Prediction of annual runoff in ungauged basins - Runoff Prediction in Ungauged Basins: Synthesis across Processes, Places and Scales - page 76

Catchment (biological) processes
Vegetation cover affects both the wetting and the drying
phases of the catchment

a)

Vegetation - Ecosystem
Engineering

Competition
Runoff

s response to precipitation. Vege-
tation adjusts, acclimatises and adapts its physiology to
conditions of different water availability on time scales
ranging from individual storm events (of the order of
minutes to hours) to evolutionary change (of the order of
decades to millennia). Vegetation links water availability
to geomorphic and pedological change, so that on long
time scales the physical and biological components of
catchments co-evolve.
In the wetting phase, the dominant effect of vegetation
cover is to reduce water availability for catchment wetting
through interception by plant leaves and by leaf litter
(Gerrits et al., 2007 , 2010 ). The proportion of precipitation
lost due to interception is of the order of 10

'

Runoff

Infiltration

Infiltration

b)

Competition sustains
bare “runoff catchments”

30%, and may
be the component of the catchment water balance that is
most sensitive to vegetation change (Brown et al., 2005 ).
Interception losses tend to be inversely proportional to
precipitation intensity and directly proportional to density
of the vegetation (Muzylo et al., 2009 ). In catchments with
low precipitation intensity and dense vegetation cover,
interception losses are higher, e.g., if savanna ecosystems
in Botswana receive less than 400 mm of annual precipita-
tion, interception loss is close to 100% of the annual water
balance (Savenije, 2004 ).
During the drying phase, vegetation imposes significant
changes on the dynamics of evaporation, first through a
trade-off between bare soil evaporation (dominant where
vegetation cover is sparse) and transpiration (dominant
when canopies close and vegetation cover is extensive)
(Laurenroth and Bradford, 2006 )( Figure 5.7 ). Vegetation
shades soil surfaces, increases near-surface humidity, and
increases the aerodynamic roughness of the land surface,
suppressing evaporation beneath vegetation canopies (Kel-
liher et al., 1993 ). The presence of vegetation modifies
catchment drying in two important ways. First, plant root
systems extend throughout the subsurface and allow
reserves of water that would otherwise be isolated from
significant evaporative demand to be connected to the
atmosphere. Phreatophytic plants, for instance, directly
tap groundwater reserves. Second, transpiration of water
is regulated by stomata. Stomata allow plants to regulate
transpiration in two ways. During periods when conditions
are unfavourable for photosynthesis, plants shut stomata.
Thus, transpiration tends to be very low during low-light
conditions or overnight, for example. More significantly,
even when conditions are suitable for photosynthesis,
plants may shut stomata to prevent unfavourably low water
potentials occurring in their canopy. Thus, even when
atmospheric conditions become favourable for drying,
vegetation stomatal responses may inhibit transpiration,
slowing the rate of catchment drying.

-

c)

Figure 5.7. Vegetation ' engineers ' its environment, often creating
conditions that sustain further vegetation growth. The bare areas
generate Hortonian runoff, which runs downslope and infiltrates in
vegetated sites (a and b). Competition for water between individual
plants leads to vegetation growing in organised patterns (c), allowing
higher biomass and transpiration than could occur in the absence of
the feedback between vegetation presence and soil properties.

Vegetation also displays adaptive features on seasonal
and inter-annual time scales. For instance, many plants
grow fewer leaves or actively shed leaves during periods
of water stress, creating a relationship between leaf area
and water availability, and reducing the transpiring surface
area. These relationships may allow water balance parti-
tioning to be inferred from observation of metrics of
catchment

such as the normalised difference
vegetation index (NDVI). For instance, the Horton index

' greenness '

-
a ratio of evaporation to plant-available water on annual
time scales (Troch et al., 2009 ) appears to be signifi-

cantly and negatively related to catchment-scale annual
NDVI in most water-limited basins (Brooks et al., 2011 ),
and to mean annual NDVI across 320 test basins in the
USA (Voepel et al., 2011 ). Ultimately, vegetation strat-
egies modulate transpiration to prevent negative extremes
in plant water potentials, and in doing so reduce variability
in catchment water balance. When water balance is meas-
ured in terms of the Horton index, for instance, inter-
annual variability is markedly damped, primarily being
expressed in terms of the partitioning of water to rapid
runoff generation (Troch et al., 2009 ), and with the greatest
sensitivity in water balance arising in the fast flow com-
ponents that are effectively sequestered from vegetation

-