Ices regulate much of the chemistry during star formation and account for up to 80% of the available oxygen and carbon. In this paper, we use the Spitzer c2d ice survey, complimented with data sets on ices in cloud cores and high-mass protostars, to determine standard ice abundances and to present a coherent picture of the evolution of ices during low- and high-mass star formation. The median ice composition H2O:CO:CO2:CH3OH:NH3:CH4:XCN is 100:29:29:3:5:5:0.3 and 100:13:13:4:5:2:0.6 toward low- and high-mass protostars, respectively, and 100:31:38:4:-:-:- in cloud cores. In the low-mass sample, the ice abundances with respect to H2O of CH4, NH3, and the component of CO2 mixed with H2O typically vary by <25%, indicative of co-formation with H2O. In contrast, some CO and CO2 ice components, XCN and CH3OH vary by factors 2-10 between the lower and upper quartile. The XCN band correlates with CO, consistent with its OCN- identification. The origin(s) of the different levels of ice abundance variations are constrained by comparing ice inventories toward different types of protostars and background stars, through ice mapping, analysis of cloud-to-cloud variations, and ice (anti-)correlations. Based on the analysis, the first ice formation phase is driven by hydrogenation of atoms, which results in a H2O-dominated ice. At later prestellar times, CO freezes out and variations in CO freeze-out levels and the subsequent CO-based chemistry can explain most of the observed ice abundance variations. The last important ice evolution stage is thermal and UV processing around protostars, resulting in CO desorption, ice segregation and formation of complex organic molecules. The distribution of cometary ice abundances are consistent with with the idea that most cometary ices have a protostellar origin.

Icy grain mantles are commonly observed through infrared spectroscopy toward dense clouds, cloud cores, protostellar envelopes and protoplanetary disks. Up to 80% of the available oxygen, carbon and nitrogen are found in such ices; the most common ice constituents - H2O, CO2 and CO - are second in abundance only to H2 in many star forming regions. In addition to being a molecular reservoir, ice chemistry is responsible for much of the chemical evolution from H2O to complex, prebiotic molecules. Combining the existing ISO, Spitzer, VLT and Keck ice data results in a large sample of ice sources (\sime80) that span all stages of star formation and a large range of protostellar luminosities (<0.1-105 L\odot). Here we summarize the different techniques that have been applied to mine this ice data set on information on typical ice compositions in different environments and what this implies about how ices form and evolve during star and planet formation. The focus is on how to maximize the use of empirical constraints from ice observations, followed by the application of information from experiments and models. This strategy is used to identify ice bands and to constrain which ices form early during cloud formation, which form later in the prestellar core and which require protostellar heat and/or UV radiation to form. The utility of statistical tests, survival analysis and ice maps is highlighted; the latter directly reveals that the prestellar ice formation takes place in two phases, associated with H2O and CO ice formation, respectively, and that most protostellar ice variation can be explained by differences in the prestellar CO ice formation stage. Finally, special attention is paid to the difficulty of observing complex ices directly and how gas observations, experiments and models help in constraining this ice chemistry stage.