Western North American Naturalist 62(3), © 2002, pp. 266–279 TREE GROWTH AND REGENERATION RESPONSE TO CLIMATE AND STREAM FLOW IN A SPECIES-RICH SOUTHWESTERN RIPARIAN FOREST Donna M. Galuszka1 and Thomas E. Kolb1,2 ABSTRACT.—We studied the influence of climate variables, stream flow, and topography on regeneration and growth of several riparian tree species (Acer negundo, Alnus oblongifolia, Fraxinus velutina, Juglans major, Platanus wrightii, Populus fremontii, Salix spp.) at an unregulated perennial stream, West Clear Creek, in central Arizona. A pulse of seedling regeneration occurred for Alnus, Fraxinus, Platanus, Populus, and Salix in 1995 and 1996 following high winter and spring surface flows in 1993 and high spring surface flow in 1995. In contrast, little regeneration occurred for Acer and Juglans under these conditions. Most seedlings occurred at the active channel topographic location, and few seedlings occurred at abandoned channel, gravel-boulder bar, and bench locations. Relationships between environmen- tal variables and annual radial stem growth varied among species and between constrained and unconstrained reaches. High spring or winter surface flows were negatively related to growth of Acer, Alnus, and Platanus, whereas high spring surface flow was positively related to growth of Fraxinus. Positive relationships between precipitation and growth occurred only for Fraxinus and Juglans, suggesting greater use of surface soil water by these species. Annual radial growth was high for Platanus and Alnus, medium for Acer and Fraxinus, and low for Juglans. Overall, the tree species in our study responded individually, rather than collectively, in regeneration and growth to changes in stream flow and cli- matic variables. Key words: Arizona, Acer negundo, Alnus oblongifolia, climate, dendrochronology, Fraxinus velutina, Juglans major, Platanus wrightii, Populus fremontii, regeneration, riparian, Salix, tree ring, stream flow. Riparian areas are physically linked to and factors that influence the regeneration and dependent upon perennial, ephemeral, or growth of riparian trees need to be better intermittent surface or subsurface waters (Ari- understood. zona State Parks 1994). These areas provide Flooding and resulting geomorphic changes critical habitat for wildlife and vegetation in riparian ecosystems often cause changes in across the United States (Brinson et al. 1981) their biotic communities (Vannote et al. 1980). and are important for maintaining regional bio- For example, recruitment of riparian Populus diversity (Naiman et al. 1993). Riparian areas species has been linked to high surface flows in the southwestern United States support in western North America (Everitt 1968, Brady forests that resemble the mesophytic humid et al. 1985, Howe and Knopf 1991). High forests that were widespread in the Early Ter- flows scour openings along stream banks, tiary era (Minckley and Brown 1982). The widen floodplains, and deposit alluvium, thus development of a drier climate after this era in creating extensive open fluvial surfaces for the Southwest limited riparian forests to sites recruitment of pioneer trees such as Populus with dependable water sources (Minckley and (Stromberg et al. 1991, 1993, Scott et al. 1997, Brown 1982). More recently, southwestern Stromberg 1997). Moreover, seedfall for many riparian forests have been depleted by human riparian trees coincides with spring flooding impacts such as flow regulation, dewatering, (Everitt 1968, McBride and Strahan 1984, Fen- groundwater pumping, and grazing (Johnson ner et al. 1985, Bock and Bock 1989, Siegel and and Haight 1984, Swift 1984, National Research Bock 1990, Stromberg 1993, Friedman et al. Council 1992, Rood and Mahoney 1993, Strom- 1995). Further, overland flows created by spring berg 1993, Busch and Smith 1995, Stromberg flooding can promote seedling establishment et al. 1996). Given the high value and increas- by providing a moist seedbed for germination ing scarcity of southwestern riparian forests, (Bock and Bock 1989, Stromberg 1997). 1School of Forestry, Northern Arizona University, Box 15018, Flagstaff, AZ 86011-5018. 2Corresponding author. 266 2002] TREE GROWTH AND REGENERATION RESPONSE 267 Survival of riparian tree seedlings can be the greatest densities occurring in abandoned influenced by environmental factors such as channels, low benches, and adjacent to the temperature, pH, and salinity (Siegel and active channel. Third, seedling density and Bock 1990), as well as their location on the annual radial growth of all species would be floodplain. Optimal germination sites for long- greater in unconstrained areas of the flood- term survival of riparian trees are at elevations plain compared with bedrock-constrained areas. high enough on the floodplain to protect the Fourth, annual radial growth of all tree species substrate and new seedlings from scouring, would be correlated with stream flow and cli- yet low enough to provide direct contact with mate characteristics. Specifically, we expected moist sediments (McBride and Strahan 1984, positive correlations between growth rate and Stromberg et al. 1991, 1993, Scott et al. 1997, surface flow and precipitation based on the Stromberg 1997). Riparian forests also have premise that tree growth in the southwestern been shown to develop successfully in aban- United States is often strongly limited by doned channels created by fluctuating seasonal water availability (Fritts 1974). We also expected flows (Brady et al. 1985). Further, the topogra- negative correlations between growth and air phy of the floodplain may also affect long-term temperature during the growing season be- survival of riparian trees. Trees growing in cause high temperatures would limit photo- broad floodplains may suffer lower mortality synthesis due to high vapor pressure deficit, as during high flows than those growing in narrow, shown for Acer negundo at a perennial stream constrained reaches due to high flow velocity near our study site (Kolb et al. 1997). Last, we and inundation in narrow reaches (Palik et al. hypothesized that radial growth rate would 1998, Friedman and Auble 1999). differ among species. Most research on southwestern riparian forests has concentrated on low-elevation METHODS forests dominated by trees in the genera Popu- Study Site lus and Salix and has focused on the impacts of stream flow or groundwater regulation by The study location is West Clear Creek, a humans (e.g., Brady et al. 1985, Stromberg and free-flowing perennial stream within the sub- Patten 1990, 1991, 1996, Howe and Knopf 1991, Mogollon region (Minckley and Brown 1982) Stromberg et al. 1991, 1993, 1996, Stromberg of the Colorado Plateau in central Arizona 1993, 1997, Busch and Smith 1995, Cooper et (Fig. 1). West Clear Creek flows approximately al. 1999, Shafroth et al. 2000, Horton and 48 km through a deep canyon to its confluence Clark 2000, Horton et al. 2001). Factors that with the Verde River. The geology of the water- affect regeneration and growth of species-rich shed comprises sedimentary rocks overlain with riparian forests of higher elevations in the basalt in some areas. Elevations in the West Southwest (Brown 1982, Szaro 1990) are poorly Clear Creek watershed range from approxi- understood, and yet these forests are wide- mately 2300 m to 945 m. Our study site is spread in the region and may be increasingly located at the Bull Pen U.S. Geological Survey threatened by anthropogenic impacts as human gaging station (lat. 34°32′19″, long. 111°41′36″, populations grow in the Southwest. Our study in NW1/4, NW1/4, sec 11, T13N, R6E, Yava- addressed the influences of topographic loca- pai County, Hydrologic Unit 15060203) 14.5 tion, stream flow, and climatic variables on km east of Camp Verde and 17.7 km upstream regeneration and growth of several riparian of the mouth in the Coconino National Forest tree species (Acer negundo, Alnus oblongifolia, at 1106 elevation (Fig. 1). Fraxinus velutina, Juglans major, Platanus Surface flow at West Clear Creek varies wrightii, Populus fremontii, and Salix spp.) seasonally. Between 1969 and 1992 average growing in a relatively pristine, species-rich monthly base flow varied between 0.42 and riparian forest along an unregulated perennial 0.57 m3 sec–1 from May through January stream in central Arizona, West Clear Creek. (Galusky 1994). Surface flows were highest in We addressed the following hypotheses. February and March due to runoff from snow- First, high stream flow promotes seedling melt and averaged 1.13 m3 sec–1 (Galusky 1994). regeneration of all these tree species. Second, Temperatures at highest elevations in the water- seedling density would be distributed unevenly shed (Happy Jack Ranger Station, Coconino across different topographic locations, with National Forest) average 7.8°C annually, 268 WESTERN NORTH AMERICAN NATURALIST [Volume 62 Fig. 1. Location of West Clear Creek and the study area in the Verde River Watershed, Arizona. 17.2°C in July, and –2.8°C in January (West- A variety of biotic communities occur in ern Regional Climate Center). Average annual the West Clear Creek watershed. In upland precipitation at the highest elevations is 68.6 parts of the watershed (elevations >1823 m), cm, which includes an average of 223.5 cm of Pinus ponderosa dominates (Plan West Associ- snowfall. Temperature at Beaver Creek ates 1994). Between elevations of 1822 m and Ranger Station (Coconino National Forest), approximately 1372 m, Pinus edulis, Juniperus which is located within approximately