在样地面积固定条件下，采用样方设计和样带设计，探讨了样方数目与面积之间的关系，以及样方数目与面积的变化对种间联结的影响。结果表明：正负关联比、平均物种出现率、多物种间X2值、物种对间X2值与显著率以及物种间联结系数的变化范围均可作为确定样方数目与面积的参数。在实际分析中应当综合考虑这些参数，进而选择最佳样方数目及面积进行种间联结分析。在样地总面积均为1 200 m2条件下，样带设计在各项指标上规律性比较明显，样方数为50，面积为25 m2最为合适；样方设计的各项参数稍差一些，需要进一步增加样方总面积才能得到更合适的样方数。分别按选定的最适样方数进行物种对联结性分析，对比分析结果显示样带设计对联结性和关联程度的检测效果好于样方设计，它的研究结果更与实际情况相符合。因此，在总样地面积有限的情况下，样带设计是进行种间联结分析的较好选择。
Increasing quadrat area or quadrat number for the study of interspecific associations in plant communities aims at adding overall sample area. The ovell sample area that was chosen by many scholars on studying interspecific association were between 3 000 m2 and 40 000 m2, and the quadrat areas were 25 m2、100 m2 and 400 m2. The overall sample areas were limited in some cases (e.g endangered species or monodominant communities) because the distribution areas are often limited. Therefore overall sample area for studying interspecific association could not be determined arbitrarily. Under condition of fixed overall sample areas, the sampling strategy needs to be explored.The typical sampling area of Quercus liaotungensis forest in Dongling Mountain in Beijing was chosen. The transect that its overall area was 1 250 m2 （5 m ×250 m） was designed. Three quadrats that its overall area were 1 200 m2 (3 ×20 m×20 m) was chosen for the study. The transect was segmented equably according to long border, the area summation of all quadrats was equal to that of the overall transect. Because three quadrats were discontinuous, every one was divided according to definite border or area, and the area summation of all divided quadrats was equal to the overall area of three quadrats. The sample area was divided into some grids. The coordinate and quantity of every individual was measured in all grids.Under conditions of given sample area, some interrelations of quadrat number and area and affects of interspecific association were discussed based on quadrat and transect design. The results showed that positive and negative interspecific association ratio, species average appearance ratio, total multi-species χ 2-test, χ2-test and significant ratio of species pairs and variation range of interspeific association coefficient could be adopted and applied synthetically to choose optimal quadrat number and area. Under condition of 1 200 m2 sample area, regularity of those parameters was showed obviously on transect design for sampling technique of Quercus liaotungensis forest, and the choice of 50 quadrats and 25 m2 for each quadret were optimal quadrat number and area for analysis of interspecific association in this area. The quadrat design was revealed some shortages, and total sample area must be added for gained optimal quadrat number and area. Based on study results of interspecific association analysis according to optimal quadrat number, effects of transect design were better then that of quadrat design. Therefore, under condition of limited total sample area, transect design was better choice for analysis of interspecific association.