The coastal red mangrove, Rhizophora mangle L., was introduced to the Hawaiian Islands from Florida 100 yr ago and has spread to cover many shallow intertidal shorelines that once were unvegetated mudflats. We used a field survey approach to test whether mangroves at the land-ocean interface could indicate watershed inputs, especially whether measurements of leaf chemistry could identify coasts with high nutrient inputs and high mangrove productivities. During 2001-2002, we sampled mangroves on dry leeward coasts of southern Moloka'i and O'ahu for 14 leaf variables including stable carbon and nitrogen isotopes (delta13C, delta15N), macronutrients (C, N, P), trace elements (B, Mn, Fe, Cu, Zn), and cations (Na, Mg, K, Ca). A new modeling approach using leaf Na, N, P, and delta13C indicated two times higher productivity for mangroves in urban versus rural settings, with rural mangroves more limited by low N and P nutrients and high-nutrient urban mangroves more limited by freshwater inputs and salt stress. Leaf chemistry also helped identify other aspects of mangrove dynamics: especially leaf delta15N values helped identify groundwater N inputs, and a combination of strongly correlated variables (C, N, P, B, Cu, Mg, K, Ca) tracked the mangrove growth response to nutrient loading. Overall, the chemical marker approach is an efficient way to survey watershed forcing of mangrove forest dynamics.