«Summer Establishment of Sonoran Desert Species for Revegetation of Abandoned Farmland Using Line Source Sprinkler Irrigation BRUCE A. ROUNDY ...»
Arid Land Research and Management, 15 :23È39, 2001
Copyright # 2001 Taylor & Francis
0890-3069 /01 $12.00 1.00
Summer Establishment of Sonoran Desert Species for
Revegetation of Abandoned Farmland Using Line
Source Sprinkler Irrigation
BRUCE A. ROUNDY
Department of Botany and Range Science
Brigham Young University
Provo, Utah, USA
University of Tehran
STEVEN E. SMITHSchool of Renewable Natural Resources University of Arizona Tucson, Arizona, USA
MARK PATERUnited States Department of Agriculture Natural Resources Conservation Service Plant Materials Center Tucson, Arizona, USA A line source sprinkler was used to determine water requirements of adapted species for revegetation of abandone d farmlands in the Sonoran Desert of southern Arizona.
Six grass and seven woody species were seeded on a Ðne-sandy soil initially at Ðeld capacity water content during the summers of 1992 and 1993 in T ucson, Arizona.
T he line source sprinkler created a greater di†erence in soil water availability between irrigated and unirrigated soils than within the irrigation gradient itself.
Initial irrigation was followed by periods of summer rainfall in both years. Species emergence and establishment varied with amount and timing of irrigation and summer rainfall. After the initial 14 days of irrigation in 1992 and 11 days in 1993, soil water availability was intermittent in the surface soil but was consistently high at depths greater than 18 cm. Emergence was more sensitive to the irrigation gra- dient than was plant survival. Most species established successfully with at least 210 mm of irrigation plus precipitation. A possible strategy for establishing many of these species on abandoned farmland would be to Ðll the soil proÐle to Ðeld capacity by irrigating on consecutive days until emergence is observed. Direct seeding without irrigation or water concentration is not recommended due to the erratic and Received 21 April 2000 ; accepted 5 July 2000.
Address correspondence to Dr. Bruce A. Roundy, Department of Botany and Range Science, 401 WIDB, Brigham Young University, Provo, UT 84602, USA. E-mail : bruce–firstname.lastname@example.org 23 B. A. Roundy et al.
24 limited nature of summer rainfall on abandoned farmland in the lower Sonoran Desert.
Keywords revegetation, seedling emergence, survival, water requirements, drought tolerance, soil water availability Since the early 1950s, over 400,000 ha of once-irrigated farmland have been abandoned in Arizona (Charney and Woodward 1990). Abandonment is the result of variability in costs of production and also in demand and returns for agricultural products over the years (Shapiro 1989). When ArizonaÏs 1980 Groundwater Management Act restricted pumping in certain Active Management Areas, cities began purchasing farmlands for water sources (Gelt 1993). Future purchase of such socalled ““water farmsÏÏ has now been prohibited by the 1990 Groundwater Transportation Act, but large tracts of retired croplands used as water farms still lie abandoned. Problems with abandoned farmlands include wind erosion and hazardous dust storms that have resulted in highway fatalities, expense of controlling weeds such as Russian thistle, Salsola iberica Sennen & Pau, and loss of ecological and aesthetic value (Meitl, Hathaway, and Gregg 1983).
Much of the abandoned farmland in Arizona is in the south-central part of the state in the Santa Cruz Valley and along the Gila River between Phoenix and Yuma. These lands are characterized by Ðne-textured soils and annual rainfall often less than 250 mm (Jackson, McAuli†e, and Roundy 1991 ; Gelt 1993). Reinvasion of these lands from native shrubs such as creosotebush, L arrea tridentata, and desert saltbush, Atriplex polycarpa (Torrey) S. Wats., is limited by lack of dispersal vectors as well as aridity, salinity, poor soil physical conditions, and disruption of former natural drainage patterns (Allen and Jackson 1992 ; Jackson 1992). Some abandoned farmlands with comparatively coarser-textured soils and higher annual rainfall may be colonized by desert broom (Baccharis spp.) and burroweed, Isocoma tenuisecta Greene, while others remain bare or dominated by Russian thistle over 25 years after abandonment (Karpiscak 1980). These lands are classiÐed by the U.S. Department of Agriculture, Soil Conservation Service (1981) as lower and upper Sonoran Desert Shrub, Phoenix Desert Shrub, and Chihuahuan Semidesert Grassland.
Revegetation of abandoned farmland without irrigation generally requires some form of water harvesting and concentration because sufficient rainfall to naturally establish desert perennials is infrequent (Bainbridge and Virginia 1990 ; Cox et al.
1982 ; Shanan et al 1970). The Natural Resource Conservation Service (Munda
1993) and Jackson and others (1991) have successfully established native shrubs between runo† areas and catchment berms near Redrock and Eloy, Arizona, respectively. Ripping to increase inÐltration has also increased plant establishment on abandoned farmlands with compacted, Ðne-textured soils (Munda 1993).
Use of functioning irrigation systems to establish adapted plants during the Ðrst year after abandonment is a promising approach to revegetating abandoned farmlands in arid areas with erratic precipitation (Cox and Thacker 1992). However, establishment and persistence of irrigation-establishe d plants is highly dependent on site conditions and drought tolerance of the seeded species. For example, Cox and Madrigal (1988) failed to establish forage grasses permanently on a silty clay loam soil at the San Xavier Indian Reservation, but Cox and Thacker (1992) succeeded in establishing grasses on sandy loam and clay loam soils in the Avra Valley in southeastern Arizona. In the latter study, plant establishment required sufficient irrigation to keep the soil surface moist until seedling roots were 5 to 15 cm long (four to six weeks of irrigation). To successfully establish, many warm-season grasses may require rather extended periods of available soil moisture at the soil surface to develop adventitious roots (Roundy et al. 1993, 1997). Establishment of woody species in the desert is also dependent on rapid root growth to stay below the soil drying front (Bainbridge and Virginia 1990 ; Jackson, McAuli†e, and Roundy 1991).
Sonoran Desert Revegetation 25 Transplants with deep roots grown in tall pots have survived well in the SonoranMojave transition zone of southern California (Holden 1992). However, direct seeding is usually necessary to revegetate large-scale disturbances (Roundy 1999).
If existing irrigation systems can be used to revegetate farmlands the Ðrst year after abandonment, we need to identify adapted plant materials and their associated water requirements for establishment. The line source sprinkler system (LSS) produces a gradient in applied water that has been used mainly to determine crop production response to irrigation (Miller and Hang 1980). LSS has the advantage of providing a continuous gradient in irrigation from excess to no irrigation within a small area as distance increases from the line source (Hanks et al. 1976 ; Fernandez 1991). The objective of our research was to use LSS to determine potential for establishment of direct-seeded grasses, shrubs, and trees in the Sonoran Desert.
Materials and Methods The study site was at the Tucson Plant Materials Center of the Natural Resources Conservation Service. Elevation is 773 m and mean annual precipitation is 293 mm with 54% (159 mm) falling between July and October (Sellers, Hill, and SandersonRae 1985). Soils are of the Anthony series, loamy Ðne-sand, mixed, calcareous, thermic family of Typic TorriÑuvents.
We seeded six grasses and seven shrubs or trees considered adapted to Phoenix Desert Shrub climatic conditions (Jordan 1981) in 1992 and 1993 in July prior to the summer rainy season (Table 1). Laboratory germination tests were conducted to determine the pure live seed percentage of each seed lot and seeds of woody species that did not imbibe water were scariÐed by hand with sand paper and retested.
Grasses, shrubs, and trees were seeded at rates of 66, 33, and 16 pure live seeds per m of row, respectively, in rows 0.41 m apart using a no-till planter. Rows were seeded perpendicular and to a distance of 16.5 m on both sides of the line source sprinkler. The experimental design was a randomized complete block with treatments arranged as a split plot. Blocks were replicated six times, with three blocks on each side of the LSS. Species were whole plots and irrigation levels were subplots.
The soil was preirrigated in both years prior to sowing to Ðll the soil proÐle to Ðeld capacity to a depth of 1 m and to create similar antecedent soil moisture conditions throughout the Ðeld. An extra sacriÐce row of each species was seeded next to each row in two of the six total replications for root measurements of selected shrubs and trees.
The LSS had nine sprinklers 6 m apart on 1.6 m high risers with each sprinkler producing an overall wetted radius of 13.5 m. Irrigation water was applied daily for at least 10 days after sowing to maintain available moisture in the upper 3 cm of soil nearest the line source. Applied water was measured in catch cans after every irrigation and plant density was counted in one or two subsamples, 1 m long within the seeded rows, every one to three weeks until November at distances of 1.5, 4.5, 6, 7.5, 10.5, 13.5, and 16.5 m from the line source. Soil matric potential (MPa) was measured every minute and hourly averages recorded using gypsum cells connected to Campbell ScientiÐc, Inc. CR-10 electronic microloggers. Two gypsum cells were placed at depths of 1È3, 8È10, and 18È20 cm, while one cell was placed at depths of 38È40 and 58È60 cm in each of three blocks at distances of 1.5, 6, 10.5, and 16.5 m from the line source.
Plant height was measured every one to two weeks on the same Ðve plants of four woody species ( jojobaÈSimmondsia chinesis ; cat claw acaciaÈAcacia greggii ;
velvet mesquiteÈProsopis juliÑora var. velutina ; and blue palo verdeÈCercidium Ñoridum) at distances of 1.5, 7.5, and 13.5 m from the line source. Tap root length of excavated plants was measured at the same distances and for the same species as plant height.
26 TABLE 1 Grasses, shrubs, and trees direct-seeded in a line source sprinkler irrigation experiment at Tucson, Arizona, in the summers of 1992 and 1993 Common name Family, tribe, or subfamily Species Grasses Purple threeawn a Agrostideae Aristida purpurea Nutt.
Cane bluestem Andropogoneae Bothriochloa barbinodes (Lag.) Herter Spike dropseed Agrostideae Sporobolus contractus Hitchc.
Galleta Zoysieae Hilaria jamesii (Torr.) Benth.
Arizona cottontop Paniceae Digitaria californica (Benth.) Henr.
Lehmann lovegrass Festuceae Eragrostis lehmanniana Nees Shrubs Creosotebush Zygophyllaceae L arrea tridentata (DC.) Cov.
Triangleleaf bursage Ambrosieae Ambrosia deltoidea (Torrey) Payne Jojoba Buxaceae Simmondsia chinensis (Link) Schn.
Cat claw acacia Mimosoideae Acacia greggii A. Gray Trees Ironwood Papilionoideae Olneya testota A. Gray Velvet mesquite Mimosoideae Prosopis juliÑora (Swartz) DC. var.
velutina (Wooton) Sarg.
Paloverde Caesalpinoideae Cercidium Ñoridum Bentham aSeeded in 1992 only.
Sonoran Desert Revegetation 27 SigniÐcance of block and species e†ects on plant density, tap root lengths, and shoot height was determined by univariate analysis of variance. Irrigation treatments using LSS are not randomized (Hanks et al. 1980). Repeated measures multiple analysis of variance was used to determine signiÐcance of irrigation and irrigation 3 species interactions for plant density (Potvin and Lechowicz 1990 ;
Fernandez 1991 ; Kuehl 1994). SigniÐcance of MauchlyÏs criterion and failure to meet appropriate Huynh-Feldt conditions precluded use of univariate tests and dictated the use of multiple analysis of variance for these tests (Potvin and Lechowicz 1990). ProÐle transformation in the repeated measures analysis of variance was used to compare plant density at one irrigation level with the adjacent and next lower irrigation level (Fernandez 1991). Grasses and woody species were analyzed separately and a separate analysis was conducted for each sampling date.
Irrigation, W ater Availability, and Overall Plant Response Natural precipitation after sowing for July and August totaled 76.2 and 68.1 mm in 1992 and 1993, respectively (Figure 1). Total irrigation plus precipitation for July and August after sowing ranged from 356.7 to 106.9 mm in 1992 and 285.2 to 106.6 mm in 1993 from 1.5 to 13.5 m from the line source. Potential evapotranspiratio n calculated from the Arizona Meterological Network (2000) weather station for Tucson using a modiÐcation of the Penman equation for July and August totaled 414 mm in 1992 and 412 mm in 1993. Potential evapotranspiratio n from the start of irrigation to the end of August was 304 mm for 1992 and 255 mm for 1993. Potential evapotranspiratio n during the irrigation period was 53 mm greater in 1992 and 30 mm greater in 1993 than the total rain plus irrigation at 1.5 m from the line source (Figure 1). From the start of irrigation to the end of August, air temperatures averaged 29.1¡C in 1992 and 29.9¡C in 1993, while relative humidity averaged 48% in 1992 and 36% in 1993. For both years, irrigation was applied frequently right after sowing in July, while natural precipitation was most frequent in August (Figure 2). This pattern of water inputs, as well as the initially wet soil proÐle, mainly resulted in measured di†erences in soil water availability between irrigated and unirrigated soils, rather than di†erences among irrigation levels (Figure 3). Soil water at the 1È3 cm depth from 1.5 to 10.5 m from the line source was highly available through mid August, while unirrigated soils were dry at that depth until natural precipitation fell in early August. At 8È10 cm, soil water potential on unirrigated soils declined prior to the initiation of summer rains in 1992 and 1993, and after cessation of rainfall in mid August in 1993. Soil water was available below a depth of 18 cm through July and August for both irrigated and unirrigated soils.