Premium
Relationship of kinds of seed dormancy with habitat and life history in the S outhern K alahari flora
Author(s) -
Kos Martijn,
Baskin Carol C.,
Baskin Jerry M.
Publication year - 2012
Publication title -
journal of vegetation science
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.1
H-Index - 115
eISSN - 1654-1103
pISSN - 1100-9233
DOI - 10.1111/j.1654-1103.2012.01415.x
Subject(s) - dormancy , seedling , biology , seed dormancy , stratification (seeds) , soil texture , soil water , multinomial logistic regression , agronomy , water content , germination , ecology , botany , horticulture , mathematics , geology , statistics , geotechnical engineering
Questions Is kind of seed dormancy important in environmental filtering at the local scale? Are there differences among kinds of seed dormancy in life‐history traits like seed mass and adult life span? Location Southern Kalahari ( B otswana, N amibia and S outh A frica). Methods Using mainly literature data, we determined kind of dormancy for each species in the Southern K alahari flora and combined this data set with data on seed mass, life form and soil texture. Soil sand content was used as an environmental gradient because the risks associated with seedling establishment increase with soil sand content and would favour seed dormancy as a bet‐hedging strategy. Also, the rapid descent of water in very sandy soils may exclude species with morphological ( MD ) or morphophysiological ( MPD ) dormancy that require relatively long moist periods for dormancy breaking. Multinomial logistic regression and phylogenetically independent contrasts were applied to determine relationships of kind of dormancy with life‐history variables and soil texture. Results The proportion of species with some kind of seed dormancy was not correlated with soil sand content. However, the proportion of species with MD or MPD decreased with increasing sand content, and proportion of species with some kind of dormancy decreased with increasing seed mass and life span. Life form and seed mass, but not soil sand content, had significant effects in a multinomial regression model predicting the presence of kind of dormancy. With increasing seed mass, the probability of a species having physical dormancy ( PY ) increased relative to the probability of having physiological dormancy ( PD ). With an increase in life span, the probability of a species lacking dormancy increased relative to the probability of having PD . Conclusions We found little evidence that kind of seed dormancy plays a direct role in environmental filtering. Our results suggest that traits related to dormancy do not evolve independently and are part of a syndrome of co‐adapted plant traits.