Variable Urban Irradiance and Shade Acclimation in Norway Maple Street Trees

Materials and Methods

Thirteen sites in Seattle, Washington where Acerplatanoides ‘Emerald Queen’ was planted in 1977 were selected for investigation. Most members of the Aceraceae in North America are considered to be moderately to highly shade tolerant (11), and A. platanoides, a widely naturalized species tolerant of many conditions, is unlikely to be an exception. In addition Emerald Queen maple is widely planted in Seattle, thus offering sites with a range of solar exposures. The selected sites ranged from nearly unobstructed horizons in commercial-residential districts to those with buildings 30-150 m in height in the urban core. Each site was a city block with trees (n=10-12) growing in a concrete parking strip contiguous with the sidewalk. Each tree was located in a 1.3 x 1.3 m cutout in the pavement with exposed soil surrounding the tree base. Trees in the urban core were selected to represent as wide a range of irradiance levels as possible, while those outside the core were randomly selected. Trees next to either bus stops, light posts, corners, or small structures that could create confounding effects were not considered for selection. The only maintenance on these trees was periodic pruning to maintain pedestrian clearance.

Global shortwave radiation (irradiance) received at street level at each site was assessed after the method previously validated by Kjelgren and Clark (10). In September 1992 the horizon topography of each site was determined at 2 m height above street level by measuring paired azimuth and elevation angles of surrounding buildings in 10° increments with a clinometer (Suunto Corp., Finland) and compass. These data were then used to characterize the light environment of each tree by estimating potential daily irradiance (MJ/m²/day) from April 1 to September 30 using the algorithm of Flint and Childs (4). Average daily irradiance was then summed to give potential seasonal irradiance. This was repeated for a full-sun, completely unobstructed horizon to define maximum potential irradiance. Seasonal totals divided by the maximum potential gave percentage irradiance by site.

Tree responses to variable irradiance levels were measured atthe same time as determination of horizon topography. Specific leaf area was determined from 10 leaves taken from three first-order shoots per tree for 30 total leaves. A 2 cm² disk was removed from each leaf, and all disks from each shoot were dried at 60°C for one day and then weighed. Specific leaf area per shoot was calculated as the ratio of leaf-disk area to weight. Trunk diameter growth was determined from increment cores taken at 2 m height from each tree as the average yearly diameter increment for the previous five years. Crown density (m≡/m³) was determined with a leaf area index meter (Model 2000, LI-COR Inc. Lincoln NE). Four paired unobstructed and below-canopy (at the base of the tree crown) readings of light transmittance were collected in four directions around the trunk that was repeated three times for each tree. The data were analyzed describing the relationship between growth responses and potential seasonal irradiance by fitting each response to irradiance level by using commercial software (TableCurve Windows 1.0, Jandel Scientific, San Rafael CA). The curve that yielded the highest coefficient of determination with the lowest number of parameters (high F statistic) established the most efficient best-fit equation describing each relationship.

Results and Discussion

Percentage of potential irradiance varied widely among sites as a function of differences in horizon topography. Figure 1 shows a hemispherical photo of a typical urban horizon topography on a northwest-southeast trending street in Seattle. This was overlaid with tracks marking the path of the sun at different times of the year for 50° north latitude. While this was several degrees north of Seattle’s actual latitude, the two locations were close enough to illustrate the effect of horizon topography on the number of hours of direct sunlight. Tic marks and numbers along the edge of the circle define the azimuth angles along the horizon, and the concentric circles indicate height above the horizon. Building height obscured the sun until noon, whereupon a window of direct sunlight lasted until late afternoon. This gave a total of approximately five hours of direct sunlight at midsummer as compared to a potential of 14 hours at an unobstructed site. During mid-winter penetration of direct sun would be limited to a small road gap in the horizon topography that would result in less than an hour of sunlight.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Hemispherical photograph of an urban canyon in Seattle, Washington, overlaid with a polar grid delineating the path of the sun for different seasons. The top of the figure is oriented towards true north.

The pattern of exposure to direct sunlight varied within very short distances in the business core due to street orientation and proximity to buildings (Fig. 2). Sites 1 and 3 in the core were on either side of an east-west street across from one another (approximately 16 m apart), but site 1 received much more sunlight than site 3. Site 1 had a southern aspect, hence far enough away from the buildings to the south that building height obscured direct sunlight only for a brief late-morning period. Site 3, by contrast, had a northern aspect immediately adjacent to a large building with minimal setback from the sidewalk that obscured direct sunlight except for early morning and late afternoon illumination through street gaps. Site 4 was similar in orientation to that in Figure 1, such that the north-south trending street located in a rather narrow urban canyon resulted in a midday gap of direct sunlight. Irradiance was lower at sites 1,3, and 4 than for the unobstructed horizon during direct sunlight. This was because buildings obscured large portions of the sky away from the path of the sun that reduced receipt of diffuse radiation.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.

Potential global shortwave radiation estimated for June 22 (summer solstice) for three sites in Seattle, Washington, and for an unobstructed horizon.

Estimated potential seasonal irradiance was lower in the business core than outside the core due to a more developed horizon topography (Table 1). Potential irradiance varied for sites 1-9, all within a three-block area in the urban core, between 27-67%. The lack of sites between the range of 70-90% irradiance was likely due to their infrequency in the urban core and the limited number of maples at such sites. The lowest irradiance occurred at sites with northern aspects next to tall buildings that limited irradiance to morning and evening penetration through street gaps at low solar elevation angles. Sites 10 and 11 were located next to low buildings that obscured direct sunlight only through mid-morning, reducing seasonal irradiance by 10%. Sites 12 and 13 were either next to very low buildings or had a full southern exposure that only reduced seasonal irradiance by less than 5%.

Decreasing irradiance significantly affected growth (Fig. 3). The responses of specific leaf area, trunkgrowth, and crown density to decreasing seasonal irradiance were best described with dosage-type curves. The dosage-type curve yielded the highest correlation and F statistic, indicating that that particular equation type described the data most efficiently using the fewest number of parameters. There was one anomalous low irradiance tree having crown density equal to the full sun trees (Fig. 3c). Apart from that outlier, all three measures of growth shifted to different levels between 70-85% of potential irradiance that indicated shade acclimation. Specific leaf area in particular is considered to be a sensitive indicator of foliar shade acclimation (2), and these results are similar to those reported for sweetgum (10). Specific leaf area could have been described with a linear response (10) but at a lower r². Such a fit, however, would have been inconsistent with the dosage responses of trunk growth and crown density.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3.

Growth responses, specific leaf area, trunk growth, and crown density of thirteen Emerald Queen Norway maple street trees to different potential seasonal irradiance levels resulting from variable horizon topographies. All three responses were fitted to a dose-response function that had the form y=a+b(1+x/c^d), where parameters a, b, c, and d, as well as the coefficient of determination, are listed on each graph. The hollow outlier data point in (c) was not included in the calculations.

While their responses differed in form, both these maples and the sweetgums (10) were shade acclimated at about 60% irradiance. Since sweetgum is considered to be shade intolerant (5), acclimation responses by other broadleaf species of intermediate to low shade tolerance (8) to variable irradiance may occur at a similar level. The effect of shade acclimation on leaf thickness, trunk growth, and crown density at low irradiance allows broadleaf trees in urban canyon settings to balance carbohydrate supply and demand. Photosynthesis is maximized under shaded conditions by allocating a limited supply of carbohydrates to fewer and thinner leaves to increase illumination and decrease self-shading (2). This results in g reater specific leaf area and lower crown density, and fewer carbohydrates availablefortrunkgrowth (1).

Whether or not reduced trunk growth and lower crown density should be a management concern depends on expectations. While lower crown density would not be considered a desirable trait in shade trees, the trees in this study were nonetheless healthy, with intact foliage and no branch die-back. Shade-stress responses in these trees would not be detectable without full-sun trees present for comparison. Consequently, stress responses such as lower crown density and less trunk growth need not be considered limiting in species selection of broadleaf shade trees for urban canyons since appearance would probably not be perceptibly affected. Shade tolerance may be a factor to consider if interactions with stresses such as compacted soil are possible (9).