Environmental Analogs in the Search for Stress-Tolerant Landscape Plants

Climatic factors

We can begin simplifying complex environmental relationships by examining various climatic and edaphic features in light of previous research on woody plant adaptation. Since trees have extended juvenile periods, the evolution of climatic adaptation of tree populations in nature is affected both by long-term, seasonal climatic cycles that are, to varying degrees, predictable and by the frequency and severity of relatively infrequent, extreme conditions, which may be impossible to predict without extensive data.

The annual photoperiod regimen, a direct function of latitude, is one of the most predictable climatic features. The photoperiod regimen may serve as a phenological signal to trees and shrubs for the induction of vegetative growth and flowering (22) and subsequent cessation of growth leading to autumnal cold hardening (16). [These phenomena are reviewed in detail by Salisbury (36).] Cultivation of natural populations for evaluation or direct use in latitudes significantly removed from their provenance often interferes with evolved phenological patterns, resulting in poor growth and winter injury (25,30). Attempts to cultivate natural tree populations that overlook a provenance’s latitude risk failure.

Low winter temperatures predicted from medium to long-term (20-100 year) records are widely used to relate tree adaptation to low-temperature injury (14). The most common measure, mean annual minimum temperature, serves as the basis of important hardiness zone maps for North America (6,34) and for Europe (14). Other studies have evaluated January mean temperatures in comparison with mean annual minimum temperature to predict the survival of selected woody plants in the north-central United States (50) or incorporated the monthly mean of daily minimum temperatures of the coldest month (MMDM) into multiple regression equations and maps for general woody plant adaptation in Canada (28, 29). The MMDM also was the climatic statistic that was most congruent with boundaries of forest plant communities in Florida (13).

Extreme low temperatures may be at least as important as mean low temperatures in causing injury. Temperatures below -40°C disrupt a common physiological adaptation, winter survival through freezing avoidance produced by free water supercooling in xylem tissue (33) and overwintering buds (35). In plant tissues, supercooling fails at or slightly above -40°C, the spontaneous nucleation point of supercooled water. The frequency of such extreme events has been correlated with the northern limits of the natural ranges of many trees in North America (12,35). Some genera, such as Prunus (33), include both species that rely on supercooling and others with poorly understood mechanisms that allow survival below -40°C.

Photoperiod reduction, decreasing autumn temperatures, and the physiological status of the plant can interact to acclimate woody plants to winter conditions (11). Record low temperatures during the normal acclimation period producing massive tree mortality have been well documented in trade literature (2,43). Such losses can be characterized by monitoring changes in low-temperature tolerance during the autumn and early winter in relation to on-site temperature data (26).

Whereas the timing and severity of low temperatures are limiting factors in the northern distribution of woody plants in North America, the limiting role of insufficient warmth during the growing season in Scandinavia has been well documented (40). The effects of insufficient warmth on plant distribution can be observed both on vegetative growth and development, and on sexual reproduction [see especially the works of Pigott and Huntley (31,32) on Tilia cordata]. These effects are not limited to Scandinavia, but also occur at other northern or montane sites. Ouellet and Sherk (28) reported that the mean frost-free period and the mean daily maximum temperature of the warmest month were statistically significant factors in multiple regression analyses of the relationship between climatic factors and woody plant adaptation across Canada. In contrast, my colleagues and I (50) found no significant effects of July mean temperatures on the survival of woody plants from Yugoslavia in the north-central United States.

Extremely high summer temperatures also induce stress, either directly to species that have evolved in alpine or otherwise cool conditions (9) or, indirectly, in concert with low relative humidities (23) or inadequate soil moisture, producing drought injury. A broad range of mathematical formulas has been proposed (7) for estimating the potential of temperature, moisture, and other factors, such as wind and insolation, to effect transpiration and thus to induce possible stress. These measures are generally known as potential evapotranspiration (PE).

Long-term moisture deficits and surpluses can be analyzed by comparing PE with mean annual precipitation. One method of expressing the balance between PE and precipitation, Mather and Yoshioka’s (24) moisture index, has helped predict the distribution of plant communities (24) and is a statistically significant variable, along with January mean temperatures, in predicting the survival of woody plants from Yugoslavia at test sites in the north-central United States (50).

Mean monthly or seasonal precipitation records have also been analyzed in conjunction with temperature data to predict woody plant survival in Canada (28) and the distribution of natural plant communities in North America (21,41). The relationship of woody plant adaptation to precipitation extremes producing serious droughts or floods has not been studied systematically, except for specific phenomena, such as relationships between leaf abscission and drought (1). Drought severity can be quantified with the Palmer Drought Severity Index or with Palmer’s Z-lndex (18), but no reports where these indices are compared with plant distribution have been found. However, Borchert’s (4) classic study of climatic factors corresponding to the distribution of grassland vegetation in the Great Plains and Midwest described interesting relationships between plant communities and patterns of July precipitation and temperature in drought years. His findings suggest that variability in the balance between PE and precipitation may be as important as is the overall balance. The effects of abnormally high precipitation are problematic, depending on the physiological status of the plant (winter floods vs. summer floods), and on drainage and other soil conditions that can influence root distribution (20).

Clearly, the photoperiod regimen, low temperatures, and interaction of high temperatures and moisture are all important determinants of woody plant adaptation. As we consider these factors in our search for tough trees, we would be wise to base our search on actual conditions faced by plants in our target environments (pre-identified managed landscapes), which may differ widely from conditions recorded at nearby weather stations (46), and on plant performance data previously collected under these or similar environments.

Edaphic Factors

Soils of managed landscapes, especially those in urban areas, have been greatly modified by human activity and often bear little resemblance to nearby undisturbed soils (8,47). Human disturbance creates soils that may be saline from deicing salts, alkaline from irrigation, calcareous from discarded construction materials, or compacted by construction machinery or other traffic. Finally, natural soil profiles are often disrupted when topsoil is removed before new construction. These disturbances produce soils with low oxygenholding capacities and chemical compositions unlike those commonly found in natural forest soils (3,8,42,45).

Although direct analogs to human-disturbed soils may be rare in nature, the processes of alkalization, calcification, salinization, and hardening all can occur during soil genesis (5) and are associated with particular soil types. When appropriate soil types are identified and detailed maps are available, it may be possible to employ this information to focus the search for tough trees, as has been proposed by Ware (45).

In the absence of detailed soil maps, knowledge of topographic and general landscape features [such as karsts (dissected limestone near the surface), extensive flood plains, or salt flats] can help focus the search. Certain plant groups are also indicators of soil drainage or chemistry: willows (Salix spp.) prefer poorly drained, seasonally flooded soils and deciduous azaleas (Rhododendron spp.) prefer nutrient-poor, acidic soils. Detailed distribution information for such indicator species may partly substitute for direct soil mapping data.

Case Study

With this very long “preface” complete, I will now describe an ongoing effort to locate and acquire potentially useful trees and shrubs native to Eastern Europe, for testing in the north-central United States. This project began in a rather unsystematic fashion in the mid-1970s, with the introduction to the United States of a diverse set of woody plant populations from the former nation of Yugoslavia (50). Many of these populations were evaluated at sites throughout the north-central United States as part of the NC-7 Regional Ornamental Trials, a long-term project to evaluate new landscape plants and to increase the future diversity of well-adapted plants found in commerce (48).

The 10-year evaluation results of Yugoslavian landscape plants indicated that only about a third of the populations survived and generally performed well throughout the north-central United States, another third failed at the colder or drier sites, and the remaining third failed at all sites (50). Statistically significant multiple-regression models, based on low winter temperatures and moisture conditions at test sites, explained 84% of variation for first-year survival and 56% of variation for overall survival across all sites (50). From these results and an analysis of climatic conditions in the former nation of Yugoslavia, my colleagues and I (50) developed a set of four criteria (Table 1) for locating Eastern European sites with environments more closely analogous to those found in the north-central United States.

These criteria are now being employed to focus future plant acquisition in Eastern Europe. First, it was necessary to analyze climatic and topographic data to determine which sites, if any, met the four criteria. I found that much of central and northern Ukraine, and adjacent portions of Belarus, the Russian Federation, and Moldova, along with two small areas in the foothills of the southern Carpathian Mountains in central Romania met all criteria (49). A literature review of natural plant communities was then performed to ensure that the criteria did not identify grassland sites lacking useful woody plants. Patterns of plant communities in these regions of Eastern Europe include mixed and deciduous woodlands, grasslands, and transitional communities similar to those found in Minnesota and Wisconsin (49).

It has not yet been possible to find soil maps for the Eastern European regions, and fine-scale vegetation maps are only now being prepared for publication (27). Thus, current efforts are focusing on developing contacts with botanical gardens in these regions known to collect and share seeds from natural populations (15) and identifying promising species. Nineteen botanical gradens have been identified for contact in the near future. And relevant floras (10,19,37,44) have been consulted to develop a comprehensive list of woody plants native to the region. This comprehensive list, along with habitat descriptions, aids our effort to identify promising species in three ways.

First, the list is being checked for those European landscape plants that are commonly cultivated in our region, with an emphasis on those that are poorly adapted to colder and drier areas, such as the northern Great Plains. Collections from Ukraine and surrounding areas, which meet the climatic criteria, should be better adapted to the harsher climates of the north-central United States than are those from sites farther south or west in Europe. Second, habitat information is being checked for species that are potential indicators for calcareous, alkaline, saline, and poorly drained soils. Finally, the list gives some indication of patterns of genetic diversity as reflected in species diversity with in genera. Genetically diverse genera, although perhaps unsuitable for direct introduction, may be particularly interesting subjects for future breeding and selection. A few examples identified by these analytical methods are listed in Table 2.

The true test of this program remains to be conducted. How will plants identified in this fashion perform at diverse sites in the north-central United States? Perhaps we will discover important limiting factors overlooked initially. Alternatively, we might be too successful. Our network of trial site cooperators will be warned that these introductions could be so well adapted that they might invade natural plant communities; they should manage these plants accordingly. We hope to begin acquiring and evaluating these plants during the next few years and ultimately find a broad array of tough new trees and shrubs.