Filtered nutrient forms gave a smaller point decline than total N:P in two Gulf river load estimates from 1980-1996 to 2009-2025

We compared total nitrogen to total phosphorus with filtered nitrate plus nitrite to orthophosphate in USGS annual load estimates for the Mississippi and Atchafalaya rivers. Under stationary flow normalization, the combined total molar ratio fell by 27.75% from 1980-1996 to 2009-2025. The selected filtered ratio fell by 14.64%. A rise in the nitrate-plus-nitrite share of nitrogen and a fall in the orthophosphate share of phosphorus offset 51.30% of the total log decline in exact load accounting. Generalized flow normalization, annual loads, separate river tests and alternative late periods retained the smaller filtered point decline. However, a box formed from source annual marginal limits included zero for the filtered ratio change and for the offset. Thus source uncertainty does not resolve these signs. The result is a nutrient-form benchmark, not a causal explanation or a test of coastal nutrient limitation. Separate fitted loads, annual time resolution and missing joint source errors limit inference. The package includes the CC0 source, hashes, code, all comparisons, a separate audit and a figure.

nutrient forms · Mississippi River · Atchafalaya River · load ratios · flow normalization · uncertainty