Report on Vegetation Conditions

ICWD Home > Research and Reports > Annual Reports > 2004-2005 Monitor >

by Sara J. "Sally" Manning, Ph.D., Vegetation Scientist
July 13, 2005, revised October 26, 2005

In 2004, vegetation conditions in areas monitored by Inyo County Water Department (ICWD) generally showed either a pattern of holding constant relative to recent years (e.g. since 2001) or of declining.

ICWD monitors vegetation conditions throughout Owens Valley each summer. In the early 1990s, the Technical Group agreed to re-inventory plant cover and species composition in selected vegetation parcels to compare each year's conditions with the Water Agreement's baseline data. The baseline data were collected during the mid 1980s by LADWP staff. ICWD staff began collecting field data in 1991, and the program has continued since that time.

The recent elimination of ICWD seasonal staff positions (due to budget concerns) resulted in a cut back in the re-inventory work during 2003 and 2004. Ninety-six parcels were re-inventoried in 2002, but only 68 and 73 parcels were monitored in 2003 and 2004, respectively.

In addition to tracking changes since the baseline period, the re-inventory data are also analyzed for parcel water table and vegetation status with regard to the requirements of the Drought Recovery Policy (DRP). Following each summer season, data on depth to water table are obtained from ICWD hydrologists, analyses are performed, and a status report on conditions is presented to Inyo County and LA decision makers. The report on conditions through 2004 is entitled: Status of Re-Inventoried Vegetation Parcels According to the Drought Recovery Policy, 2004, by Sara J. Manning, April 27, 2005. (Available in pdf here.)

Data from 1991 through 2004 show that, in parts of most wellfields, water table and perennial vegetation have failed to recover since the 1987-1992 drought when LADWP pumped large amounts of water from Owens Valley. In terms of the DRP, of the 73 parcels re-inventoried in 2004, 29 parcels were classified as still subject to the management constraints imposed by the policy. This means that conditions have persisted below the Water Agreement baseline in these affected areas of the valley at least since 1991 (when monitoring began) and probably for a few years longer (prior to 1991).

Most of the other parcels re-inventoried in 2004 were classified either as Controls or as "Free" from the management restrictions of the DRP. Twenty parcels were designated as Controls, that is, not affected by pumping during the 1987-1992 drought. Although water tables and perennial plant cover have varied in the Control parcels since the baseline period, the range of variation in water table has been relatively small compared to the water table variation in the monitored wellfield parcels, and perennial cover in the Control parcels tends to be equal to or above baseline levels. The 21 parcels classified as "DRPfree" showed clear evidence that water tables recovered since 1992, and perennial plant cover responded to this recovery. In all cases, this "recovery" occurred during the period from 1995-2001. That is, no monitored "DRP" wellfield parcel has exhibited recovery since 2001. In addition, since 2001, many of the DRPfree parcels have exhibited declines in water table and perennial cover and some exhibited cover statistically significantly below baseline in 2004. Three additional parcels were re-inventoried in the Laws area beginning in 2003. Because of the 14-yr gap in field data, these three parcels have not been classified according to the DRP. However, all three showed extremely low cover, significantly below baseline levels.

In addition to data collected in the field by ICWD, ICWD obtains total green plant cover estimates that are derived annually from Landsat satellite data. These data are purchased and analyzed by Dr. Andrew Elmore of Dartmouth College. During a 3-yr collaboration with ICWD scientists, Dr. Elmore and other university researchers developed a technique for assessing live plant cover throughout the valley based on a late summer satellite scene. Dr. Elmore has continued to analyze cover on an annual basis, such that ICWD has a complete late summer record of vegetation cover for 1986-2004. For ICWD's re-inventoried parcels, the Landsat-derived cover estimates were presented in Manning's 2005 report (referenced above).

The various forms of vegetation cover data have been used to study relationships between water availability and cover. To perform these analyses, best available estimates of each site's depth to water table (DTW) and cumulative annual precipitation were obtained, and relationships were examined using linear regression. Linear regression detects whether an incremental change in an environmental factor (e.g. a change in DTW) is associated with a predictable incremental change in a response factor (e.g. vegetation cover).

Despite many limitations and potential errors in the available data, some strong relationships have been detected between water availability and cover. These results are helping develop a clearer picture of the interaction between water table, precipitation, and phreatophytic plant cover in Owens Valley, and this information should be useful to water managers. Some highlights are discussed below.

The data show that, in Control areas, where water tables have fluctuated minimally during the past 19 years, plant cover does not appear to respond to the small fluctuations in DTW. In some cases, a minor correlation is detected between precipitation and cover (with cover increasing somewhat in high rainfall years) in Control areas. It is assumed that, because these sites experience a constant supply of water from the water table, the plants are not very "water-limited." Thus, when their cover varies, it is probably due to changes in other factors, such as disturbance or herbivory. Furthermore, as mentioned with regard to the field data, cover tends to remain high in Control sites.

In areas affected by pumping or by water spreading, water availability can become a constraint on vegetation cover. Analyses revealed that when management activities resulted in fluctuations in the water table (typically more than 2m during the 19-yr period), vegetation cover responded to changes in water availability. Results show that when the water table fluctuated in the root zone of the dominant Owens Valley phreatophytic species, vegetation cover was correlated with DTW. However, once the water table dropped below and stayed below root zones, further fluctuations in cover were not correlated with changes in DTW. At the point where groundwater is no longer available, total plant cover at the site is much lower than it had been when roots had access to the water table, and total plant cover begins to respond, weakly, to precipitation. (See case study for two parcels, Blackrock 99 and Blackrock 94.)

These results have implications for the parcels ICWD re-inventories. Analyses show that most of the wellfield parcels classified as "DRPfree" show a correlation between cover and change in DTW. That is, as the water table rises, cover increases, and as the water table drops, cover decreases. Groundwater levels associated with DRPfree parcels typically fluctuate within or just below the root zone, and the analyses strongly suggest that these fluctuations drive changes in cover. The relationship between DTW and cover is strongest for meadow DRPfree parcels located near the lower edge of the Sierran alluvial fans and in Laws. In contrast, for the DRP parcels, DTW typically remains below or deep within the root zone, and cover is not correlated with changes in DTW. Most DRP parcels exhibit low total cover values. Analyses often reveal that the meager cover is significantly correlated with precipitation; that is, cover increases somewhat in high rainfall years. However, the results suggest that even a very large rainfall event has little chance of returning the site cover to baseline levels and/or re-establishing functioning baseline conditions.

Case Study: Two Wellfield Parcels

Southwest of the Blackrock Fish Hatchery in the Thibaut Sawmill wellfield, parcels Blackrock 99 and Blackrock 94 are located next to each other near the toe of the Sierran alluvial fan (Figure 1, below). In 1986, LADWP mapped both parcels as alkali meadow and measured baseline perennial plant cover in parcels 94 and 99 as 41% and 48%, respectively. Blackrock 94 lies west of the aqueduct, and most of Blackrock 99 is also west; a small portion was mapped by LADWP on the east side of the aqueduct. Landsat cover estimates for both parcels are available for the entire period from 1986 (the baseline year) through 2004. In addition, relatively good quality data exist for estimating each parcel's DTW and cumulative annual precipitation each year.

Figure 1: Features of example parcels: Blackrock 94 and 99 in Thibaut Sawmill wellfield.

Linear regression reveals that cover in Blackrock 99 (as measured in late summer each year by satellite) is correlated with DTW and that precipitation contributes no measurable influence on cover. During the period since baseline, the water table has fluctuated beneath Blackrock 99, but the average April DTW beneath this parcel has remained within reach of roots of the phreatophytic vegetation (i.e., shallower than 3m) (Figure 2, below). In contrast, the data show that, since 1988, cover in Blackrock 94 is correlated with precipitation and not with DTW. Since 1988, water tables have been below the root zone of this meadow parcel (at 4m or deeper) (Figure 3, below).

Figure 2A: Blackrock 99 - DATA. The charts above show (top) the total green plant cover measured in late summer by Landsat each year from 1986 through 2004 and (bottom) the estimated April depth to water table (DTW) beneath the parcel each April, 1985-2004. In these graphs, the data are arranged sequentially by year. The cover measured by satellite during 1986 was 58% and is highlighted by the bold line. The bottom graph shows that water table fluctuated during the 20-yr period, but did not drop below 4m. (The Green Book set root zones for management at 2m for grasses and 4m for phreatophytic shrubs. These are shown in the bottom graph by green and yellow shading.)

Figure 2B: Blackrock 99 - ANALYSIS. In these graphs, Landsat cover for Blackrock 99 was rearranged and graphed based on increasing depth to water (top) and increasing precipitation (bottom). The relationship between cover and DTW was analyzed using linear regression, and a significant correlation resulted (p = 0.00). The significant result is shown by the line drawn through the data points and the equation in the bottom corner of the graph. These results indicate that as water table declined in this parcel, cover also declined (and as water table increased, cover increased). The coefficient of determination is R2 = 0.77; this means that fluctuation in April DTW accounted for 77% of the change in the cover data. When cover was compared with precipitation, no significant relationship was detected by linear regression (p = 0.09). Therefore, the analysis did not reveal a relationship between changes in amount of precipitation and changes in vegetation cover in Blackrock 99.

Figure 3A: Blackrock 94 - DATA. The charts above show (top) the total green plant cover measured in late summer by Landsat each year from 1986 through 2004 and (bottom) the estimated April depth to water table beneath the parcel each April, 1985-2004. In these graphs, the data are arranged sequentially by year. The cover measured by satellite during 1986 was 47% and is highlighted by the bold line. The bottom graph shows that water table fluctuated considerably during the 20-yr period. By April 1988, the water table was below the 4m root zone, and according to these estimates, it remained below 4m through 2004. For the analysis, cover and DTW data for years when the water table was in the root zone were omitted (the omitted data are shown by the gray shading).

Figure 3B: Blackrock 94 - ANALYSIS. For this analysis, only data from 1988 onward were used. By eliminating 1986 and 1987, the analysis focuses on the roles of water table and precipitation in situations when the water table is below the management root zone. In these graphs, 1988-2004 Landsat cover for Blackrock 94 was rearranged and graphed based on increasing depth to water (top) and increasing precipitation (bottom). Linear regression did not detect a significant correlation between water table and vegetation cover (p = 0.12). Although the water table variation 1988-2004 went from 4m to nearly 8m, the pattern of vegetation change was not correlated with these large fluctuations (top graph). However, when cover was compared with precipitation, a statistically significant relationship was detected by linear regression (p = 0.004) (bottom graph). The analysis shows that, once the water table is below the root zone, precipitation may influence cover. In this example, fluctuation in annual precipitation amounts accounts for 43% of the change in the cover data. It is noteworthy that, without the presence of groundwater in the root zone, the highest cover achieved 1988-2004 was 26%, thus well below the 1986 baseline value of 47%. In fact, the regression result indicates that, with an Owens Valley average rainfall of about 14 cm, approximately 18% cover can be sustained at this site.