There have been few studies of water saving technologies in urban environments, especially involving tree pits. A technology to potentially reduce the effect of drought stress, particularly on seedlings and saplings, could be the hydrophilic, cross-linked polymers known as super absorbent polymers or hydrogels. The term hydrogel is sometimes used for these compounds because when the dry crystals absorb water, they take on the consistency of a gel [13] . Hydrogels can retain up to 400 times their weight in water when saturated, but will not dissolve in water [14] . The hydrophilic functional groups attached to the polymeric backbone enable hydrogels to absorb water, while the cross-links between network chains make hydrogels resistant to dissolution [15] . Ideally, at least 95% of the water stored in hydrogels is available to plants [16] .

Hydrogels can also retain nutrients when incorporated into the soil, making them available for plant growth whenever required. Hydrogels are hydrophilic and contain carboxylic groups, enabling them to bind cations and adsorb water [17] . The water and nutrients stored in hydrogels are released gradually for plant growth under water limiting conditions, whereas under non-water limiting conditions, they are reported to enhance nutrient uptake for plant growth [18] . This is of special interest for urban foresters, because water stress and nutrient deficiency are two important growth-limiting factors for street trees.

Different factors affect the absorptive capacity of hydrogels for water. These factors include the tolerance of hydrogels to ionic solutions, the tensions at which they bind water, and the speed with which they degrade in the field [19] . According to Orzolek [20] , under field conditions, polymers will lose 10% to 15% of their activity each year. The degradation of polymers can be due to microbial activity [21] , chemical decomposition [20] , or physical modification over time. In spite of the beneficial aspects of hydrogels in the soil and their effects on plant growth, an overabundance of hydrogels can have negative consequences in some cases [13] . For instance, Sarvaš et al. [22] observed high mortality (64%) of pine seedlings caused by an overdose of hydrogel. Shooshtarian et al. [23] attributed a high seedling mortality rate in hydrogel-amended soil to a reduction of free air space resulting from the swelling of the hydrogel and the consequent reduction in soil aeration.

In general, the main function of hydrogel in soil is to preempt the negative effects of water deficiency on plant growth and productivity. However, its effectiveness varies depending on the situation. Accordingly, the objective of this research was to observe the effects of differential proportions of hydrogel, under conditions of different water availability, on the growth of Siberian elm and silver maple in greenhouse experiments.

The project was performed in a greenhouse located on the Macdonald Campus of McGill University, Sainte-Anne-de-Bellevue, Canada, under humid summer conditions of 28˚C - 30˚C, 65% relative humidity, and a 16-h photoperiod. The experiment was run in two phases of six weeks each: the first from Aug. 18^th to Sep. 30^th, 2014, and the second from Oct. 2^nd to Nov. 11^th, 2014. Siberian elm (Ulmus pumila) and silver maple (Acer saccharinum) saplings were potted in sieved soil (tamisé) (Laniel Prodamex, Pierrefonds, QC). The soil in each pot was thoroughly mixed with one of two proportions (0.5 and 1.0% dry weight) of dry Super ABA 200 (Iramont Inc., Laval, QC), a potassium polyacrylamide hydrogel, as compared to the control soil (0% of hydrogel). Water-soluble fertilizer (20-20-20, Passion Jardins, Vaudreuil-Dorion, QC) was added to the soil at the beginning of the experiment.

In the first phase of the experiment, the trees were arranged in a full-factorial, completely randomized design with twelve replicates, totaling 72 pots. The trees were irrigated regularly and the pots were randomly rearranged at each watering. At the end of this phase, 18 trees, i.e. three replicates per treatment, were sacrificed and analyzed for growth and physiological parameters.

The second phase used the remaining trees (Siberian elm and silver maple, potted in soil with one of three proportions of hydrogel), but also included three different irrigation schedules, replicated three times, totaling 54 trees. The trees were either watered daily (no stress), weekly (moderate stress), or every second week (severe stress). In order to keep the water availability similar within treatment levels, the same amount of water was added to each pot in that level at the time of watering.

The measured growth parameters were total height from the soil surface to the top of the crown and trunk diameter at the soil surface, measured every second week. Normalized difference vegetation index (NDVI) was measured using a Crop Circle™ ACS-430 active crop canopy sensor (Holland Scientific, Lincoln, NE). Leaves were tested using a SPAD 502 Plus Chlorophyll Meter (Spectrum Technologies, Aurora, IL). The SPAD-502 meter measures leaf transmittance in the red (650 nm) and infrared (940 nm) bands. The SPAD meter value, typically between 0.0 and 50.0, is proportional to the amount of chlorophyll in the leaf [24] . Both NDVI and SPAD readings were measured and recorded every second week.

A Kolmogorov-Smirnov test showed that none of the data were normal, so a two-step transformation was used to rank the values by percentile, giving uniformly-distributed probabilities, and then map them onto normally-distributed z-scores using an inverse-normal function [25] . A general linear model was performed to diagnose the significant interactions between independent variables (hydrogel concentration, watering frequencies and the species). Analysis of Variance (ANOVA) was then applied to the transformed data to test for significant effects of tree species, hydrogel concentration, and watering frequency on the transformed values of growth, NDVI, and SPAD readings (α = 0.05) (SPSS Statistics v.21, IBM Corporation, Armonk, NY). Mean values of the measured variables for each hydrogel concentration were compared among watering frequencies using the Tukey method (α = 0.05) [26] . A bivariate correlation test was also performed between the dependent variables (stem height, stem diameter, NDVI, and SPAD readings) and the independent variables (watering regime and hydrogel concentration).

Hydrogel had no observable effect on well-watered trees, as measured during the first phase of the experiment. This is consistent with previous studies [27] . The experimental factors did have significant effects on the water-stressed trees, as observed during the second phase of the experiment. Significant interactions were observed between tree species and hydrogel concentration, and between tree species and watering frequency (Table 1). This result indicates that the elms and maples each responded differently to the hydrogel treatment and watering frequency. Therefore, the rest of the statistical analysis was performed separately on the two tree species.

According to Rais et al. [28] , drought-induced water stress can influence tree growth even before it has an evident effect on physiological processes such as photosynthesis. We observed that change in stem height was a more sensitive indicator of drought stress than stem diameter in both species (Table 2), as have other authors [28] . Bouriaud et al. [29] surmised that tree height is sensitive to water stress because of restricted movement of water to the actively growing extremities. Moreover, water stressed trees might favor root extension over crown expansion in an attempt to compensate for the water deficit [28] [30] .

We observed a negative effect of watering frequency on the height of silver maple (r = −0.29, p < 0.05). In corroboration with this result, Elhadi et al. [31] also found tree height increased more rapidly when saplings were watered every nine days than when they were watered every three or six days. They concluded that frequent irrigation is not always profitable for plant growth because it can negatively affect root respiration and growth. As a result, water and nutrient uptake by roots diminishes and tree growth declines [32] .

Under water stress, 0.5% hydrogel concentration increased stem height in both tree species (r = 0.48, p < 0.05, Figure 1, Table 2 and Table 3). We also observed a significant interaction between hydrogel and watering frequency on