Arbuscular mycorrhizal symbiosis alters stomatal conductance of host plants more under drought than under amply watered conditions: a meta-analysis
Augé, R. M., A. M. Saxton, and H. D. Toler.  2014.  Mycorrhiza, 25: 13-24.

Stomata regulate rates of carbon assimilation and water loss. Arbuscular mycorrhizal (AM) symbioses often modify stomatal behavior and therefore play pivotal roles in plant productivity. The size of the AM effect on stomatal conductance to water vapor (gs) has varied widely, has not always been apparent, and is unpredictable. We conducted a meta-analysis of 460 studies to determine the size of the AM effect under ample watering and drought and to examine how experimental conditions have influenced the AM effect. Across all host and symbiont combinations under all soil moisture conditions, AM plants have shown 24 % higher gs than nonmycorrhizal (NM) controls. The promotion of gs has been over twice as great during moderate drought than under amply watered conditions. The AM influence on gs has been even more pronounced under severe drought, with over four times the promotion observed with ample water. Members of the Claroideoglomeraceae, Glomeraceae, and other AM families stimulated gs by about the same average amount. Colonization by native AM fungi has produced the largest promotion. Among single-AM symbionts, Glomus deserticola, Claroideoglomus etunicatum, and Funneliformis mosseae have had the largest average effects on gs across studies. Dicotyledonous hosts, especially legumes, have been slightly more responsive to AM symbiosis than monocotyledonous hosts, and C3 plants have shown over twice the AM-induced promotion of C4 plants. The extent of root colonization is important, with heavily colonized plants showing ×10 the gs promotion of lightly colonized plants. AM promotion of gs has been larger in growth chambers and in the field than in greenhouse studies, almost ×3 as large when plants were grown under high light than low light, and ×2.5 as large in purely mineral soils than in soils having an organic component. When AM plants have been compared with NM controls given NM pot culture, they have shown only half the promotion of gs as NM plants not given anything at inoculation to control for associated soil organisms. The AM effect has been much greater when AM plants were larger or had more phosphorus than NM controls. These findings should assist in further investigations of predictions and mechanisms of the AM influence on host gs.