Iceberg calving is a major component of glacier mass ablation that is not well understood due to a lack of detailed temporal and spatial observations. For better understanding, it is critical to examine processes occurring on the time scale of calving processes, sub-daily to sub-hourly. Current satellites are not able to observe the same location at time scales small enough to measure sub-daily phenomena. This research aims to increase the temporal resolution of ice speed and elevation measurements during the calving season to allow for analysis of short-term variations that are otherwise unobserved. We measure glacier speed and surface elevation at 3-minute intervals using a portable radar interferometer at three marine-terminating glaciers in West Greenland over two summer field campaigns. We detect diurnal variations in glacier speed caused by tidal height changes that propagate far inland, the effect of which varies by glacier but are consistent with simple models where basal stress is tidally modulated. We find no speed up from ice shedding off the calving face or the detachment of floating ice blocks, as expected. We detect a 30% speedup within a few hundred meters of the ice front that persists for days when calving removes full thickness grounded ice blocks. Within one ice thickness from the calving front, we detect strain rates 2 to 3 times larger than observable from satellite data, which has implications for studying iceberg calving as a fracturing process, in particular to select an appropriate value of the threshold tensile stress necessary for ice cliff failure.

Organismal complexity is often reduced to individual systems, but organisms function as an integrated whole and reductionist studies cannot address the constraints imposed by systems working together to perform a function. In this dissertation, I use integration between locomotor and feeding performance during prey capture in fishes as a model system for understanding complex behaviors and their ecological relevance. First, I demonstrate the empirical utility of integration for describing emergent differences between species. I utilize two species of Pacific marine sculpins capturing live amphipod prey, and confirmed that species were similar in feeding behaviors but different in their use of locomotion during prey capture. This resulted in differences in integration that reflect ecological divergence that would not be apparent in feeding behaviors alone. Second, I demonstrate that differential capture success is due to differences in predator accuracy. I utilize centrarchid sunfishes to develop a non-invasive model of suction volume and accuracy and apply this model to 3D feeding kinematics of three predators capturing two prey types. Not only did accuracy vary across species, but so did the ability to modulate the shape of the ingested volume of water, leading to a direct effect on predator capture success. Finally, I expand the techniques for quantifying behavioral integration to multivariate space and assess the causes and consequences of integration using a single centrarchid predator capturing two prey types. Partial least squares correlations describe multivariate integration and demonstrate that predators rely on performance variables differentially for each prey type. These differences are then reflected in patterns of integration and predator accuracy across prey types. This dissertation advances our understanding of how organismal integration and complexity act to drive performance and ecology. I demonstrate that performance integration is real and can be quantified and establish the empirical and ecological relevance of performance integration. Integration and organismal complexity may be one of the next scientific frontiers, and this dissertation provides a first step in exploring this new direction.