Abstract: [Objective] We constructed empirical models for leaf area(LA) of eight broadleaf species in Xiaoxing'an Mountains in order to provide technical support for rapid and accurate determination of LA and their dynamics of broadleaf plants with different leaf shapes.[Method] We took eight broadleaf species in a mixed broadleaved-Korean pine (Pinus koraiensis) forest in the Xiaoxing'an Mountains, including Betula platyphylla, Tilia amurensis, Acer mono, Acer tegmentosum, Betula costata, Fraxinus mandschurica, Ulmus japonica and U. laciniata, to measure their leaf length (L), width (W) and LA in the leaf growth periods (late May), stable period (middle July) and early leaf-fall period (early September). Based on the median of the ratio of L to W, the eight species was divided into two groups. A general LA empirical model for predicting different leaf growth periods of different species was constructed, and the prediction accuracy (FC) of the empirical model was evaluated. Finally, we tested the applicability of empirical models for predicting LA of other broadleaf plants by taking two shrubs as examples.[Result] The optimal model types for predicting LA of eight broadleaf species in different leaf growth periods were all power functions. For species group with ratio of L to W less than 1.5 (median), the empirical model for predicting LA was LA=0.618×L^1.197W^0.806 in May and July and FC was 81%-96%, and LA=0.581×L^1.332W^0.671 and FC was 83%-94% in September. For species group with ratio of L to W more than 1.5, the empirical model for predicting LA was LA=0.627×L^0.975W^1.063 in Mayand FC was 91%-95%, and LA=0.705×L^0.876W^1.146 and FC was 92%-96% in July,and LA=0.607×L^0.970W^1.076 and FC was 92%-95% in September. The FC of these models for predicting Corylus mandshurica (ratio of L to W lower than 1.5) and Syringa reticulata var. Amurensis (ratio of L to W larger than 1.5) in different leaf growth periods ranged from 88% to 94%.[Conclusion] For different leaf growth periods, different empirical models should be constructed to predict LA. It is feasible to classify a variety of broadleaf species according to the ratio of L to W, and the empirical models constructed based on trees can also effectively predict the LA of shrubs. The result can provide technical support for quick and efficient LA prediction of other broadleaf plants.

Key words: leaf area, length, width, the ratio of length to width, empirical model