More efficient and reliable High-Throughput Field Phenotyping (HTFP) approaches are essential for the development of plant breeding and carbon storage studies, as well as the improvement of yield estimation in the food production sector. The use of ground-based platforms in combination with other data sources such as UAVs and satellites addresses constraints related to payload capacity restrictions and reduced below-canopy data collection. This study describes an early approach to the deployment of agile robots for HTFP that aims to estimate height, diameter at breast height (DBH), and volume for forty-three unique trees corresponding to two different species (E. variegata and F. altissima) occupying an urban-park. The data acquisition system consists of an agile robot from Boston Dynamics and a navigation enhancer LiDAR module from the same company. In order to obtain a point cloud using this system, it is necessary to overcome two challenges: a reduced vertical FoV of the LiDAR and limited management of the LiDAR module. A multiway registration approach is implemented to reconstruct a low-density digital twin of the experiment site, which is later georeferenced using points surveyed with a GNSS system. Subsequently, the point cloud is manually segmented using CloudCompare software to obtain individual tree point clouds. Three different algorithms are implemented to obtain height, DBH, and tree volume estimates from the individual point clouds. The results are statistically analyzed by species in order to characterize sources of error. The height estimation method had a Median Percentage Error (MPE) of 1.4% for E. variegata and 1.2% for F. altissima. The DBH estimation had an MPE of 20.1% for E. variegata and 13% for F. altissima. The volume estimation model returned an R2 of 0.86 for E. variegata and 0.98 for F. altissima. Finally, all three feature estimations are mapped into a GEOJson file. These findings, combined with the numerous advantages of using agile robots as mobile platforms in HTFP, enable more efficient and reliable estimation of important parameters such as aboveground biomass and carbon storage sequestration, as well as delivery the potential for improvements in crop monitoring and yield estimation.