Abstract:We studied leaf and twig functional traits, namely, leaf thickness(LTS), leaf area(LA), specific leaf area(SLA), leaf dry weight of every leaf (LM), leaf dry matter content(LDMC), leaf tissue density(LTD) and twig length(TSL), twig diameter(TSD), leaf area ratio(LAR), leaf/stem mass ratio(LSMR), Twig stem wood density (TSWD) and Huber value(HV) of Populus euphratica Oliv. in Tarim extremely arid area. We analyzed the variation characteristics of these functional traits along the groundwater depth gradient (GWD), and the tradeoff between leaftwig traits and functional traits combination by Oneway ANOVA, Pearson correlation and principal component analysis(PCA), in order to reveal the ecological strategy of P. euphratica to adapt the extremely arid desert environment. The results showed that: (1) the variation coefficient was from 8.61% to 59.31%, the HV was the largest (59.31%), LDMC and LAR were the smallest (8.61%, 9.75%) and the most conservative traits. (2) Variance analysis showed that thirteen traits of P. euphratica had significant difference between GWD 1.5 m and 4.8 m, and there were significant difference of LTS, LTD, TSL and HV between GWD≤2.4 m and GWD≥3.5 m. (3) Correlation and principal component analysis (PCA) showed that LA, SLA, LDMC, LTS, LTD, LAR, TSWD and HV could be used as important indicators to reflect the characteristics of P. euphratica twigleaf traits. There were significant correlation between 28 pairs of leaftwig traits(P<0.05), especially HV was significantly negative correlation with TSL and positive correlation with TSD(P<0.05), and TSL was negative correlation with TSD, as well. HV was significantly negative correlation with LAR, LSMR, LA, LM(P<0.05). Also, leaf display efficiency(LAR and LSMR) were significantly negative correlation with LDMC(P<0.05), and TSWD was negative correlation with LA and SLA. It indicated that there was a potential tradeoff mechanism between machinewater safety and carbon availability of P. euphratica in arid areas. P. euphratica adopted the ecological strategy, such as reducing LA, SLA and leaf display efficiency (LAR and LSMR), increasing LTS, LDMC, TSWD and HV, which were beneficial to reduce water loss, storing resources (nutrients and water) and enhancing resilience, to adapt the aridbarren desert environment. Plant economic spectrum also existed in desert ecosystems. With global warming and regional water shortage, it generally shifted towards the conservative strategy of slow investmentreturn.