-day heat treatment, average temperatures ranged from 7.23 °C (SD = 0.43) for the 5 °C treatment to 28.40 °C (SD = 1.83) for the 30 °C treatment (Table 2 and Figure 1A). After the heat treatment, the average temperature across all experimental tanks was 10.3 °C (SD = 1.04) for the 2-day recovery period.

At the conclusion of phase 1 and 2 of the kelp priming experiment, 21 out of 24 HOBO pendants successfully logged across the 17 days. Temperatures remained close to the target temperatures for each phase (Table 3 and Figure 1B).

Kelp performance experiment

Across all performance measures we found optimum patterns typically between 10–15 °C and a clear threshold response when temperatures exceeded 20 °C (Figure 7). In fact, for spool coverage, relative growth weight, and final dry weight we did not include the 25 and 30 °C temperature treatments due to cyanobacterial growth but did include them for average blade count where we could indicate zero blades and average Fv/Fm where we took readings from the few blades that survived and included zeros when blades were not present. The best fit for the proportion of mini-spool coverage on the final day was the second order polynomial regression (AICc = −17.01, AICcWt = 0.57; R 2 = 0.31, F -stat = 24.41 on 2 and 21 DF), peaking at 15 °C at 0.79 (±0.06) cover (Figure 7A). For relative growth rate, the best fit was also the polynomial regression (AICc = −33.30, AICcWt = 0.79; R 2 = 0.41, F -stat = 6.6136 on 2 and 13 DF) with the highest growth occurring closer to 10 °C, which had an average of 30.2% d −1 (±0.02) (Figure 7B). Finally, the best fit for final dry weight was a linear model (AICc = −56.74, AICcWt = 0.64; R 2 = 0.10, F -stat = 2.608 on 1 and 14 DF). The dry weights where highest for the 5 °C and 10 °C treatments (0.12 g ± 0.01 and 0.15 g ± 0.02, respectively), then decreased by an average 0.02 g (±0.01) per ca. 5 °C as temperatures increased (Figure 7C). The reduction in range (excluding 25 and 30 °C), in addition to a truncated dataset only examining these measures on day 17, likely contributed to low R 2 values for the proportion of coverage on day 17 and dry weight.

Patterns at the spool (population) level mimicked those at the individual level. For average blade count, the best model fit was the polynomial regression (AICc = 198.47, AICcWt = 0.95; R 2 = 0.67, F -stat = 4.369 on 2 and 21 DF). The 15 °C treatment had the highest blade count, with an average of 49 blades per 15 cm, however the counts were similar for the 10 °C treatment (Figure 7D). The best fit for Fv/Fm was a polynomial regression (AICc = −28.82, AICcWt =; R 2 = 0.88, F -stat = 87.7 on 2 and 21 DF) (Figure 7E). Any bleaching was noted qualitatively but was avoided when determining the final Fv/ Fm values, as we were interested in the intact photosynthetic tissues. Several bleached spots were tested, resulting in a zero every time. This does not inform the overall photosynthetic potential of a mini-spool. The average Fv/Fm for the cooler temperature treatments (5–20 °C) ranged from 0.649–0.669, with a precipitous drop off for the higher temperature treatments. For photosynthetic capacity at 25 and 30 °C, where cyanobacteria were present, we cannot say definitively how the kelp would have fared across these metrics without the cyanobacteria growth and our results are likely at or below the reduction in photosynthetic efficiency.

Modeling the full extent of the proportional coverage, the best fit model, based on the AICc value and weight, (AICc = −116.21, AICcWt = 0.78; see Supplementary Figures S1, S2 and Supplementary Table S1) included day and treatment and the interaction between the two. There was specifically a significant interaction between the final day (day 17) and two temperature treatments, 10 °C and 15 °C (p -values = 0.007 and < 0.001, respectively), indicating that coverage evenness was significantly higher between these two thermal conditions after recovery (day 17), but not before.

There was a clear threshold delineation between the highest temperature treatments (25 °C and above), including coloration (Figure 8). Across the 5–20 °C treatments, the majority (88%) of mini-spools decreased in color intensity (i.e., became darker), whereas all warmer treatments saw an increase (Figure 8A). The cooler treatments (≤ 20 °C) formed a tight aggregation between 0.640 to 0.680 Fv/Fm and displayed a difference in color intensity less than 22 (Figure 8B). In comparison, the warmer treatments had readings below 0.400 and above the 22-coloration difference threshold. A post-hoc analysis comparing Fv/Fm of these two clusters found significant differences (Kruskal-Wallis; p -value < 0.001), with Fv/Fm decreasing by an average of 0.370.

Overall, blade growth across all metrics was restricted at temperatures at or above 20 °C (Figure 9), similar to the broader mini-spool measurements trends. Average surface area was highest at 15 °C and lowest at 25 °C (Figure 9A; one-way ANOVA, p -value = 0.0137). Average height decreased at higher temperatures starting at 20 °C (Figure 9B; one-way ANOVA, p -value = 0.0012), while average width decreased at temperatures above and below 15 °C (Figure 9C, one-way ANOVA, p -value = 8.72 x 10 −5). We also found fewer blades at higher temperatures, with only three blades found total on the 25 °C mini-spool tails and no blades in the 30 °C treatments (i.e., 100% mortality).

There were not significant differences between treatments for percent carbon (one-factor ANOVA, p -value = 0.0964; Figure 10A) or percent nitrogen (Kruskal-Wallis, p -value = 0.06017; Figure 10B), but there were differences in the carbon to nitrogen ratio (one-way ANOVA, p -value = 0.0016, Figure 10C). Reflective of the broader trends, the C:N ratio was the highest at the 15 °C treatment and the lowest at 5, 10, and 30 °C. The average ratio, however, at 15 °C was only 0.76 parts carbon to nitrogen larger than 30 °C treatment (Figure 10C). There were also significant differences in δ 13 C (one-way ANOVA, p -value = 0.0072) where the 15 °C treatment is higher than the 25 and 30 °C treatments (Figure 10D). δ 15 N had significant differences across treatments (one-way ANOVA, p -values = 0.0032) with the 20 °C treatment lower than the 5, 25 and 30 °C treatments (Figure 10E).

As expected, percent nitrogen increased linearly with percent carbon (R 2 = 0.91, F -stat = 202.8 on 1 and 20 DF, Figure 11) with values ranging from 4.14 to 8.02% for N and 26.11 to 43.75% for C. We performed a post-hoc cluster analysis to examine patterns in this relationship across temperatures, however extreme heat did not stand out (see Supplementary Figure S2). For all nutrient and isotope analyses, it is important to note two samples were not included (one from both 5 °C and 20 °C treatments) as the samples were too small for isotopic analysis.

Kelp priming experiment

The best fit model to examine differences in mini-spool cover over time included day, nutrient treatment, and heat treatment as fixed factors and their interactions (AICc = −242.8, AICcWt = 0.54; see Supplementary Table S3), with a significant three-way interaction between day, nutrient treatment, and heat treatment (p -value < 0.001, Figure 12 and Supplementary Table S4). On the final day, fully primed mini-spools had higher coverage than expected from the additive effects of nutrient and heat priming. There were no significant differences in proportion of cover at the start of the experiment (two-factor ANOVA, p -value > 0.05) or between phase 1 and phase 2 (T -test, p -value > 0.05).

Fully priming (nutrients and heat) kelp sporophytes resulted in higher growth rates than single priming treatments (Figure 13A), and all priming led to higher dry weights than the control (Figure 13B). All factors were significant, and there was an interaction between factors when comparing relative growth rate (glm; p -value < 0.01), with a significantly higher rate of growth for the fully primed mini-spools (mean = 0.478 % d −1) compared to only heat (mean = 0.266% d −1) or nutrient primed mini-spools (average = 0.303% d −1; Figure 13A). There were significant differences across treatments for final dry weight (linear model; p -value < 0.001). The unprimed treatment (control) had the lowest final dry weight out of all treatments, with the fully primed and nutrient primed treatments weighing over 2x more (Figure 13B). All factors were significant for average blade count per 15 cm (glm; interactive p -value < 0.001) with the unprimed and fully primed treatments (phase 1) having more blades per 15 cm (Figure 13C). The average Fv/Fm also had a significant interaction between the factors (p -value < 0.001), with the fully primed and unprimed treatments (phase 1) having ~8 times higher blade counts (Figure 13D). Post-hoc analyses found an effect of phase for each metric.

There were not significant differences between treatments for percent carbon (linear model, p -value > 0.05), percent nitrogen (linear model, p -value = 0.1063), or carbon to nitrogen ratio (linear model, p -value = 0.405; Figures 14A–C). Post-hoc analyses found a phase effect for the carbon to nitrogen ratio; however, all values were between 5.9 and 7.4 (difference of 1.5). All factors were significant and there was a significant interaction between factors for δ 13 C (glm, interactive p -value < 0.001). The heat primed and nutrient primed (phase 2) had higher values than the other treatments (Figure 14D). Post-hoc analysis confirmed a phase effect for δ 13 C (Kruskal Wallis, p -value < 0.001). δ 15 N values were positive for the phase 2 treatments and negative for the phase 1 treatments (Figure 14E). A Kruskal Wallis comparing phase 1 with phase 2 found significant differences (p -value < 0.001).

We make several important contributions toward understanding the temperature tolerance and capacity for priming in M. pyrifera at the juvenile sporophyte stage. First, fully priming this species in a hatchery setting resulted in overall increased performance during a MHW, lending support that priming can be a viable intervention for aquaculture practitioners growing this species. Second, temperatures above 20 °C had a significant negative impact on sporophytes during this early stage. Furthermore, our growth, survival, and physiological findings generally support the current outplanting strategy in California of waiting for temperatures to reach 15 °C but suggest cooler conditions (closer to 10 °C) would be more favorable. These findings broaden our understanding of this species and support further research on priming seaweeds for climate stressors.

Fully priming (nutrients and heat combined) M. pyrifera mini-spools during the hatchery stage resulted in significantly higher coverage and relative growth after the simulated 18 °C MHW, providing the first supportive evidence that priming could be a climate tool for aquaculture and restoration of the species. Although a different life stage and species, our results are similar to (Gauci et al. 2024), where authors found that Saccharina latissima (sugar kelp) sporophytes, another brown kelp common in aquaculture, grew 30% faster when gametophytes were primed for 4 weeks before spores were created. Gauci et al. (2024)′s results lend support to potentially longer lasting priming, possibly through epigenic DNA-methylation; though this was not explicitly tested. Our priming was considerably shorter—days instead of weeks—which is more common in the literature (Jueterbock et al., 2021). We also focused on juvenile sporophytes given the dearth of information on this critical transitional stage and the uncertainty of developmental reset. Even so, we still found comparable growth (average of 0.35% day −1 across treatments) to Gauci et al. (2024), almost a 15% increase in spool coverage for fully primed spools, and better evenness across the spools, indicating it was not a function of just selection of the best growers. Gauci et al. (2024) did not, however, test nutrients, which we found was likely important in improving tolerance of our species, similar to Fernández et al. (2020) and in support of the nutrient rescue hypothesis in M. pyrifera. The nutrient rescue hypothesis states high nutrient availability can alleviate impact from stressors, allowing individuals to compensate for the energetic cost of the stress (Aristi et al., 2016; Simonin et al., 2018).

For both the kelp performance experiment and kelp priming experiment, the nitrogen levels (5.3% ± 0.18) were well above reported values in other studies of M. pyrifera (Wheeler and North, 1981; Fernández et al., 2020, 2021), a function of a farmed vs. wild setting. Bunting et al. (2024) reported nitrogen averaged below 1.8% across all treatments. Our experiment surpassed nitrogen levels above the saturation point found by Wheeler and North (1980), where external nitrite exceeded 25 μg of nitrogen L −1. This was true across both experiments, which each had 1 week with hatchery level nutrients or half-hatchery level nutrients and no added nutrients during the outplanting phase. The increased level of nitrogen may have contributed to overall improved thermal performance of this species, as other research on M. pyrifera adult sporophytes also found nutrients to enhance growth and photosynthesis at higher temperatures (Fernández et al., 2020). This suggests that the industry standard has more than enough nitrite for M. pyrifera hatcheries, with levels remaining elevated for at least 10 days post outplanting and even up to 2 weeks (Gerard, 1982). However, further research should be conducted to determine which species of kelp experience nutrient rescue at higher temperatures, as increased tolerance does not appear to be ubiquitous. Elevated nutrients did not buffer the effects of high temperatures for gametophyte or early sporophyte stage Nereocystis lutekeana (bull kelp), a kelp that can be found from central California up into Alaska (Weigel et al., 2023) and has seen significant declines caused by MHWs (Rogers-Bennett and Catton, 2019). Another study on the sporophyte blades of N. lutekana and S. latissima found added nitrogen did not improve growth at 21 °C for either species, and in fact found increased biomass at lower temperatures (13 and 16 °C) when grown under “low” nitrogen conditions (Fales et al., 2023). Timing of nutrient addition might also play a role, as Fales et al. (2023) found that blades grew more biomass across temperatures when sourced from a site with previously higher dissolved inorganic nutrients. This suggests a possible lag effect and future studies should focus on the timing of nutrient addition on temperature amelioration.

We found temperatures above 20 °C had major negative impacts across all measures, affecting growth, survival, and physiology of juvenile M. pyrifera sporophytes at the individual and population (spool) level. Although not the focus, we also seemed to find a temperature mediated interaction between juvenile kelp and cyanobacteria that has important implications for wild and farmed systems under climate change, which should be investigated further. At the highest temperature treatments (25 and 30 °C) cyanobacteria growth was found across the spools. Few blades were able to survive at temperatures above 20 °C, these blades had reduced photosynthetic capacity, were smaller in size, and showed clear evidence of bleaching. Similar temperature thresholds have been reported at other life stages of M. pyrifera. For example, in New Zealand, gametophytes experience inhibited fertilization above 18.8 °C and complete secession of fertilization at 23.6 °C (Le et al., 2024). Similarly, a study on populations in California and Chile found gametophytes survived at 20 and 18 °C, respectively but Chilean gametophytes produced no eggs, while Californian gametophytes produced some eggs, but no sporophytes developed (Hollarsmith et al., 2020). A study modeling future climate conditions predicts precipitous declines in the future at southern and northern California sites, with complete extirpation of M. pyrifera at the southern site by 2100 (Alanis et al., 2021). Furthermore, the thermal optimum (based on growth and photosynthesis) for adult sporophytes range between 14 and 19 °C, with CTmax of 26 °C (Fernández et al., 2020). In our study, however, even 20 °C resulted in some metrics declining (coverage, growth rate, dry weight, surface area, height and width). These findings provide evidence that (unprimed) juvenile sporophyte sensitivity maybe more similar to gametophytes than their adult sporophyte counterparts. Given this stage is the one that experiences real ocean temperatures first in aquaculture and restoration settings, this higher sensitivity will be important in considering the climate change consequences on the species in the future.

While our findings of the performance curves generally support the current practice for M. pyrifera in California aquaculture of outplanting in November or December at 15 °C (Ocean Rainforest personal correspondence), the increasing occurrence and intensity of MHWs in the region highlights impending threats (Chan et al., 2024). We did find the proportion of mini-spool coverage, the relative growth rate, and the average blade count to be highest at 15 °C; although some measures were equivalent or better closer to 10 °C (e.g., RGR, final dry weight, Fv/Fm), suggesting cooler temperature are better for this stage. Notably, these were unprimed measures and we found metrics began to decline even under 20 °C. Chan et al. (2024) recently assessed spatial and temporal MHW patterns in the Santa Barbara channel, finding anomalous thermal conditions occur year around, including 2–3 °C average above normal in the winter months in more years. In fact, the average near-surface conditions during November and December are trending above 15 °C, reaching more sustained levels closer to 18 °C during the 2015 blob; likely again during the 2026 “super El Nino” (Acosta et al., 2026). Long term, Fong et al. (2024) projected a decrease in seaweed aquaculture performance under future climate conditions, however these projections do not account for climate extremes or changes in farming practices. Thus, as ocean temperatures continue to rise and heatwaves become more common, maintaining the current outplanting schedule will likely need to be reassessed. In the near future, outplaning may need to be delayed into January and/or other adaptive measures adopted, including priming. Genetic selection could theoretically help (Charrier et al., 2015), but is currently prohibited in the West Coast of the U.S. and thus presently not a viable option unless policies change.

We found that blade color directly reflects the performance of photosynthesis in M. pyrifera and could potentially be used on kelp farms as an indicator of health. The measurable lightening of color intensity we observed in the higher temperature treatments corresponded to a threshold response with significantly lower Fv/Fm when difference in color intensity exceed 22. As it may be more expensive and time consuming for farmers to directly test photosynthesis, color grading could be a simple and non-invasive way for farmers to assess their crop over time via visual inspection. Color grading is extremely common in terrestrial agriculture, including for strawberries, tomatoes, apples, and soybeans, to name a few (Syahrir and Connsynn, 2009; Liming and Zhao, 2010; Alfatni et al., 2011). In fact, Nori (Pyropia spp.)—seaweed commonly used in sushi for direct human consumption—has a color system in place (SushiSushi, n.d.). Color scores such as these are also used in assessing coral health, both by researchers and public science initiatives (Marshall et al., 2012; Hackerott et al., 2025). Thus, a simple color chart could be developed from our work and used for a quick visual assessment tool, with the added benefit of knowing the relationship to photosynthetic performance.

Despite no significant differences in bulk percent C or N, the temperature-driven changes in C:N and stable isotopes in the performance experiment provide evidence that thermal stress alters carbon acquisition strategy rather than total nutrient content, and priming may be beneficial. The elevated C:N and higher δ 13 C at 15 °C suggest more efficient inorganic carbon uptake consistent with active carbon-concentrating mechanisms (CCMs) under near-optimal conditions (Raven et al., 2002; Sterner and Elser, 2003). In contrast, the more negative δ 13 C values at 25–30 °C suggest increased isotopic discrimination (fractionation) during photosynthesis, associated with greater reliance on passive CO 2 diffusion and/or internal CO 2 recycling as CCM efficiency declines under heat stress (Raven et al., 2002). In fact, all but one replicate in the 20–30 °C range had δ 13 C < −30, which falls into the third category described by Velázquez-Ochoa et al. (2022), CO 2 diffuse entry. The majority of replicates from the 5–15 °C treatments had lower δ 13 C values, falling into category 2 (−10 < δ 13 C > −30), using a mix of CO 2 and HCO 3 - uptake (Velázquez-Ochoa et al., 2022). This mechanistic interpretation aligns with Bunting et al. (2024), who demonstrated that marine heatwave intensity and duration suppress growth in young M. pyrifera, likely reducing CCM operation and tissue turnover, a process further supported here by elevated δ 1 5N at thermal extremes, indicative of stress-limited nitrogen assimilation (Evans, 2001; Bunting et al., 2024). Yet, compared to the performance experiment isotopic signatures, we found less negative δ 13 C and stable %C, %N, and C:N across treatments, indicating that priming did not substantially alter overall nutrient balance. These results do not necessarily confer priming buffers to negative CCM impacts from heat stress, but rather there is no added burden and reductions in efficiency occurs at the most extreme temperatures (>20 °C). In all, marine heatwaves push kelp into a low-efficiency metabolic state by suppressing growth and carbon-concentrating mechanisms, which can undermine the resilience of kelp forests in the absences of buffering processes, that may include priming.

While our study was limited due to space and time with confounding factors arising from our two-phase approach, it was an important step forward for seaweed priming. We did find evidence of a “phase effect” for several of the metrics (e.g., RGR, average Fv/Fm) that confounds some of our findings. While hatcheries have protocols in place to mitigate the chances of a bacterial or disease outbreak, this presents a real-world issue within seaweed and other aquaculture ventures (Rehman et al., 2017; Ward et al., 2020). Although we did not see continued bacteria growth once phase 2 began at UC Santa Barbara, it is possible there were fewer sporophytes during this phase due to early impacts from the bacteria or possibly self-thinning—where small individuals crowd each other as they grow, and progressively die as competition increases (Scrosati, 2005)—as these blades were slightly larger than phase 1 blades. Increased Fv/Fm, although minimal between phases, could also be a result of differing age (Dring et al., 1996). Regardless, while the exact effect size of the fully vs. just heat primed spools has higher uncertainty, the evidence is in favor of priming. Indeed, even with this phase effect, heat priming did not have a negative impact on any metric, underscoring that priming is still beneficial. For M. pyrifera, the next steps should focus on optimal duration of a priming, recovery, and timing of the trigger. Our trigger event was soon after the priming stress, but the best-case scenario for seaweed aquaculture practitioners would be if the priming stress contributed to longer-term (outplanting to harvest = 3–6 months) resilience against MHWs. Studies should also focus on epigenetic testing to understand the mechanisms underlying priming for seaweed species, as this has huge implication for the longevity of the priming effects (Jueterbock et al., 2021). Ultimately, determining the best and most feasible protocols to ensure future resilience in a real-world setting will give aquaculture practitioners confidence in applying these techniques in their own hatcheries.

Our findings suggest that adding a heat and nutrient priming intervention during the hatchery stage may improve outcomes for farms that experience marine heatwaves after outplanting. Importantly, the “high nutrient” priming treatment we used in the study matches the industry standard for seaweed aquaculture, and thus adding a heat priming intervention would be a relatively low effort and expense during the hatchery stage. As noted, this is a promising development as nutrient rescue has not been successful in N. lutekeana (Weigel et al., 2023). While further studies should explore how long priming in M. pyrifera will result in heatwave protection and if other kelp species can be successfully primed, the applications for priming M. pyrifera could extend outside of aquaculture and into kelp conservation efforts.