In this paper the authors compiled an impressive telemetry dataset from 7 brown bear populations across North America to test 1) whether brown bear ‘track the green wave’ but selecting habitat in spring based on spatially ordered vegetation quality vs quantity, and 2) whether variation in resource tracking is explained by landscape-level characteristics (e.g. rate of vegetation green-up) and individual- or population-level traits (e.g. diet, sex). They found that many bears do track waves of spring green-up, but the extent of this is modified by several landscape characteristics as well as the amount of protein in the bear’s diet (reducing the need to rely on vegetative forage). This study is novel is that it is one of the first to demonstrate the extent and context of resource tracking based on intermediate forage biomass (high vegetation quality) in an omnivore. The paper was extremely clear, well-motivated, and robust methodologically. I believe this will make a valuable contribution to the literature, and I only have minor comments.

There seems to be some moderate conflation in the writing between resource tracking, the Forage Maturation Hypothesis (FMH), and the Green Wave Hypothesis (GWH). The FMH predicts that herbivores will select vegetation at intermediate biomass to balance energy intake rate and digestibility. In L112-113, the authors state they “test whether brown bears select forage resources at intermediate phenological states (supporting the GWH)”, however this is actually supporting the FMH and would not necessarily support the GWH without further information. L154 makes a similar conflation: “To test if bears were tracking the green wave versus maximum biomass”, however tracking of intermediate biomass does not necessarily mean green wave tracking. The GWH is a specific case of resource tracking (movement to track phenological variation in resources over space, which can be in any spatial configuration) in which that phenological variation is spatially ordered across the landscape in a wave-like pattern (e.g. Aikens et al. 2017). However not all resource tracking is green wave tracking, and some of the examples given in the introduction are of resource tracking but not green wave tracking (e.g. Bowersock et al. 2021). The authors should clarify these distinctions further in the introduction.

Given the large range of fix rates of your study individuals, did you do any sensitivity checks to ensure fix rates do not affect results? For example you could subsample the hourly fixes where available and compare selection coefficients for a given population at various fix rates.

In the discussion, the two metrics – use & selection – are frequently treated indistinguishably as “tracking behavior”, though they represent different processes and yielded different results. It would be more transparent to be clear about which process is being discussed (a la L367), and provide some explanation on what the different metrics represent ecologically, why they might have differed, and the ecological significance/interpretation of those differences.

The abstract suggests that the findings from this paper could guide conservation and habitat restoration efforts, but there is no treatment of this in the discussion. A paragraph on the conservation implications of this work would be valuable.