1. Fig. 5 – MODIS: Summertime stratus cloud cover over Santa Cruz Island – % of cloud-free days at 10:30AM from July 3 to Sep. 30 for 2000 – 2004. Black polygons represent approximate locations of P. muricata. Pines exist at elevations where mountainsides tend to intercept cloud cover. Fig. 4 - Average cumulative change in trunk circumference of P. muricata measured with dendrometer bands. Sample size grew from n=2 in June 2004 to n=39. Change in circ. Since May ’04 (mm) Fig. 6 – 24-hr running potential evapotranspiration on western Santa Cruz Island from Dec. 19, 2003 – Aug. 4, 2005. Gray areas highlight extended periods of summer cloudiness/fog. 4 Despite zero rainfall during most summers, P. muricata manages some of its fastest growth rates through August (Fig. 4). The relatively low r-values of the correlations with growth, 13 C, and rainfall suggest that winter rain is not the only major source of moisture in this strongly water-limited environment. Figure 5 shows that stands of P. muricata do not necessarily grow in the areas with the most cloud cover, but at the elevations where the stratus marine layer often collides with mountainside. Additionally, figure 6 shows prolonged stretches of decreased potential evapotranspiration (water stress) at one of our weather stations due to summertime cloud cover. The ultimate goal of this research is to use tree ring analysis to reconstruct past climate in areas that do not have extensive instrumental climate records. In particular we hope to reconstruct fog history. This research will be valuable in forecasting future changes in the cloud/biosphere interaction and subsequent effects on biodiversity. Our preliminary research has been on the California Channel Islands (Santa Cruz & Santa Rosa Islands). Results from the first tree ring width and carbon isotope analyses on these islands show significant correlation with winter rainfall. However, ground observations and satellite data from MODIS/Terra suggest there must be important summer fog information stored in tree rings as well. 1 Fig. 3 – P. muricata core segment from Santa Cruz Island - False banding within years may indicate a drought-induced slowing in photosynthesis before revival due to summer fog. 1960 19701965‘61 As a result of rainless summers, these pine trees have abnormal (false) banding patterns, indicative of photosynthetic inconsistency. A good example is the 1964 ring below. We believe that heavy summer fog often prolongs the growing season and kicks a tree back into earlywood growth. Earlywood Latewood 1964 3 Fog and Vegetation on the California Channel Islands: A Tree Ring and Satellite Analysis 1 A.P. Williams, 2 C.J. Still, 3 D.T. Fischer, and 4 S.W. Leavitt Geography Department, 3611 Ellison Hall, UC Santa Barbara, Santa Barbara, CA 93106; 1 williams@geog.ucsb.edu, 2 still@icess.ucsb.edu, 3 fischer@geog.ucsb.edu 4 Laboratory of Tree Ring Research, 105 W. Stadium Bldg. 51, Tucson, AZ 85721; sleavitt@ltrr.arizona.edu Fig. 2 - Linear correlation (r) with total precipitation over different combinations of consecutive months during a 24-month period when p < 0.05. a) Annual latewood 13 C (P. murcata, Santa Cruz Island, 25 years). b) Annual ring width (P. torreyana, Santa Rosa Island, 96 years). Importance of winter precipitation to tree growth a) b) Earlywood and latewood tree rings (illustrated in figure 3) were separated out from four Bishop Pine (Pinus muricata) cores and 13 C analysis was performed on α-cellulose extracted from tree rings grown from 1979-2003. δ 13 C values were normalized by subtracting annual δ 13 C of atmospheric CO 2 (La Jolla, CA). 13C is enriched in plants that undergo water stress (no fractionation for cellulose formation was applied). We also collected 33 cores of Torrey Pine (Pinus torreyana) from Santa Rosa Island in 2005 and created a ring width chronology for 1908-2004. We do not yet have enough data to include fog in this correlation analysis. Both records correlate significantly with annual precipitation in the Santa Cruz Island Central Valley (Fig. 1). Figure 2 shows that only winter & spring (Nov. – April.) rainfall is significant to ring width and 13 C. Fig. 1 – a) Earlywood and latewood annual tree ring 13 C discrimination (‰) in P. muricata trees on Santa Cruz Island. b) Annual ring width index (detrended) for P. torreyana on Santa Rosa Island. c) Annual (Jul-Jun) precipitation in the Santa Cruz Island central valley. a) b) c) 2 Precip. (in) 87% falls Nov. – Mar. 6 Fig. 8 – Correlation of Santa Rosa Torrey Pine ring width with 619 ring width chronologies from the western US, Mexico, and Canada, interpolated using inverse distance weighting. (red=pos. blue=neg.) If there is significant spatial variation in water stress in a given region, such as Santa Cruz Island, the western US, or along a tropical transect that passes through cloud forest, extensive analysis plants that photosynthesize in the summer rainless period. This means that the frequency of summer fog should be recorded within tree rings. We should also be able to distinguish among different water sources (not just water stress) through tree ring H and O isotope analysis because fog and rain are isotopically unique. We have been collecting extensive data on fog and plant response to drought and moisture influence for the past two years. Future goals are to combine this data with remotely sensed imagery to reconstruct a limited history of cloudiness/fog over the islands, cross-check tree ring data with these records, and finally hind-cast variation in cloud cover over the past century using the tree ring information. of tree core data from throughout that region can help us describe those spatial differences and how they have varied in the past. This research suggests that while winter precipitation is essential to vegetation on the Channel Islands, summer fog is equally important to 5 north-facing slopes and in areas with heavy summer cloud cover where direct sunlight is limited. We will next analyze inter- annual NDVI data and compare this with dendrometer and leaf gas exchange (Licor 6400) data that we are collecting currently. We are also measuring monthly xylem pressure potential to assess seasonal changes in pine water status. -1 - 0 0 -.1.1 -.2.2 -.3.3 -.4.4 -.5.5 -.6.6 -.7.7 - 1 NDVI Values Fig. 7 – Monthly Normalized Difference Vegetation Index (NDVI) during 2004 from 250m MODIS/Terra data JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC Chaparral plants and arid southern CA ecosystems are typically dormant during the long summer dry season (last rain in 2004 was March 1). The exceptions to this trend are on