Discussion

The importance of forested catchment area in hydrological cycle is a well explored topic which has been under extensive reviewing from time to time (Bruijnzeel 2004; Chomitz & Kumari 1998; Hamilton 2008; Makarieva et al. 2006). Forest type (native/plantation/primary/secondary), edaphic factors and ground water recharges are some of the major factors which have been investigated by many workers (Brauman et al. 2011; D’Odorico et al. 2010; Kagawa et al. 2009; Muñoz-Villers et al. 2011; Smerdon et al. 2009). Forest cover has definite role in water conservation by soil moisture retention (through humus and thick litter layer) and maintaining microclimate thus reducing evapotranspiration demand (Sikka & Selvi 2006; Thomas & Sankar 2006). Similarly studies have also addressed the effect of afforestation and land conversion in relation to hydrological cycle (Purandara et al. 2006; Sikka & Selvi 2006).

Ecologically, the sacred groves in central Western Ghats region often represent relic primary forests or other old growth forests, usually characterised by large trees (higher basal area), rich litter cover and less compact soils. However, their ecological status is too much varied as per management condition and changing land use practices in the region from British colonial time period (Chandran & Gadgil 1998). Both sacred and non-sacred hill slopes of the current study sites, separated from each other by a single hill, are part of the central Western Ghats. Therefore, overall temperature, precipitation and gross geological pattern are similar for both the study areas. However, differences would certainly exist in microclimate level as reflected in plant species diversity, endemism, periodic soil moisture level and ground water recharge.

Karikan sacred grove occupies south-west slope of the hill and is the source of perennial streams in the locality. The old growth evergreen to semi-evergreen nature of the Karikan forest is in sharp contrast to its neighbouring forested areas which are secondary with more deciduous tree elements. These deciduous elements in a heavy rainfall zone are the results of slash and burn cultivation system of the past. Ever since this system was prohibited by the British over a century ago, the practice of setting fire to the forests diminished in the region, favouring the return of the evergreens in a big way as is evident from the greater numbers of the endemic evergreen trees like Hopea ponga and Diospyros candolleana. Nevertheless, the tree basal areas in the secondary forests of Sambegadde are far less from that of the Karikan hill. NMDS analysis also reveals distinct species composition pattern in two areas. Tree population study in Karikan shows considerable presence (31%)  of high girth class members (> 90 cm GBH) and upper strata at ≥ 30 mt., typical structural characteristic of tropical old growth forest as reported from other parts of Western Ghats and elsewhere in tropics (Chandrashekara & Ramakrishnan 1994; Pascal & Pelissier 1996; Uuttera et al. 2000). Although not saturated, the diversity study has captured 74.66% and 81.21% (Jackknife 1 and Chao1) of species diversity from the study area. High species richness among trees with 38% Western Ghats endemism indicates favourable microclimate especially for diverse endemic species. Presence of typical wet evergreen species like, Dipterocarpus indicus, Polyalthia fragrans, Diospyros saldanha, Knema attenuata, Myristica malabarica etc. (Western Ghats endemics), in Karikan also confirms this contention. Although the non-sacred forest site in the Sambegadde hill has shown less species richness and endemism (22%), estimators assume more species to be found. Tree population structure shows dominance of younger members (86%) i.e. girth class < 90 cm GBH and  lower height level (~ 20-25 mt.) which indicates comparatively younger age of the forest stand, in contrast to the large girth trees with several reaching 30-40 m in Karikan. The dominant species are Hopea ponga, Diospyros candolleana, Aporosa lindleyana, Schleichera oleosa etc. (as per IVI value) of which Hopea ponga (though an evergreen dipterocarp) and Aporosa lindleyana are well recognised disturbance indicators of the forests of the high rainfall areas of South Indian Western Ghats (Gokhale 2005; Pascal & Ramesh 1997).

CCA has found close association between Karikan middle and upper region with soil moisture which could be justified by presence of undisturbed interior, large sized evergreen trees and litter covered forest floor whereas, the difference in lower region is mainly due to disturbances like, widespread tree cutting, lopping and encroachment (for plantation crop establishment) etc. Although this lower region has some wet evergreen species like Sageraria laurifolia, Hydnocarpus laurifolia, Polyalthia fragrans and Pterospermum diversifolium they are more restricted to adult population with minimal representation in the juvenile group. Non-sacred sampling sites in Sambegadde forest are present in close association with each other in relation to factors like bulk density and soil organic carbon but standing apart in soil moisture conditions. Similarly, tree species also form separate clusters corresponding to soil moisture. Wet evergreen species like Dipterocarpus indicus, Myristica malabarica, Dyospyros saldanha, Cinnamomum macrocarpum, are clustered in the vicinity of soil moisture variable which confirms their inclination towards moist microclimatic conditions of Karikan. These species are more abundant in southern Western Ghats which enjoy more rainy months (usually 7 to 9) than in the central Western Ghats of Uttara Kannada with 5-7 rainy months (Pascal & Ramesh 1997).  

Dry season soil moisture profile for top soil layer (0-20 cm) was analysed as it is one of the important determinant factors for species composition in any area (D’Odorico et al. 2010; Jirka et al. 2007). Moisture level at top soil layer is an indicator of water retention potential of entire soil mass especially at lean season. In clayey soil, the top soil layer (10-30 cm) is considered as soil moisture control section which plays important role in maintenance of above ground biomass (Anonymous 1999). As long as the top soil layer is wet there is a clear indication of water availability in the area. The study shows significantly higher soil moisture profile throughout the pre-monsoon season in Karikan which could be attributed to its dense canopy cover of evergreen trees and rich litter cover on the soil, not tampered with at least in the sacred grove proper. Many studies have established linkages of old growth trees in tropical forests as adapted to several physiological changes like less water usage, decreased stomatal conductance and protection of soil from direct solar radiation which promote water conservation in the nearby localities (Kagawa et al. 2009; Macfarlane et al. 2010; Singh & Mishra 2012; Vertessy et al. 2001).

The typical old growth forest structure (i.e. higher basal area, height, high evergreenness and endemism) and soil moisture profile in Karikan have obviously combined effect on water availability in downstream area. Bangarmakki shows comparatively slower decline in water table in dry months than Sambegadde indicating higher recharging capacity. The landscape with good vegetation cover allows higher infiltration of water and groundwater recharge. This also helps in sustaining water in the streams during lean seasons due to lateral flow from vadoze zone and from aquifer. The partial presence of water in stream bed in May in the Bangarmakki valley indicates stable presence of ground water in the area even in dry summer period, despite having heavy use for horticultural crops. Increment in water table is more drastic in Sambegadde compared to Bangarmakki due to its completely dry stream bed and sharp decline in the ground water table during summer.

Here we find the hydrological significance of a natural sacred site (Karikan) benefitting immensely the farming community in the valley downhill. The predominance of water demanding plantation crops at Bangarmakki indicates year long water availability in the region whereas, Sambegadde shows dominance of rain-fed cultivation mainly of paddy, (64% of total cultivable area) with less plantation crops (29% of total cultivable area).  The availability of water for longer period (lean seasons) is also reflected in water usage pattern. Monthly water utility in dry period (January to May) shows that Bangarmakki has more water availability as reflected in the usage (1862- 136 Klit ha^-1) than Sambegadde (947-37 Klit ha^-1). Similarly, a conservative monetary assessment on farm production has also highlighted the contrast between the areas. Constant availability and higher market prices of horticultural products keeps the Bangarmakki farmers in better economic situation in comparison to Sambegadde farmers. The farmers in Sambegadde, obviously, not backed to that extent by supporting forests like a primeval sacred grove, have to subsist mainly on rain fed rice. They are also required to grapple with fast depleting water table in the dry months throwing challenges for maintenance of cash crops in their gardens.