Abiotic factors vary along altitudinal gradients, and this may influence plant morphology, physiology and function. This study aimed to test the hypothesis that leaf δ¹³C,  a common proxy for water use efficiency, was indirectly influenced by morphological adjustments with changing climatic factors along an altitudinal gradient on Mt. Kenya. We sampled leaves of Dendrosenecio keniensis and Lobelia gregoriana using seventy-two 10 × 10 m plots situated every 100 m starting from 3600 to 4300 m. We determined leaf δ¹³C using stable isotope mass spectrometry. We also quantified the following morphological factors; leaf area, leaf mass per area, specific leaf area and leaf thickness. Climate data included mean annual temperature and precipitation, diurnal temperature range and water vapor pressure. Our results revealed that there was a leaf δ¹³C enrichment of 1.76 ‰km^-1 and 1.62‰ km^-1 with altitude for D. keniensis and L. gregoriana respectively. Leaf δ¹³C was enrichment by 0.01‰ mm^-1 with mean annual precipitation along the altitude gradient for D. keniensis and 0.02‰ mm^-1 for L. gregoriana. D. keniensis and L. gregoriana have high water use efficiency, an adaptation for surviving near freezing alpine temperatures and high diurnal range. Leaf δ¹³C exhibited a depletion of -0.37‰ per ^o C increase of mean annual temperature along the altitude gradient for D. keniensis and -0.34‰ per ^o C increase for L. gregoriana. Our results also showed a negative relationship between pCO₂ and leaf δ¹³C and positive relationship between pCO₂ and ∆¹³C for both species. Low temperatures led to the increase in leaf thickness and specific leaf area for these two species, factors that influenced leaf δ¹³C and ∆¹³C.

Data collection of Dendrosenecio keniensis and Lobelia gregoriana was done using a total of seventy-two 10 × 10 m plots were set up along the altitudinal transect with 9 plots being set up in each altitudinal belt starting from 3600 to 4300 m at 100 m altitude intervals. The leaf thickness of at least 5 fresh fully expanded leaves from different D. keniensis and L. gregoriana in each 10 × 10 m plot were also measured using a digital vanier caliper before the leaves were plucked from the stems. The fresh leaves of at least 5 mature fully expanded leaves from plants occurring in each plot were picked and put in a collection bag for later scanning using a digital A4 scanner EPSON perfection V33/V330. The images were used to calculate the respective leaf area in RStudio using package LeafArea (Katabuchi, 2015) and ImageJ (Schneider et al., 2012). The leaves were later oven dried at 70o C to a constant weight and the final weight was recorded as the leaf dry weight. Specific leaf area was calculated by dividing each leaf area with its corresponding oven-dry weight.