Blackberry and Raspberry Cold Hardiness Update in the Pacific Northwest

A line graph titled “Blackberry Cold Hardiness – Sakuma, Victoria, 2025–2026.” The x‑axis shows dates from early September through January 19. The left y‑axis shows temperature in °F ranging from –20°F to 70°F. The right y‑axis shows an LT₉₀ scale ranging from roughly –20°F to 0°F. A vertical black line labeled “Today” marks the current date near January 19. Four time‑series lines are plotted: Blue line: Minimum 2‑meter air temperature, fluctuating widely between about 25°F and 60°F. Yellow line: LT₁₀, staying near 20–30°F with small variations. Orange line: LT₅₀, ranging roughly from 5–15°F. Red line: LT₉₀, ranging from near 0°F to –10°F, showing more variability than the other hardiness lines. Dotted extensions of all four lines appear just past the “Today” marker, indicating projected values. The background is lightly shaded green, and a color‑coded legend below the chart describes cold‑risk categories: No Risk (green), Moderate (yellow), High (orange), and Extreme (red). The bottom right corner cites “Source: AgWeatherNet.”
Figure 4. Screenshot of new cold hardiness model.

Volume 14 Issue 1

Authors: Pedro Rojas-Barros, Emma Rogers, Clark Kogan, and Lisa Wasko DeVetter

Cold hardiness is a dynamic process that strongly depends on air temperatures. As temperatures cool down in the fall, buds gradually acclimate and gain cold hardiness. In contrast, as temperatures rise in the late winter and spring, buds deacclimate and start losing hardiness. Deacclimation can occur during the winter in response to warm spells, although buds may partially reacclimate if cold temperatures return.

The ability of buds to reacclimate depends, in part, on chill accumulation. Chill is typically defined as the accumulation of temperatures between 32 and 42 °F. Once sufficient chill has been accumulated, buds transition into a “ready-to-grow” stage, and their capacity to regain hardiness after warm periods becomes limited. 

Because these processes vary among species, cultivars, and environments, it is important to monitor cold hardiness across different conditions. This information helps us move closer to our long-term goal of developing a regionally relevant predictive cold hardiness tool that growers can use to better manage freeze risk.

What We Measured

We estimated cold hardiness using a laboratory-based freeze assay, where buds are exposed to controlled freezing temperatures and then scored for damage (discussed in “Chilling Insights -Blackberry Cold Hardiness in the Pacific Northwest,” September 2025) . From these tests, we calculate the temperature at which 50% of fruiting buds are damaged (LT50). More negative LT₅₀ values indicate greater cold hardiness.

This season, we monitored:

  • Blackberry: ‘Victoria’ and ‘Celestial’ [Skagit County and British Columbia (BC) sites]
  • Raspberry: ‘Kulshan’ and ‘Meeker’ (Whatcom County)

Samples were collected every week for the Skagit County samples and every two weeks for BC and Whatcom County samples, allowing us to track how cold hardiness changed over time.

Comparing Two Seasons in Blackberry

When we compare the current Fall-Winter season to last year for ‘Victoria’ and ‘Celestial’ blackberry growing in Skagit County, several important patterns emerge.

1.     Fall acclimation:

Sample collection in the 25-26 season started in early September instead of mid-October and we observed that buds of both cultivars gradually gained cold hardiness from early September through November as temperatures cooled. The LT₅₀ values during overlapping time periods were similar compared to the previous 24-25 season in both cultivars. 

2.     Early-Winter temperatures and cold hardiness

The 24-25 season was colder than the current season, with greater chill accumulation and a higher frequency of freezing temperatures (Fig. 1). Chill accumulation progressed more steadily through mid-winter during the 24-25 season, whereas accumulation slowed during the 25-26 season, coinciding with warmer temperature periods (Fig. 1A). The distribution of daily mean temperatures was also shifted toward colder values in 24–25, while the 25–26 season exhibited a higher frequency of warmer daily means (Fig. 1B). In contrast, the 25-26 season to date has been characterized by lower chill accumulation and fewer freezing events, although freezing temperatures were more intense when they occurred (Fig. 1C). Freeze events in 24-25 were more frequent but generally shorter, whereas the 25-26 season experienced fewer events with longer durations and lower minimum temperatures (Fig. 1D). 

Regarding cold hardiness, during the 24-25 season, maximum cold hardiness in ‘Victoria’ was reached in late January (LT50 = 3.0 °F), while ‘Celestial’ reached a LT50 of -7.6 °F during the same period (Fig. 2). During the 25-26 season, maximum cold hardiness in ‘Victoria’ was reached earlier, in late December, and a similar pattern was observed in ‘Celestial’. However, a warm spell during January affected the two cultivars differently. ‘Victoria’ exhibited a sharp deacclimation during mid-January, whereas ‘Celestial’ was less affected. [LD2] The mid-January freezing temperatures observed in the past week promoted a quick reacclimation in ‘Victoria’ (LT50 = 3.0 °F), while ‘Celestial’ continued showing similar LT50 values as previous weeks.

3.     Late winter deacclimation:

      As temperatures warmed in late January and February, both cultivars began to lose cold hardiness. During the 24-25 season, deacclimation was more pronounced and occurred earlier in ‘Victoria’, whereas ‘Celestial’ maintained greater cold tolerance for a longer period before rapidly deacclimating towards the end of winter.

A multi‑panel figure showing comparative cold‑weather metrics for two periods, labeled “24–25” in turquoise and “25–26” in red. Panel A: A line graph of cumulative chill hours (≤45°F) from October 16 to January 20. Both years show steadily increasing chill accumulation, with the 25–26 line rising slightly faster after mid‑December. Panel B: Overlaid density curves showing the distribution of daily mean temperatures. Both periods form a roughly bell‑shaped curve centered near the mid‑40s °F, with 25–26 showing slightly more density at cooler temperatures. Panel C: A bar chart showing total hours within temperature intervals (
Figure 1. Canopy air temperature conditions during the study period, including chill accumulation (A), daily mean temperatures (B), cold exposure intervals (C), and freeze event characteristics. Data were collected from an in-field sensor in Skagit County, WA.
A line graph titled “Bud Cold Hardiness (LT₅₀) Across Seasons” showing weekly changes in LT₅₀ (°C and °F) for blackberry buds. The x‑axis is “Week of collection,” ranging from weeks 1 to 26. The y‑axis on the left shows LT₅₀ in °C (0 to –21), and the y‑axis on the right shows the equivalent °F values (32 to about –6). Two cultivars are displayed: Celestial shown in solid green for 2024–2025 and dashed green for 2025–2026 Victoria shown in solid pink for 2024–2025 and dashed pink for 2025–2026 Across both seasons, LT₅₀ values decrease through mid‑winter (indicating increasing cold hardiness) and then rise again toward the end of the collection period. Celestial typically reaches lower LT₅₀ values (greater hardiness) than Victoria, especially around weeks 12–18. Dashed lines for 2025–2026 show similar seasonal patterns but with slightly warmer (less hardy) mid‑season values for both cultivars. A dotted horizontal line at 0°C is included for reference.
Figure 2. Comparison of cold hardiness between the 24-25 and 25-26 seasons, showing lethal temperature in which 50% of fruiting [LD4] buds showed damage (LT50) in ‘Victoria’ and ‘Celestial’ blackberry grown in Skagit County, WA. Lower LT50 values indicate greater cold hardiness.

Expanding the Study: Raspberry and British Columbia Sites

This season, we also included raspberry cultivars (‘Kulshan’ and ‘Meeker’) and sampling sites of ‘Victoria’ in British Columbia (Fig. 3) Bud cold hardiness increased through the fall and early winter in all cultivars, but the timing and extent of cold hardiness differed among cultivars and locations (Fig. 3). In blackberry, ‘Victoria’ showed less cold hardiness variation during the Fall across locations, but differences were evident towards the end of December, with higher cold hardiness in buds collected from Skagit County. Regarding raspberry, ‘Kulshan’ showed a similar response to cold temperatures during the Fall-Winter period compared to ‘Meeker’. By the end of January, ‘Meeker’ remained slightly more cold hardy than ‘Kulshan’ (Fig. 3).

A two‑panel line graph titled “Bud Cold Hardiness 25–26 Season.” The left panel shows blackberry cultivars; the right panel shows raspberry cultivars. The x‑axis in both panels is “Collection Date,” spanning September through January. The left y‑axis displays LT₅₀ values in °C from about –2 to –24, and the right y‑axis shows the equivalent °F scale from about 32 to –11. Blackberry panel: Multiple cultivars are shown with distinct colors and symbols: Celestial (green) Victoria (pink) Kulshan (orange triangles) Meeker (purple squares) Victoria BC Locations 1, 2, and 3 (yellow/orange dashed lines with different markers) All blackberry lines show decreasing LT₅₀ values from early fall into mid‑winter (indicating increasing bud cold hardiness), reaching lows near –18 to –22°C. Celestial and Victoria show the steepest declines. Some cultivars exhibit slight mid‑winter fluctuations, with Victoria showing a noticeable warm‑hardiness spike in January. Raspberry panel: Two cultivars are shown (orange and purple dashed lines). Both exhibit the same fall‑to‑winter pattern: LT₅₀ values decrease from early autumn into December, reaching about –14°C, then level off or slightly increase by January. A dotted horizontal line at 0°C appears at the top for reference. The legend identifies each cultivar and symbol style.
Figure 3. Changes in bud cold hardiness during the 25–26 season for blackberry (left) and raspberry (right) cultivars, including samples collected in Washington and British Columbia (BC). Lower LT₅₀ values indicate greater cold hardiness.

Blackberry Cold Hardiness Model (AgWeatherNet)

Figure 4 shows estimated blackberry bud cold hardiness based on a temperature-driven beta model. The model uses hourly air temperature data from AgWeatherNet to estimate lethal temperature thresholds (LT10, LT50, and LT90), which represent the temperatures at which 10%, 50%, and 90% of fruiting buds are expected to be injured.

Model outputs are updated automatically as new temperature data become available. The solid lines represent estimated cold hardiness thresholds, while the minimum air temperature is shown for reference. A short-term projection based on the weather forecast is also displayed to indicate potential changes in cold hardiness.

Access the Blackberry Cold Hardiness Tool:

👉 https://test.awn.cahnrs.wsu.edu/ The model is available for free but requires first-time users to create a username and password. Once logged into WSU AgWeatherNet, log into “Models” ® “Blackberry”. Be sure to select the weather station closest to your field of interest. 

Note this model is still undergoing refinement and we value your feedback. If you have any comments or questions regarding the model on AgWeatherNet, please fill out this form: AgWeatherNet New Portal Feedback – Beta testers – Fill out form

A line graph titled “Blackberry Cold Hardiness – Sakuma, Victoria, 2025–2026.” The x‑axis shows dates from early September through January 19. The left y‑axis shows temperature in °F ranging from –20°F to 70°F. The right y‑axis shows an LT₉₀ scale ranging from roughly –20°F to 0°F. A vertical black line labeled “Today” marks the current date near January 19. Four time‑series lines are plotted: Blue line: Minimum 2‑meter air temperature, fluctuating widely between about 25°F and 60°F. Yellow line: LT₁₀, staying near 20–30°F with small variations. Orange line: LT₅₀, ranging roughly from 5–15°F. Red line: LT₉₀, ranging from near 0°F to –10°F, showing more variability than the other hardiness lines. Dotted extensions of all four lines appear just past the “Today” marker, indicating projected values. The background is lightly shaded green, and a color‑coded legend below the chart describes cold‑risk categories: No Risk (green), Moderate (yellow), High (orange), and Extreme (red). The bottom right corner cites “Source: AgWeatherNet.”
Figure 4. Screenshot of new cold hardiness model.