While it is generally accepted that primary production in temperate forest ecosystems tends to be N-limited, some have suggested that co-limitation between N and other nutrients (P or Ca) is more likely.  Precipitation pH, soil mineralogy, soil depth, forest age, and land-use history vary dramatically across the northern hardwood forest region of the northeastern US and southern Canada.  In addition, deposition rates of N, base cations, and P have all changed over the past century.  The forests of the northeastern US are well-studied from both silvicultural and ecological perspectives.  Here, I ask:  How strong is the evidence for limitation of primary production by N or other nutrients in the northern hardwood region, and to what degree does nutrient limitation vary among sites?  To answer this, I conducted a meta-analysis using data from all 21 published forest fertilization experiments (1939-2008) on glaciated soils across the northern hardwood region.

Results/Conclusions

The evidence is strongest for N limitation (suggested by 15 of 17 studies; mean effect 50±14%), but this does not rule out co-limitation, which most studies are not designed to detect.  Forest productivity also tends to increase with additions of Ca and P (mean effect = 44% and 16%, respectively), though the supporting evidence is weak so far.  The designs of most of these studies do not allow strong conclusions with regard to the nutrient(s) limiting net primary production at the scale of stands or individual trees.  Chief among these is the lack of full factorial design in terms of added nutrients.  The forms in which these nutrients are added can also be problematic.  P is typically added as triple superphosphate, which contains almost as much Ca as P.  Ca and N are typically added in forms that may affect soil pH and therefore change the availability of other nutrients.  Future studies should add nutrients independently and in factorial combination, and in forms that minimize changes in soil pH, which should be investigated separately.