Back to Articles List

Garden Under Stress – Correlating Plant Deaths and Weather in a Red Hill Garden

By Ros Walcott

Aug, 2026

---

I have recorded details of each plant in our garden, such as month planted, short description of each plant, plant health and month of death or removal of plant and the reason for it, from January 2004 until the present, just over 8,000 plants. We have removed many plants in our garden that grew too big for their position or became unattractive. Two thousand plants were planted in our garden from June 2003-December 2003 in our original planting by Able Landscaping, making a total of just over 10,000 plants. Over recent months I have gathered details of ‘month planted’ and ‘month died or removed’ into Excel files. I have also recorded the rainfall in our garden month by month over the same period.  This data is in tables at the end of the text.

I want to correlate weather patterns, such as drought and rainfall, with plant deaths. Thanks to the application of AI (Claude) to these data, I have been able to recognise certain statistical trends.

The record divides into three broad eras, a relatively quiet early period through the Millenium Drought years, a middle period driven by heat waves and a more volatile later period marked by both La Nina wet years and renewed heat stress.

Total yearly rainfall on its own is a poor predictor of total yearly plant deaths. The five driest years (2018, 2019, 2009, 2008, 2013) averaged 188 deaths; the five wettest years (2016, 2020, 2010, 2021, 2022) averaged 214 deaths — both close to the overall average of 209. Claude calculates only a weak correlation of r = 0.16.

When you look at which months tend to be wet versus deadly across the dataset, the relationship is stronger — wetter months skew toward higher death counts, (moderate correlation (r ≈ 0.48). This points toward wet-season stresses (waterlogging, root rot, fungal disease, storm/hail damage) mattering more than drought stress in this garden. August is the clearest outlier: only moderate rainfall but the second-highest death total of any month — frost damage is the culprit here. I find that plants stay green and healthy during the winter, then turn brown and die in August.

The two standout years 2022 (1,252 mm, the wettest year on record) also had the second-highest death toll (389) — consistent with flooding/fungal stress. Eremophilas in particular find the wet years problematic in our garden. I have calculated that we have lost 80% of the eremophilas we have planted, mostly in or after the wetter La Nina years. 2025 had the highest death toll (349) despite close to average rainfall (618 mm), but the culprit here was five frosty nights the last week in June 2025, with temperatures of -5 degrees or less, ( Walcott, Canberra ANPS Journal 2025).

The fact that very few deaths occur in June and July, Canberra’s driest and coldest months, demonstrate that plants are mostly not dying from the cold and frost. The exception here is 2025 as noted above. Death in the garden is mainly driven by heat and moisture levels, either drought or flood and heat waves.

The necrology tracks remarkably closely with documented climate events affecting the ACT. The major spikes are not random — each corresponds to a period of documented extreme stress.

2005

Heat and Drought 87 deaths

An early spike, occurring mid-way through the Millennium Drought (2000–2009) that drove drier-than-average conditions across the ACT for nearly a decade. Losses were concentrated in August (22) and October (25), suggesting late-season heat stress after a dry winter.

2009

Heat and Drought 155 deaths

The southeastern Australia heatwave of January–February 2009 was one of the most extreme in the region’s history, with temperatures consistently 10°C above average across the ACT and surrounding region. The garden record shows April (50) and June (31) as the peak months — reflecting plants that were mortally stressed by the summer heat but only declared dead by autumn, a common lag pattern in woody and perennial species.

2011

Floods and La Nina 174 deaths

The first of two La Niña events (2010–11 and 2011–12) that broke the Millennium Drought brought exceptionally wet and cool conditions. The spike in April (62) likely reflects waterlogging, root rot, and cold damage in the wetter-than-usual autumn — a paradoxical but well-documented pattern where excess moisture kills drought-adapted species. January (23) and February (35) also elevated, possibly from summer flooding events.

2013

HEAT 195 deaths

The summer of 2012–13 was the hottest on record for Australia at the time, with a continent-wide heatwave peaking in January 2013. The garden’s February peak (79 deaths) is the highest single-month total for any month before 2022, suggesting massive losses in heat-sensitive species following the January stress event. March (40) extends the damage. 2013 was Australia’s hottest year on record at the time of recording.

2014

HEAT 245 deaths — the highest pre-2022 total

The summer of 2013–14 followed directly on from 2013’s record heat. Back-to-back extreme summers compounded stress on plants that had never fully recovered. August (65) is a notable spike — mid-winter losses at this scale suggest plants that were severely weakened from summer damage but collapsed during the cooler months, a pattern consistent with species that exhaust root reserves trying to recover from heat stress.

2017

HEAT 232 deaths

Part of the drier-than-average stretch from 2013–2019 (with the exception of 2016). January (46) and February (53) point to another summer heat event, with August (52) again showing the mid-winter lag of summer-weakened plants.

2022

FLOODs TRIPLE LA NIÑA 330 deaths — the record year

The most lethal year in the record, coinciding with a rare triple-dip La Niña (2020–21, 2021–22, and 2022 spring). Canberra broke its October rainfall record in 2022, reaching 163.4mm against an average of 62.4mm — nearly triple the norm. The data tells a distinctive story: September (60), October (88), and November (59) form a catastrophic spring cluster — the inverse of the heat-stress pattern. These are almost certainly drowning losses: waterlogging, fungal disease, and root anaerobia in species adapted to Canberra’s normally dry springs.

2025

HEAT 212 deaths

January (50) and September (60) as the peaks suggest a double-stress year: summer heat followed by a wet spring. Australia recorded its fourth-warmest year on record in 2025, following its second-warmest in 2024. However there were many frost-related deaths from the very cold last week of June 2025, with temperatures of -5 degrees over five nights.

2026

HEAT 279 deaths (Jan–Jun only)

With data only through June, 2026 is already tracking at 279 deaths — which would be the second highest year on record if the second half tracks near average. March alone recorded 186 deaths, a single-month figure that dwarfs anything in the previous 22 years.

 

The conclusion is inescapable that our garden is under increasing stress. The weather, with climate change, is more variable and more extreme and is becoming even more extreme with time. In the first five years, 2004-2008, the average annual death toll was 57. In the most recent five years, 2019-2023 the death toll was 137, more than double. The pattern of planting has not changed very much, apart from the spurt recorded in the first six months. Since 2004 we have planted an average of 357 plants per year, so the doubling of the average annual death toll is significant.  The palette of plants we use has stayed experimental over the years. We have not planted ‘tried and true’ plants in the garden but rather used plants from every part of Australia to see if we could get them to thrive in Canberra. Our main planting groups are Acacias, Banksias, Brachychitons, Callistemons, Correas, Eremophilas, Eucalypts, Grevilleas, Hakeas, Leptospermums, Melaleucas, Prostantheras and Westringias.

We are now facing possibly the most severe El Nino event for decades in 2026 Australia. We must be prepared, as gardeners, to plant wisely and to expect losses in the garden under these severe and uncertain conditions.

The following are links to the data tables

Plant deaths, removals:https://nativeplantscbr.com.au/wp-content/uploads/Plant-Deaths^J-Remvals-Month-and-Year.pdf

Planted month and Year:https://nativeplantscbr.com.au/wp-content/uploads/Planted-Month-and-Year-21.pdf

Rainfall: https://nativeplantscbr.com.au/wp-content/uploads/Rainfall-2026.pdf

 

---

Back to Articles List