Early Survival and Growth of Climate-Ready Trees and Native Oaks in an Urban Canopy Planting Program

Study Site and Community Canopy Tree Planting Program

The study was conducted in Davis, California, USA, a small city in California’s Inland Valley climate region with a population of around 67,000. Davis has hot, dry summers and cool, wet winters. Precipitation falls almost exclusively between the months of October and May, with an average annual precipitation of 48.3 cm over the last 30 years (1995 to 2024)(PRISM Group 2025). The region experienced a pronounced drought for the years 2020 to 2022 (National Centers for Environmental Information 2026).

In 2019, the City of Davis received a grant for $840,000 from the California Department of Forestry and Fire Protection (CAL FIRE) to implement a tree planting program in partnership with Tree Davis, the local urban forestry nonprofit. The funded program, called the Community Canopy program, aimed to plant 1,000 new shade trees, about half of which would be climate-ready species and half more commonly planted urban tree species and native oaks. Funding included stewardship activities (e.g., inspecting, pruning, watering, mulching, adjusting stakes) for 5 years after trees were planted.

In preparation for the grant, a group of stakeholders including City of Davis urban forestry managers, nonprofit representatives, and researchers worked together to produce a species palette that would satisfy a variety of landscapes. The group developed a list that included some well-known, commonly planted species in the area that were widely available at nurseries, which we refer to as “common” trees, as well as the area’s 3 native oak species. The list also included a set of climate-ready trees that were selected based on several trials in Davis and nearby Sacramento, California, USA: a drought-tolerant tree trial, started in 1999 (McPherson et al. 2017); the National Elm Trial, started in 2005 (McPherson et al. 2009); and the Climate-Ready Trees experimental field trial (McPherson et al. 2018), which was started in 2015 and underway at the time of species selection. For the Community Canopy list of climate-ready trees, the group selected some species that had been well tested and proven to perform well and several species that were in the process of evaluation or considered promising (McPherson et al. 2020). The full list of selected species is shown in Table 1. While acknowledging that each species has unique properties, we use the groupings of climate-ready trees, native oaks, and common species to allow statistical testing across a range of settings and draw some generalized comparisons.

Over 1,000 Community Canopy trees were planted between October 2019 and March 2023. All trees were 15-gallon (56.78-L) container stock when planted, purchased through a local retail nursery that sourced trees from different wholesale nurseries across the Western United States. Trees were all mulched at planting and remulched and weeded as determined necessary from monitoring visits, sometimes with the help of volunteers. Some trees also received supplemental water during the summer and periods of drought, with varying frequency depending on site context and monitoring. Trees in unirrigated, publicly managed spaces were hand-watered approximately weekly during water-scarce periods for the first 2 years, with further watering on a less regular basis, while trees planted in lawns and those with homeowner stewards were only hand-watered if a monitoring visit suggested they were not receiving enough water. Trees were also pruned as determined necessary from monitoring visits, with approximately 40% of trees receiving at least one pruning before the end of the study period. Any trees found to be dead during watering and maintenance visits were replaced during the next fall/winter planting season. In addition, trees have been systematically monitored for survival and performance beginning in 2021, as described below.

Data Collection

Near the end of the growing season in 2021, 2022, and 2023, we visited all planted Community Canopy trees to determine mortality status. In 2021 and 2023, we also took measurements of diameter at breast height (DBH) and recorded information about each tree’s location (Table 2) and condition (Table 3). DBH was measured at 1.37 m when possible, and measurements in 2023 were taken at the recorded height for 2021 when possible. Each tree received a separate score for the condition of its trunk, scaffold branches, small branches, roots, and foliage (see Table 3 for scoring criteria), and these scores were summed for a total condition score. Data were collected from 2021 August 26 to 2021 September 17; 2022 August 1 to 2022 October 1; and 2023 September 1 to 2023 November 2.

Statistical Analysis

All analyses were performed in R version 4.4.2 (R Core Team 2024). We calculated an overall survival distribution for all Community Canopy trees using Kaplan-Meier survival analysis with Turnbull estimator for interval-censored data (Fay and Shaw 2010). Interval-censored data are generated when survival status is only observed on specific data collection dates and the exact date of death is unknown. We also used permutation-based weighted logrank tests to compare survival distributions for climateready trees, native oaks, and common trees, as well as trees growing in different landscape contexts and site types, and with and without irrigation systems. To account for multiple planting and data collection dates in our study, we constructed survival distributions using the number of days between planting and observation for each tree. Replacement trees were recorded as new individuals. There were 20 trees marked as dead or unknown in 2021 and/or 2022 that were found alive in 2023, either resprouting or located slightly away from their recorded coordinates, and we assumed that these trees had been alive from their date of planting. Survival analyses were carried out with the interval package (Fay and Shaw 2010).

Because one of the prospective merits of climateready trees is their performance in low-water landscapes, we investigated whether the survival of different tree types was related to site irrigation status. Site irrigation was identified as a very important predictor of overall mortality in the survival analysis, but that approach only allowed testing outcomes for tree types and site irrigation independently. We used a regression approach to test the interaction of these variables and assess the performance of different tree types while accounting for their site irrigation status. We used a binomial generalized linear model with mortality status in 2023 as the response variable and tree type (climate-ready, native oak, and common), planting year (to account for different tree ages and cohorts), and site irrigation as predictors. We included the interaction between tree type and site irrigation, and also tested the interaction between tree type and year planted because oak tree container stock was observed to be of poor quality in the first year. Non-significant (P > 0.1) interactions were removed from the final model. Given that the study design was not balanced, the regression approach provides important context for the survival analysis results. For trees that were planted before data collection in 2021, we also used this regression approach to model the effects of total condition score and DBH in 2021 on mortality status in 2023. Tree type and site irrigation were also included as predictors. For both models, assumptions were checked with the DHARMa package (Hartig 2024), pairwise contrasts were evaluated with the emmeans package (Lenth and Piaskowski 2026), and predictor effects were calculated using the effects package (Fox and Weisberg 2019).

We also investigated whether 2023 total condition scores were related to site factors and 2021 condition scores. For trees with both 2021 and 2023 data, we used Spearman’s rank correlation tests to evaluate the relationship between total condition in 2023 and total condition and DBH in 2021. For all trees that were alive in 2023, we used Kruskal-Wallis rank sum tests to determine differences in total condition score for different tree types, landscape contexts, site types, and irrigation status. For tests with significant results, we conducted multiple comparison tests to determine significant differences between groups using the pgirmess package (Giraudoux 2024). We used non-parametric tests for these analyses because the data did not satisfy assumptions for parametric tests.

Finally, for trees with 2021 data that were still alive in 2023, we determined growth as the change in DBH between 2021 and 2023. We used a linear regression model to investigate whether growth differed by tree type and whether it was influenced by irrigation status, condition score in 2021, and size (DBH) in 2021. The model also included an interaction between tree type and irrigation status, as preliminary investigation suggested that oak growth had a different response to irrigation than the other types of trees. Variables were transformed as necessary to meet model assumptions, and pairwise contrasts within the interaction were evaluated with the emmeans package (Lenth and Piaskowski 2026). We removed 38 trees from this analysis that had missing data or a smaller DBH in 2023, generally due to resprouting.

Site Irrigation Increases Survival and Growth for Most Trees

Despite the supplemental hand watering that many Community Canopy trees received, the presence of landscape irrigation had a strong influence on survival and growth. This finding is in line with other studies linking irrigation to the performance of young trees. In particular, Gilman et al. (1998) found that frequent irrigation improved the establishment and growth of newly planted container-grown trees, while Koeser et al. (2014) found significantly greater survival and growth for urban trees planted in sites with irrigation systems. Numerous other studies point to consistent watering as an important component of urban tree survival and health, particularly in areas with limited precipitation (Roman et al. 2015; Vogt et al. 2015b; de Guzman et al. 2018). Watering may be the most important factor in the early success of many urban trees in the study area, and the importance of site irrigation was likely amplified by the severe drought conditions from 2020 to 2022. Community Canopy trees in unirrigated sites may have performed better if they had received more consistent, frequent, or extended hand watering, and supplemental water is likely to become an even more important factor with the increased drought severity expected with climate change in this region (Houlton and Lund 2018).

Site irrigation was associated with higher growth rates for common and climate-ready trees, yet the native oaks that survived from 2021 to 2023 grew more in unirrigated settings than in irrigated settings. Ours is not the first study to find that irrigation does not increase the growth of these particular oak species—a study that experimentally manipulated post-establishment irrigation levels on the same 3 native oaks found no difference in growth related to irrigation in the first 4 years after planting (Costello et al. 2005). The improved growth we observed in unirrigated settings may be related to an unmeasured relationship with site properties. Some of the unirrigated sites where native oaks were planted were natural areas that may have had less soil compaction and could have boosted aboveground growth of transplants through increased ectomycorrhizal presence in the soil (Berman and Bledsoe 1998). On the other hand, many native oaks were planted along a bike path where the soil was disturbed and compacted. Another possible explanation for our results is that the relatively shallow irrigation typical of urban landscapes could be detrimental to native oak growth, particularly Quercus lobata (Callaway 1990). After establishment, young native oaks may thrive on the less frequent and deeper supplemental water they received in unirrigated areas.

Climate-Ready Trees Had the Best Overall Performance

As a group, climate-ready trees performed comparatively well: their early survival, condition, and growth surpassed other trees in both irrigated and unirrigated settings. This successful early establishment and growth suggests that planting climate-ready trees could be a particularly good investment for future canopy cover. Although continued monitoring will be necessary to determine whether these trends continue or there are trade-offs to robust early growth, our study finds that these trees are not only surviving but are also reaching larger sizes more quickly. Furthermore, they are likely to continue to perform well under changing climate conditions and reduced landscape irrigation (McPherson et al. 2020). Our results suggest that the process for identifying and evaluating climate-ready trees (McPherson et al. 2018) has been a successful selection strategy overall.

However, there was variation in performance within the group of climate-ready trees, and differences in mature size among species will also influence long-term canopy provision. Smaller climate-ready trees, such as Cercis canadensis var. texensis ‘Oklahoma’ and Chilopsis linearis, will not provide nearly as much shade at maturity as larger trees like Ulmus ‘Morton Glossy’ or Q. buckleyii. These larger trees also had faster growth rates. One climate-ready tree species, Parrotia persica, did not perform well in terms of survival or growth. This species was the least tested of the group and may not merit further planting in this area. Thus, while considering the performance of categories of trees provides some useful insights and allows for statistical testing, the performance of individual species must ultimately be considered for planting plans.

In the same vein, some common trees had excellent survival and growth rates as well, and the performance of the group as a whole was diminished by the relatively poor performance of two species—Ginkgo biloba had one of the slowest growth rates and nearly a 25% mortality rate, which was exceeded by the nearly 60% mortality rate for Q. suber. The other common species would still be good choices for irrigated settings, and a slightly different palette of common species could have yielded results equivalent or superior to the climate-ready trees included in this project.

Native Oak Tree Performance Was Influenced by Irrigation and Nursery Stock Quality

Although native oak trees had a lower overall survival rate, they performed nearly as well as climateready trees when accounting for the fact that they were more commonly planted in unirrigated settings. The contrast we observed between the negative impact of site irrigation on native oak growth and the positive impact on survival could be explained by the importance of consistent water availability for very early establishment. Costello et al. (2005) found no benefits of irrigation for oak growth only after the trees had first been established for one year with frequent, high levels of irrigation, and Young and Evans (2005) found substantially higher survivorship for container-grown Q. lobata seedlings when they were irrigated after planting.

Another factor that probably contributed to higher mortality for native oaks was poor container stock quality noted in the first year of planting. This problem was noted only for native oak trees, and after the program switched to a different vendor with better container stock for these species, survivorship for native oak trees increased. Container-grown native oaks are generally prone to root problems (Young and Evans 2005), and it is likely that the coupled effects of water stress and poor condition at planting contributed to the relatively high failure rates observed for Q. lobata and Q. douglasii.

These higher failure rates could also reflect a climate that is becoming less suitable for native oak trees. However, while the continued climate suitability of native oaks in this area is a concern (Kueppers et al. 2005), mortality during establishment is typically high for these species in their native ranges (Tyler et al. 2006), and the constrained root architecture of container-grown saplings is likely to make them even more vulnerable to water shortages than saplings that have grown in situ (Young and Evans 2005). Research has also shown that for Q. lobata, saplings have a narrower climate envelope than mature trees (McLaughlin and Zavaleta 2012), suggesting that if the Community Canopy oaks survive to maturity, they have a better chance of withstanding changing climate conditions.

Early Condition Influences Growth and Survival

The importance of using high-quality nursery stock was further highlighted by our finding that tree condition in the first year or two after planting was predictive of survival, growth, and condition two years later. Other studies have also shown a strong relationship between condition—particularly foliar or crown condition— and survival for small trees (Roman et al. 2014b; Vogt et al. 2015b; Bigelow et al. 2024). We also found that initial size had a positive relationship to both condition in 2023 and growth rate, although it did not significantly affect survival. Vogt et al. (2015b) also found no significant effect of initial size on survival but a strong negative impact on growth rate, while Koeser et al. (2014) found that trees with greater initial calipers had greater growth rates. The effect of initial size on growth rate may depend on species, nursery stock, and site conditions.

The most common condition problem for young trees in both 2021 and 2023 was their foliage. We collected data at the end of the season, when foliage problems are most likely to have developed. The issues we observed were typically browning leaf edges, although damage from elm leaf beetles was also common for the climate-ready elm trees. However, low foliage scores became less prominent between 2021 and 2023, suggesting that some trees can recover from early foliage issues. This recovery may have been facilitated by a much wetter winter in 2022 to 2023 compared to the drought conditions preceding the summer of 2021. Another encouraging finding about foliage condition comes from a study that considered multiple size classes of urban trees and found that poor foliage condition was no longer related to survival as trees became mid-sized (Roman et al. 2014b). On the other hand, we found that more trees developed issues with their trunks over time, often due to injuries from landscaping equipment. Lower trunk damage can significantly reduce growth in urban trees (Vogt et al. 2015b), underscoring the importance of marking and protecting young trees in managed landscapes.

For about a third of trees, issues with roots contributed to lower condition scores. Some of these issues likely stemmed from poor root condition in containers (Gilman et al. 2015), but most could be attributed to poor planting practices where the tree was planted too low or too high relative to the ground surface. Deep planting can be particularly harmful to trees, as it has been associated with increased development of girdling roots and decreased survival in some species (Wells et al. 2006). A focus on training and supervision of volunteers during planting could help to avoid these problems.