Influence of chilling requirement on the southern distribution limit of exotic Russian olive (Elaeagnus angustifolia) in western North America, The

2011 Spring.Includes bibliographical references.Russian olive (Elaeagnus angustifolia L.), a Eurasian tree, is now a dominant species along rivers in western North America. The southern boundary of Russian olive distribution in western North America runs through southern California, Arizona, New Mexico and Texas. I related the distributional pattern of Russian olive to temperature regime and investigated potential temperature-dependent mechanisms that might explain this distributional limit. Specifically, I investigated whether lack of cold temperatures at the southern limit may prevent the accumulation of sufficient chilling and inhibit dormancy loss of seeds and buds, potentially constraining Russian olive's southern distribution boundary. First, I used field observations to thoroughly define the southern limit of Russian olive across western USA and related this distribution to temperature variables. I found that Russian olive occurrence was more strongly associated with low winter temperatures than with excessive summer heat. I then carried out controlled seed germination and vegetative bud-break experiments and a field survey of fruit production and seed viability. Next, I cold stratified Russian olive seeds in growth chambers with temperature regimes simulating six locations along a latitudinal gradient from Socorro, New Mexico (33.8°N latitude) to Presidio, Texas (29.56°N latitude). Both germination proportion and germination times were highest under temperature regimes simulating locations near the southern range limit and declined for temperature regimes simulating locations north and south of the range limit. This pattern only weakly supports the hypothesis that germination would decrease south of the southern range limit. I then conducted an additional controlled germination experiment containing treatments with varying levels of cold stratification, the results of which suggest that the chilling requirement for germination is partly responsible for the southern range limit. Both seed germination proportion and germination time decreased when the amount of cold stratification dropped below values typical of the southern range limit. I also carried out a preliminary bud-break experiment where Russian olive cuttings that were pre-exposed to natural chilling were subjected to various levels of additional chilling in a refrigerator. The results of this experiment suggest that a high percentage of buds burst with chilling values less than those typical of the southern limit. I then conducted a controlled bud-break experiment with cuttings that accumulated varying levels of chilling naturally. Percent bud break decreased when chilling dropped below values typical of the southern range limit, suggesting that the chilling requirement for bud-break is partly responsible for the southern range limit. In 17-65% of the years from 1980-2000, the chilling accumulated at a site near the southern range limit (El Paso, Texas) would lead to a 10% or more decrease in bud-break. The potential decline in growth could have large fitness consequences for Russian olive trees. Finally, I collected fruit production and seed viability data. While fruit production did not decline towards the southern range limit, seed viability declined with decreasing latitude. If climate change follows a warming trend, it is very likely that the chilling requirement for bud-break of Russian olive trees will not be met in some years and this combined with decreased seed viability at lower latitudes may cause its southern range limit to retreat northward. The retreat of a widespread non-native species, such as Russian olive, may present land managers and ecologists with a unique restoration opportunity