INFLUENCE OF UNDERSTORY REMOVAL, THINNING AND P FERTILIZATION ON N-2 FIXATION IN A MATURE MESQUITE (PROSOPIS-GLANDULOSA VAR GLANDULOSA) STAND

The natural abundance N-15/N-14 method was used to estimate the influe nce of silvicultural and P fertilization treatments on N accretion, N- 2 fixation and N partitioning among tissues in a mature mesquite Proso pis glandulosa var. glandulosa stand in Texas. The silvicultural treat ments consisted of understory removal, herbicide treatment of brushy r esprouts, thinning trees to single stems and 100 kg ha(-1) P fertiliza tion. The trees had a mean basal diameter of 17.8 cm with 8 to 35 cm r ange. The stand was slow growing with the increase in dry matter rangi ng from 0.465 Mg ha(-1) year(-1) to 0.701 Mg ha(-1) year(-1) for the 8 years after the treatments were applied. N accretion after 8 years ra nged from 3.1 kg ha(-1) year(-1) to 4.4 kg ha(-1) year(-1). Due to the range in delta(15)N of the leaves, twigs, branches and trunk, we used the weighted (by biomass) average delta(15)N per tree in calculations of the percent N derived from N-2 fixation (%Ndfa). There was conside rable variability in delta(15)N Of the reference plants, i.e. from 3.3 to 5.9. In contrast there was low variability in the background delta (15)N of nearby soils (7.0 +/- 1.0). As the total above-ground biomass delta(15)N Of a grass grown outside the influence of mesquite (7.8 +/ - 0.58) had the same delta(15)N as the soil (7.5 +/- 1.0), we used the grass outside the influence of mesquite and the weighted tree mean de lta(15)N to calculate % of N derived from N-2 fixation. The decrease i n intraspecific competition by thinning multistemed trees to single st emmed trees was the only treatment that significantly (p = 0.0001) inc reased growth. Interspecific competition, i.e. understory removal, did not increase growth. There were no significant differences in total N production or N fixation among treatment means. The most striking res ult was the highly positive correlation between tree delta(15)N and to tal N per tree and biomass per tree (R-2 = 0.90, F = 164.4, df. = 18, Mean square error (MSE) = 0.155, p 0.0001). This implies that the youn ger trees colonizing infertile soils relied more heavily on N-2 fixati on than larger trees which accumulated 1200 kg ha(-1) more N under the ir canopies. The percentage N derived from N-2 fixation ranged from 63 to 73% in the various treatments. Despite the high percentage of N de rived from N-2 fixation, the N-2 fixation of the stand was very low, i .e. 1.98 to 2.80 kg N ha(-1) year(-1), due to the low growth of the st and. We believe that comparisons of the whole tree weighted delta(15)N to background soil delta(15)N provides a more reasonable approach to estimate % N-2 fixation than comparisons of leaves of fixers and refer ence plants.