Abstract
Prosopis pugionata (Fabaceae) is a halophyte tree species that grows exclusively in Argentina’s arid environments, whose woodlands are mainly affected by fires and overgrazing. Here, we describe the wood anatomy and fire wounds of P. pugionata and their relationship with plant size and bark thickness. Besides, we attempt to determine the potential of P. pugionata for dendrochronological studies in order to date fire events throughout time. We tested the hypothesis that P. pugionataforms datable fire wounds, allowing its use for dendrochronological studies. The study area is located in the arid Argentine Chaco region. Seventeen individuals, varying from 15 to 65 years of age, were randomly sampled and bole disks were taken at 0.3, 1.3 and 2.3 m high. P. pugionatais a diffuse and semi-ring-porous hardwood species with growth rings delimited by marginal parenchyma bands. The bark is longitudinally fissured, with an average thickness of 0.19 cm, and a scarce increase with age. Seventy-seven percent of the samples showed fire wounds (). We identified fire scars (70%) and marks (30%), and differentiated them by wood growth interruption in the former. Sixty-one percent of fire wounds affected less than 20% of the cambial perimeter and there were no significant effects of age, bole diameter, bole height and bark thickness on the cambial damage percentage. After cross-dating, P. pugionata showed a mean correlation value between series of 0.5, which represents an accurate potential for dendrochronological studies. Twenty wildfires were dated from 1943 to 2007 in the study area with a fire frequency of 0.40 fires/year and a mean fire interval of 2.5 years. Wildfires frequency increased from 1980 to 0.48 fires/year and a mean fire interval of 2 years. Our results indicate that P. pugionata has the potential to date fires, considering its potential for dendrochronological studies, its ability to survive recurrent fires and forms identifiable wound scars in the wood.
Two-Stage Estimation for Seemingly Unrelated Nonparametric Regression Models
