Ecohydrological impacts of woody-plant encroachment: seasonal patterns of water and carbon dioxide exchange within a semiarid riparian environment

Across many dryland regions, historically grass‐dominated ecosystems have been encroached upon by woody‐plant species. In this paper, we compare ecosystem water and carbon dioxide (CO₂) fluxes over a grassland, a grassland–shrubland mosaic, and a fully developed woodland to evaluate potential consequences of woody‐plant encroachment on important ecosystem processes. All three sites were located in the riparian corridor of a river in the southwest US. As such, plants in these ecosystems may have access to moisture at the capillary fringe of the near‐surface water table. Using fluxes measured by eddy covariance in 2003 we found that ecosystem evapotranspiration (ET) and net ecosystem exchange of carbon dioxide (NEE) increased with increasing woody‐plant dominance. Growing season ET totals were 407, 450, and 639 mm in the grassland, shrubland, and woodland, respectively, and in excess of precipitation by 227, 265, and 473 mm. This excess was derived from groundwater, especially during the extremely dry premonsoon period when this was the only source of moisture available to plants. Access to groundwater by the deep‐rooted woody plants apparently decouples ecosystem ET from gross ecosystem production (GEP) with respect to precipitation. Compared with grasses, the woody plants were better able to use the stable groundwater source and had an increased net CO₂ gain during the dry periods. This enhanced plant activity resulted in substantial accumulation of leaf litter on the soil surface that, during rainy periods, may lead to high microbial respiration rates that offset these photosynthetic fluxes. March–December (primary growing season) totals of NEE were ?63, ?212, and ?233 g C m^?2 in the grassland, shrubland, and woodland, respectively. Thus, there was a greater disparity between ecosystem water use and the strength of the CO₂ sink as woody plants increased across the encroachment gradient. Despite a higher density of woody plants and a greater plant productivity in the woodland than in the shrubland, the woodland produced a larger respiration response to rainfall that largely offset its higher photosynthetic potential. These data suggest that the capacity for woody plants to exploit water resources in riparian areas results in enhanced carbon sequestration at the expense of increased groundwater use under current climate conditions, but the potential does not scale specifically as a function of woody‐plant abundance. These results highlight the important roles of water sources and ecosystem structure on the control of water and carbon balances in dryland areas.     

Ecological monitoring to support Water Sharing Plan evaluation and protect wetlands of inland New South Wales,Australia

Government‐funded flow response monitoring and modelling programmes (flow science) provided by the New South Wales Office of Water (NOW) have supported water resource management since 1997. Flow science has a core technical component defined by hypothesis‐driven long‐term monitoring and analysis, but it also represents many activities that support committees involved in environmental flow management. This is done through collaborations and contracting and has fostered considerable research and analysis into flow ecology, including modelling for the recent Murray–Darling Basin Plan. We describe the performance of environmental flows against legislated wetland objectives to improve wetland function and diversity using flow science. On‐ground monitoring at wetland sites has largely ceased but the flow science done so far indicates that the environmental flow rules written into Water Sharing Plans improve wetland diversity and function. Determination of the long‐term flow needs of NSW wetlands, including how well current Water Sharing Plans aid the delivery of environmental flows, requires finding the means to build on current flow science knowledge from across Australia.     

Water stress vulnerability of four Banksia species in contrasting ecohydrological habitats on the Gnangara Mound, Western Australia

This study investigated the interspecific differences in vulnerability to xylem embolism of four phreatophytes – two facultative phreatophytes ( Banksia attenuata and B. menziesii ) and two obligate phreatophytes ( B. ilicifolia and B. littoralis ). Species differences at the same position along an ecohydrological gradient on the Gnangara Groundwater Mound, Western Australia were determined in addition to intraspecific differences to water stress between populations in contrasting ecohydrological habitats. Stem- and leaf-specific hydraulic conductivity, as well as Huber values (ratio of stem to leaf area), were also determined to support these findings. We found that where water is readily accessible, there were no interspecific differences in vulnerability to water stress. In contrast both facultative phreatophyte species were more resistant to xylem embolism at the more xeric dune crest site than at the wetter bottom slope site. B. ilicifolia did not differ in vulnerability to embolism, supporting its classification as an obligate phreatophyte. Other measured hydraulic traits ( K _S, K _L and Huber value) showed no adaptive responses, although there was a tendency for plants at the wetter site to have higher K _S and K _L. This study highlights the influence site hydrological attributes can have on plant hydraulic architecture across species and environmental gradients.     

Ecohydrology of a seasonal wetland in the Rift Valley: ecological characterization of Lake Solai

The following research describes through an ecohydrological approach, the first assessment of the ecology of Lake Solai, with a particular emphasis on the vegetation. Lake Solai is located 50 km north of Nakuru in the Rift Valley in Kenya at E36°80'–36°84' to N00°05'–00°08'. It is a shallow lake that follows a very peculiar seasonal water regime, and that faces conflicts between agriculture and conservation water users. In the upper catchment, an overview of the agricultural practices was implemented and river water uses were identified to assess river flows. Crops/grassland and woodland/shrubland were the major land uses, covering c. 65% of the catchment. Closer to the lake, vegetation samples were collected around the lake together with samples of environmental factors such as soil and water quality. Thirteen vegetation communities were identified within four main zonations: forest, grassland, river inlet and rocky outcrop. These communities showed abundance, distribution and diversity determined mostly by the human pressures, the flooding periods and the salinity. Cynodon , Cyperus and Sporobolus genera were the most abundant.     

Integrating Remote Sensing and Ground Methods to Estimate Evapotranspiration

Evapotranspiraton (ET) is the second largest term in the terrestrial water budget after precipitation, and ET is expected to increase with global warming. ET studies are relevant to the plant sciences because over 80% of terrestrial ET is due to transpiration by plants. Remote sensing is the only feasible means for projecting ET over large landscape units. In the past decade or so, new ground and remote sensing tools have dramatically increased our ability to measure ET at the plot scale and to scale it over larger regions. Moisture flux towers and micrometeorological stations have been deployed in numerous natural and agricultural biomes and provide continuous measurements of actual ET or potential ET with an accuracy or uncertainty of 10–30%. These measurements can be scaled to larger landscape units using remotely-sensed vegetation indices (VIs), Land Surface Temperature (LST), and other satellite data. Two types of methods have been developed. Empirical methods use time-series VIs and micrometeorological data to project ET measured on the ground to larger landscape units. Physically-based methods use remote sensing data to determine the components of the surface energy balance, including latent heat flux, which determines ET. Errors in predicting ET by both types of methods are within the error bounds of the flux towers by which they are calibrated or validated. However, the error bounds need to be reduced to 10% or less for applications that require precise wide-area ET estimates. The high fidelity between ET and VIs over agricultural fields and natural ecosystems where precise ground estimates of ET are available suggests that this might be an achievable goal if ground methods for measuring ET continue to improve.     

Restoration of an Ecosystem Function to Revegetation Communities: The Role of Invertebrate Macropores in Enhancing Soil Water Infiltration

Guidelines for revegetation in agricultural landscapes may not address restoration of ecosystem functions because management is focused on the replanting stage, although certain functions are delivered by organisms that colonize revegetation months or years later. We investigated the ecosystem function of water infiltration to tree root zones and channels, delivered by invertebrates that form soil macropores. We measured macropore density and infiltration rates at revegetation sites established on retired grazing land, in relation to site age, tree species composition, and geographical location, compared with adjacent matched pastures. Revegetated sites had significantly more macropores than pastures, and revegetation sites aged 11–20 years had more macropores than sites aged 3–5 and 6–10 years. Tree species had a marginal effect, with more macropores in sites with Acacia spp. and Eucalyptus spp. than those with Eucalyptus spp. only. Besides ants, the main groups of soil burrowers were mygalomorph and lycosid spiders and also ground‐nesting native bees. Infiltration rates in revegetation sites aged 11–20 years were double those of pastures and of 3–5 and 6–10 year sites. This is the first study to quantify the rate of recovery of an invertebrate‐driven soil hydrological ecosystem function following revegetation.     

Effect of rainfall interannual variability on the biomass and soil water distribution in a semiarid shrub community

The dynamics of biomass and soil moisture in semiarid land is driven by both the current rainfall and the ecosystem memory. Based on a meta-analysis of existing experiments, an ecosystem model was used to calculate the effect of the rainfall interannual variability on the pattern of biomass and soil moisture in a shrub community. It was found that rainfall interannual variability enabled shrubs to be more competitive than grasses, and to maintain the dominant role over a longer time. The rainfall interannual variability resulted in complex soil moisture dynamics. The soil water recharge in wet years alternated with discharge in drought years.     

Land-use change and water losses: the case of grassland afforestation across a soil textural gradient in central Argentina

Vegetation changes, particularly those involving transitions between tree‐ and grass‐dominated covers, often modify evaporative water losses as a result of plant‐mediated shifts in moisture access and demand. Massive afforestation of native grasslands, particularly important in the Southern Hemisphere, may have strong yet poorly quantified effects on the hydrological cycle. We explored water use patterns in Eucalyptus grandis plantations and the native humid grasslands that they replace in Central Argentina. In order to uncover the interactive effects that land cover type, soil texture and climate variability may have on evaporative water losses and water use efficiency, we estimated daily evapotranspiration (ET) in 117 tree plantations and grasslands plots across a soil textural gradient (clay‐textured Vertisols to sandy‐textured Entisols) using radiometric information from seven Landsat scenes, existing timber productions records, and ¹³C measurements in tree stems. Tree plantations had cooler surface temperatures (?5°C on average) and evaporated more water (+80% on average) than grasslands at all times and across all sites. Absolute ET differences between grasslands and plantations ranged from ～0.6 to 2 mm day^?1 and annual up‐scaling suggested values of ～630 and ～1150 mm yr^?1 for each vegetation type, respectively. The temporal variability of ET was significantly lower in plantations compared with grasslands (coefficient of variation 36% vs. 49%). Daily ET increased as the water balance became more positive (accumulated balance for previous 18 days) with a saturation response in grassland vs. a continuous linear increase in plantations, suggesting lower ecophysiological limits to water loss in tree canopies compared with the native vegetation. Plantation ET was more strongly affected by soil texture than grassland ET and peaked in coarse textured sites followed by medium and fine textured sites. Timber productivity as well as ¹³C concentration in stems peaked in medium textured sites, indicating lower water use efficiency on extreme textures and suggesting that water limitation was not responsible for productivity declines towards finer and coarser soils. Our study highlighted the key role that vegetation type plays on evapotranspiration and, therefore, in the hydrological cycle. Considering that tree plantations may continue their expansion over grasslands, problematic changes in water management and, perhaps, in local climate can develop from the higher evaporative water losses of tree plantations.