UC Berkeley

Accurate quantification of ecosystem change is essential for effective environmental management. However, the selection of meaningful indicators of impacts to ecosystems and of benefits from restoration is not standardized. In this dissertation I investigate 1) the applicability of using sediment reduction as an indicator of the cumulative impacts of dams in the Mekong River basin; 2) review and evaluate the meaningfulness of common river restoration evaluation metrics such as macroinvertebrate diversity and richness in habitat heterogeneity projects, and 3) demonstrate the usefulness of prey availability as an indicator of restoration success in riparian restoration projects along the lower Colorado River.

1) The Mekong River, largely undeveloped prior to 1990, is undergoing rapid dam construction. Seven dams are under construction on the mainstem in China and 133 are proposed for the Lower Mekong River and tributaries. The question is what cumulative effect will these dams have on sediment movement in the watershed. There was a lack of data on sediment yields in some portions of the basin so we delineated nine distinct geomorphic regions, for which we estimated sediment yields based on geomorphic characteristics, tectonic history, and the limited sediment transport data available. We then applied the 3W model to calculate cumulative sediment trapping by these dams, accounting for changing trap efficiency over time and multiple dams on a single river system. Under a ‘‘definite future’’ scenario of 38 dams (built or under construction), cumulative sediment reduction to the Delta would be 51 percent. Under full build-out of all planned dams, cumulative sediment trapping will be 96 percent. That is, once in-channel stored sediment is exhausted, only 4% of the predam sediment load would be expected to reach the Delta. We then combined geomorphic assessments of the Mekong channel and delta with the 3W model’s results of sediment trapping to forecast geomorphic change. We expect the biggest changes to occur along alluvial reaches, though stripping of thin sediment deposits in bedrock reaches may also have significant consequences for benthic invertebrates, fishes, and other aquatic organisms dependent on the presence of alluvium in the channel. If all dams are built as proposed, the resulting 96% reduction in sediment supply would have profound consequences on productivity of the river and persistence of the delta landform itself and suggests that strategies to pass sediment through/around dams should be explored to reduce the magnitude and consequences of downstream sediment starvation. In this first case, we use sediment reduction as an indicator of watershed impairment. Though many complexities (e.g. oil, gas, and groundwater withdrawals, routing of sediment through deltas) influence coastal erosion, we found sediment reduction to be a meaningful worldwide indicator. We compiled sediment data from 24 worldwide deltas and results indicate a positive relationship of sediment reductions to deltas resulting in decreased rates of aggradation. In particular, sediment reductions of more than 80% are consistent in almost complete cessation in aggradation rates. The full-build scenario of Mekong dam building would result in 96% reduction in sediment delivery and we would then expect an almost complete cessation in sediment deposition in the delta.

2) In a search for accountability, the effectiveness of many large restoration programs has been evaluated using standard such as acres or length of stream restored per dollar, but this was recognized to be inadequate. Another common restoration metric is based on the common goal of enhancing ecosystems by creating more complex and varied habitats. Although widely implemented, there is little understanding of the success to date of such projects. There is also little agreement on the best approaches and metrics for quantifying success. We reviewed the methods of 26 peer-reviewed evaluation studies and investigated the influence of study design on evaluation results. Of the 26 studies, many did not implement rigorous study designs. For example, only 46% of the studies used quantitative measures of habitat, 62% included only one year of post-project monitoring, 46% used zero or one control (unrestored) sites, and 62% did not include reference (best potential ecological condition) sites. Studies that used more rigorous designs (e.g. sampled more years, measured habitat quantitatively) were more likely to find increased ecosystem diversity and richness in response to heterogeneity enhancement. More fundamentally, all studies used macroinvertebrate diversity and/or richness as the measure of ecological success. We question the logic of assuming that reach-scale diversity or richness is useful as a universal measure of ecosystem integrity. Monitoring and evaluation should first establish hypotheses and conceptual models based on watershed perturbations and set specific milestones towards a sustainable, dynamic, and healthy ecosystem. Restoration targets should be defined based on regional, historical, and analytical reference conditions and by conducting manipulative experiments that can help predict ecosystem responses to restoration actions. It is important to understand if habitat heterogeneity projects are succeeding, but it is not yet possible to draw general conclusions. Metrics to evaluate performance of stream restoration projects need more rigor and should be tied to project specific goals. Generic metrics may yield misleading results.

3) Below Hoover Dam, riparian vegetation along the Colorado River was extensively cleared for agriculture. Thus, large areas of habitat were lost to clearing. Moreover, the functions of the ecosystem were compromised as the connections of the river to its floodplain were severed by levees, flow reduction by dams and diversions, channel incision, and groundwater pumping. Subsequently, native species declined, including the southwestern willow flycatcher (Empidonax traillii extimus) that nests along rivers in dense riparian thickets. The Lower Colorado River Multi Species Conservation Program (MSCP) was established in 2005 to re-create habitat for 26 species including the flycatcher, but the benefits of these restoration sites for target species have not been quantified. Many MSCP projects have involved extensive plantings of willow (Salix exigua, S. gooddingii] and cottonwood [Populus fremontii) on high terraces disconnected from the river by levees. MSCP projects goals are specified as acres of habitat, but to support functioning food webs, riparian ecosystems in arid regions require a subsidy of aquatic insects. We documented prey availability for the southwestern willow flycatcher in constructed habitats as an indicator of their potential to support the species. The number of aquatic insects, proportion of aquatic insects, total number of insects, and number of insect orders all decreased with distance from the river, and the decrease occurred within the first 100 meters from the river. The MSCP cottonwood-willow plantation (more than 500 m from the river) at Cibola National Wildlife Refuge had 86% fewer total insects (p=0.055), 97% fewer aquatic insects (p=.032), and only half as many insect orders (p=0.015) as sites adjacent to the river. In the plantation, only 16% of the insects were aquatic vs. 59% aquatic at the river’s edge (p=.063). Our results suggest that restoration success (and the recovery of southwestern willow flycatcher) may be limited by prey availability and that future riparian plantings should be concentrated along the river or tributary channels. Southwestern willow flycatchers have not been nesting in MSCP plantations. Thus the metric of “acres restored” is inadequate to capture ecosystem function. More meaningful metrics would identify potential limitations in ecosystems (such as prey availability) so that habitat suitability and functionality can be assessed and adaptively managed.

Together, the chapters of this dissertation highlight different approaches and considerations for the quantification of ecosystem impacts and restoration success. The field of ecosystem quantification is still far from adopting universally appropriate indicators of change, but this dissertation seeks to highlight problems in current approaches, and to demonstrate useful models and approaches.