|
<br />60
<br />
<br />F, ZAMORA-ARROYO ET AI-
<br />
<br />Table 2. Comparison of area of nafive free habita~ ('> 1O~ P.opulus fremontii
<br />and Salix gooddingii) and shrub habitaf (Tamanx ramoslSSlma and Pluchea
<br />sericea) on the regulated strefch of the lower Colorado River in the United Sfafes,
<br />from Davis Dam fO me Normerly International ~OU~ry, ~nd o.n fhe unregulated
<br />stretch in Mexico, from the NIB to the Juncf!on With Rw Hardy
<br />
<br />Habitat type U.S. stretch Mexico stretch
<br /> % ha % ha
<br />P. fremontii + S. gooddingii > 10%:
<br />Open gallery forest 0 0 12.7 1818
<br />Closed gallery forest 0,3 98 0 0
<br />Shrub-dominated 4,3 1460 14'3 2045
<br />Total 4.6 1558 27'0 3863
<br />P. fremontii + S. gooddingii < 10%:
<br />T. ramosissima/P. serieea 54.1 18,453 73-0 10,453
<br />T. ramosissima/Prosopis 31'7 10,829 0 0
<br />Other 9-6 3273 0 0
<br />Totals 100 34,096 100 14,316
<br />
<br />Gallery forest has > 80% (closed gallery) or > 35% (open gallery) ~verstory trees. The '?ther' c~tegory for
<br />the U.S. Stretch includes emergent marsh and Atriplex (saltbush) habitat nol encountered In the river stretch
<br />surveyed in the delta.
<br />
<br />Comparison of native tree eover on U.S. and Mexico portions of the river
<br />
<br />We compared the number of hectares of native tree habitat in the delta with estimates for
<br />the regulated portion of the river (above More1os Dam) made by BaR. The results
<br />(Table 2) show that the delta supports 2'5 times as much native tree habitat as the stretch
<br />from Davis Dam below Grand Canyon, to Morelos Dam (six times more per unit area).
<br />Approximately 1800 ha of gallery forest has regenerated in the delta, compared to only
<br />a single stand of 98 ha on the regulated stretch, and th!S patch actually is in the delta of
<br />the Bill Williams River, a tributary of the Colorado River (Ohmart et al., 1988).
<br />
<br />Timing and flow rates of water releases to the delta
<br />
<br />We examined flows to the delta over the period 1992-1999 to correlate flows with
<br />vegetation data. Water releases during major releases varied in volume from less than
<br />100 m3s -I to over 1000 m3s - 1 (Fig. 6). We conducted an overflight in February 1997,
<br />when releases were 80-100 m3 s - I according to IBWC data, to document the extent of
<br />flooding from a low-volume release. We observed extensive. overbank flooding .of ~e
<br />river within the levee system, and w~ter was exiting the delta mto.the Gulf of Cahforma
<br />via the river channel and sheet floodmg of the lower delta f1oodplam, Furthermore, water
<br />was flowing into Laguna Salada, a below-sea-Ievel depression west of the del.ta. Me~ico
<br />Highway 2, which crosses the southern part of the delta, was flooded and Impassible.
<br />Progressively larger volumes of water, released 1997-199,9, flooded great~r areas of
<br />floodplain within the levees and in I.:agu~a S~lada, but dl? not flood agncul~ral .or
<br />urban areas. Discharges occurred m81n1y m wmter and spnng (February-;Apnl), With
<br />one fall release (September-December 1998) and almost no releases m summer.
<br />
<br />
<br />REGENERATION OF TREES IN RESPONSE TO FLOOD RELEASES
<br />
<br />(a)
<br />1992 ~ 1994 ( 1996 (
<br /> I ,
<br />,....r r/ i-!
<br />f
<br /> ! /
<br /> '"
<br /> . ~',
<br />1997 ( 1998 f! 1999}
<br />ri f r'"
<br /> (
<br /> / /
<br /> .. ,.
<br />
<br />
<br />
<br />
<br />
<br />% Vesel.lion
<br />
<br />Iwa..r
<br />~ ~ BaK Jllil
<br />20-2~/o
<br />30-39%
<br />40 -490/.
<br />
<br />150-59%
<br />60-69%
<br />70-79'%
<br />80-89%
<br />90-100%
<br />
<br />.t:" 600
<br />'0
<br />1;)
<br />~ 500
<br />5
<br />~ 400
<br /><Ii
<br />'iil 300
<br />~
<br />I;:;
<br />~ 200
<br />.c
<br />o
<br />~loo
<br />o
<br />;3 0
<br />",C?- ';1\~ ~ ~<I' ';1\,,<1' <!,..l ';1\,,<1' <t...,,<I' rfI,,,<I'
<br />,,,,,,,.go ,~ ...~ ,~ ....~ ...'" ...'"
<br />
<br />(b)
<br />
<br />cr.~
<br />c:;
<br />CJ..')
<br />(0
<br />r"'"
<br />~...
<br />
<br />
<br />64
<br />
<br />
<br />60
<br />
<br />.s
<br />is 56
<br />1l.
<br />'!t
<br /><f. 52
<br />48
<br />
<br />] 2 3
<br />Years of river flow
<br />
<br />4'
<br />
<br />Figure 6. Relationship between % vegetation cover and flood events in the Colorado River delta,
<br />Mexico. TM images (a) showing summer vegetation before and after major flood events (b)
<br />(arrows show dates ofTM images) were classified using NDVl to show % vegetation cover in the
<br />native tree zone (the riparian zone north of the dark line across each image). (c) is the regression of
<br />% vegetation on the number of prior years .of water discharge,
<br />
<br />Correlation between vegetation C()'l}er and flow releases, 1992-1999
<br />
<br />We conducted a change analyses of vegetation density as affected by flood flows
<br />into the delta, using satellite imagery for past years to estimate summer vegetation cover.
<br />We used reflectance-based, NDVI values to estimate % cover on six TM images
<br />covering the period 1992-1999. We restricted the analysis to the 100 km stretch of river
<br />containing native trees, from Morelos Dam to the junction with Rio Hardy.
<br />Vegetation cover, as estimated by NDVI values on satellite images of the delta for
<br />different years, showed an apparent positive response to flood flows (Fig. 6). We
<br />quantified the relationship by calculating % vegetation cover in the first 100 km of river
<br />below Morelos Dam for years before and after each flow event in the 1990s. We found
<br />a positive relationship between % vegetation and the total of the three previous years'
<br />volume (calculated from flow rates over time)(r = 0'80-0'82*). However, the strongest
<br />correlation was simply with the number of previous years of flow irrespective of volume
<br />r = 0'97***). Thus, the lowest cover (ca. 50%) was present in 1992 and 1996,
<br />years which were preceded by three or more years without river discharge
<br />(Fig. 5). Vegetation cover was ca. 53% in 1994 and 1997, following one year of
<br />discharge. These values were similar even though the 1993 discharge peaked at
<br />> 500 m3 s - I compared to < 100 m3 s - I in 1997. Then, vegetation cover increased
<br />progressively after 1997 as discharges continued in 1998 and 1999, reaching 62% after
<br />3 years of discharge.
<br />
|