The graph shows zinc in solution at 3 locations over time. I suspect the late spring / early summer spike would be due to low water levels / drought conditions. What water there was spent longer in contact with the ore minerals before running past the test stations. In normal, wetter, conditions the water has less time to dissolve metals, has less carbonic acid content from soils and CO2 solution, and is more diluted.
Do you have a rainfall graph, or a flow graph, for the same location(s)?
My guess for now until something better comes along... 😉