Figure captions
Figure 1. Meteoric Waterlines (WMWL) generated with hydrogen and oxygen isotopic composition of all rainfall events originating from the Pacific or Atlantic Ocean during the study period between 2012 and 2014. The line for water stored and flowing within volcanic rock fractures is also shown (water-rock interaction). WMWL represents the global world meteoric water line of Craig (1961). LMWL represents the local meteoric water line of Sierra San Miguelito Volcanic Complex (SSMVC) in San Luis Potosi, Mexico. Inset shows the routes of the different meteorological events that influenced the research site during the 2012-2014 study period.
Figure 2. Combined electrical resistivity tomograms (ETR) and ground penetrating radargrams (GPR) taken in October 2012 after 10 days of rain (86 mm) in (a) Pinus cembroides (pine),(b) Quercus potosina (oak) and (c) mixed pine-oak forest stands. ERT profiles reveal a close linkage between the position of roots (diameter > 2.5 cm), soil resistivity (lower resistivity implies greater water availability), and rock fractures. The soil corresponds to the top 20 cm, the layer underneath the soil (regolith) includes soil pockets and rock fractures and is depicted by the dotted band, and below the solid line the fresh bedrock extends to great depth. Open circles of increasing size depict roots of increasing diameter. Black circles indicate roots used for calibration of the ground penetrating radar (GPR). Trees marked with X indicate the location of soil psychrometer sensors inserted at 12 cm soil depth.
Figure 3. Spatial analysis of root distribution for eight root diameter classes of Pinus cembroides and Quercus potosina , in pure and mixed stands, using the Kriging interpolation method at three different soil depths: 0-10 cm, 10-20 cm, and 20-30 cm. Transversal profiles are the graphical representation of the ERT (Figure 1). Symbols within the circular plots mark pine (left), oak (middle), and co-occurring oak and pine (right) tree positions. Prediction error (average standard error): (a) 0.23, (b) 0.28,(c) 0.32, (d) 0.27, (e) 0.19, (f) 0.25, (g)0.27, (h) 0.20, (i) 0.26.
Figure 4. Spatiotemporal niches identified by the value of soil and xylem water isotope composition (δ18O, δ2H) in oak and pine trees. The size of arrows denotes the proportion of water used by each species; a cross denotes uncoupling between soil and xylem water. Colors indicate isotopic signatures of different water sources; the warmer the color the more enriched the isotopic value of the substrate (i.e. yellow → orange → green); winter rain (red → violet); fracture water (blue). Average δ18O and δ2H are the isotopic values of water in different soil layers, xylem water, and the groundwater. Identified ecohydrological periods include the following: (a)wet at the end of summer 2012, (b) dry / depletion period early autumn 2012 to late winter 2013, (c) dry /recovery period throughout spring 2013, (d) wet/ from early summer 2013 to early autumn 2013, dry / (e) depletion period early autumn 2013 to late winter 2014, (f) dry / depletion period in late winter 2014, (g) dry / depletion period from late winter 2014 to late spring 2014, (h) wet / from late spring 2014 to early autumn 2014 and (i) dry / depletion period early autumn 2014 to early winter 2014. For better interpretation please refer to Table 1.
Figure 5. (a) Leaf (Ψleaf ) and(b) soil (Ψsoil ) water potentials associated with Pinus cembroides and Quercus potosina in pure and mixed stands in a semiarid forest ecosystem in San Luis Potosí, Mexico. Closed bars present total monthly precipitation recorded in the study site between September 2012 and December 2014. The shaded and clear areas indicate the length of ecohydrological periods defined in this study. Each point represents the mean ±1 SE (n=4 ).