Leaf mass loss in wetland graminoids during senescence Corina Vernescu, Jess Coulas and Peter Ryser Vernescu, C., Coulas, J. and Ryser, P. 2005. Leaf mass loss in wetland graminoids during senescence.  / Oikos 109: 187 /195. Mass loss of senescing leaves is an important part of plant biomass turnover and has consequences for assessment of ecosystem productivity, ecosystem nutrient use efficiency, and plant nutrient resorption efficiency. Data, however, on mass loss are scarce, and often based on leaf area as the reference base. This leads to an underestimation of the mass loss, as leaf area itself shrinks during senescence. Furthermore, the few existing studies have almost exclusively used woody species. The purpose of the present study was twofold: i) to assess leaf mass loss during senescence in herbaceous species, with the example of five wetland graminoids and, ii) to compare two different methods of mass loss assessment (two species). Assuming that leaf length does not change during senescence, we assessed leaf mass per leaf length prior to and after senescence. We also estimated pre-senescence leaf mass nondestructively based on leaf length, width and thickness. For Typha latifolia and Carex stricta , two species with graminoid type leaves but contrasting leaf structure, both methods delivered almost identical results. After the first assessment of leaf mass on July 7th, T. latifolia leaf mass initially increased by 13%, and then decreased to be 12% below the original mass after senescence. C. stricta leaf mass remained stable until senescence, but decreased then by 33%. In a second experiment, the mass of 100 mm pieces of leaves was measured before and after senescence. Calamagrostis canadensis , Carex rostrata and C. stricta lost 23 /57% of their leaf mass during senescence, whereas Glyceria canadensis did not show any mass loss. We conclude that mass loss of senescing leaves of herbaceous plants can be considerable and should not be neglected in studies of productivity, nutrient use efficiency or nutrient resorption. For species with no shrinking leaf length during senescence, mass loss can be measured with leaf length as the base whereas for others, pre-senescent mass can be estimated on the basis of leaf dimensions. C. Vernescu, J. Coulas, and P. Ryser, Dept of Biology, Laurentian Univ., Ramsey Lake Road, Sudbury, ON, Canada P3E 2C6 (pryser@laurentian.ca). Resorption of mineral nutrients from senescing leaves is an important aspect of a plant’s nutrient balance (Aerts 1996, Eckstein et al. 1999). In contrast, resorption of non-nutrient-containing compounds from senescing leaves is usually thought to be insignificant (Chapin et al. 1990). Hence, litter production is often set equal to ecosystem productivity, especially in studies involving tree or ecosystem nutrient use efficiency (Vitousek 1982, Knops et al. 1997, Smith et al. 1998, Cordell et al. 2001, Norby et al. 2002, Yasumura et al. 2002, Mediavilla and Escudero 2003). Difference in pre- and post-senescence nutrient concentrations in litter is often regarded as a direct measure of nutrient resorption efficiency (Gerdol et al. 2000, Hawkins and Polglase 2000, Ochieng and Erftemeijer 2002, Distel et al. 2003), and litter mass has even been used to determine the leaf area index of a forest (Norby et al. 2001). Nevertheless, leaf mass loss during senescence has been reported to be as much as 36% (Oland 1963, Woodwell 1974, Jonasson 1989, Chapin and Moilanen 1991, Karlsson 1994, Tre ´molie `res et al. 1999). Such figures should not be ignored in any studies investigating biomass or litter production of Accepted 10 September 2004 Copyright # OIKOS 2005 ISSN 0030-1299 OIKOS 109: 187 /195, 2005 OIKOS 109:1 (2005) 187