Greenhouse experiment and phenotype measurements
Seedling traits that may be related to drought tolerance include height growth, root length, dry shoot weight, dry root weight, root-to-shoot dry mass ratio, and specific root length. Allocation to roots versus shoots can affect a seedling’s ability to take up water (Brunner & Godbold 2007; Markesteijn & Poorter 2009; Moran et al. 2017a), while the number and size of needles affect the area over which they may lose water through the stomata (Parker 1949). Two other traits related to this latter point are stomata density and the number of rows on the adaxial versus abaxial sides of the needles. However, it should be noted that lower stomatal conductance can be achieved with either fewer stomata, smaller stomata, or by closing the stomata more often (Irvineet al. 1998; Ryan 2011).
Seeds were collected from a subset of 50 genotyped parent trees in the summer of 2018. We placed 2-3 mature cones from each mother tree into paper bags and put them in a warm, dry place until seeds were released. We aimed to have ten seedlings from each of the fifty maternal families in wet and dry treatments, 1000 seedlings in total. Forty-eight of the 50 families had enough seeds in their cones to be included in the experiment (Fig. S1). During winter 2018, the seeds were stratified to break dormancy by placing them in aerated water for 48 hours, then surface-drying them and storing them in plastic bags in the refrigerator (~1.7°C) for six weeks. Because pines are wind-pollinated and outcrossing (Williams 2009), seeds from the same tree are mostly half-siblings, occasionally full-sibs.
Because the maximum first-year seedling root length observed in a pilot experiment was more than 110 cm, we used plastic tubes with an 8 cm width and 120 cm depth for planting. The bottom of each tube was capped with mesh to prevent the soil from falling out while allowing drainage. The lightweight clear tubes were wrapped in black plastic to keep roots in the dark. The planting soil was a mixture of 70% sand, 20% vermiculite, and 10% organic-rich potting mix to mimic the coarse texture of the soil of many Sierra Nevada conifer forests (Baleset al. 2011). To keep tubes upright, we used PVC pipes to build 10 frames that could each hold 100 tubes. Two seeds from each family were planted in each tube in February 2019, and two tubes from each family were randomly placed within each frame. In April 2019, we replanted more stratified seeds of the correct family in tubes without seedlings. All the tubes were watered every other day during the germination and seedling establishment period (February through June).
At the end of June 2019, all but one seedling per tube was removed, and alternating frames were assigned to the wet treatment and the dry treatment (5 frames containing up to 500 seedlings per treatment) (Fig. S1). The wet treatment group was watered twice weekly, and the drought treatment group was watered once every three weeks until mid-October (3.5 months). While wild ponderosa pine seedlings would receive little to no precipitation during the summer, this occasional watering was necessary for the greenhouse environment. Temperatures inside the greenhouse in the low-elevation environment of Merced, CA, reached as high as 37°C on the hottest days and the soil volume of the tubes was limited, with no access to groundwater, both of which made evaporation and drought stress more intense than the no-precipitation condition in the wild.
Multiple phenotypic traits were measured during and after the greenhouse experiment. Only 42 out of 48 mother trees had enough germination to carry out these measurements across both treatments. We calculated shoot growth as final height minus height at the initiation of the treatments. The length of fresh roots was measured from the soil surface to the taproot tip immediately after the harvesting to avoid shrinkage. Following harvest, needles, fresh stems, and fresh roots of all the seedlings were separately put into paper bags and dried at 75 °C for 48 hours. We measured root dry mass (RW) and shoot weight (SW, total of stem and needles). We then calculated the root-shoot ratio (R2S) as RW/SW. Specific Root length (SRL) was calculated as root length/root weight.
Before harvest, we also collected 3-4 fresh needles from living seedlings to calculate stomatal density. In pines, stomata are arranged into longitudinal rows. We put each needle on a slide and photographed it at 100x magnification using a Leica DME compound microscope equipped with a Leica DFC290 digital camera. All counts were conducted near the middle of the needle to avoid variation that might occur at the base and the tip. Approximately 1.96 mm lengths of the needle were surveyed for the number of stomata and stomatal rows on their adaxial (upper) and abaxial (lower) surfaces. Needle width was measured in magnified images using the line measure tool in the Leica software. Then we calculated the stomata density on each side as the number of stomata divided by 1.96*width of needle. An average density and number of rows were calculated for each individual across sampled needles.