Kin Recognition Alters Root and Whole Plant Growth of Split-root Cycas edentata Seedlings

Split-root techniques were used to determine if Cycas edentata seedling roots possessed identity recognition capabilities. One compartment of the split-root containers contained a competing half-sibling seedling, and the second compartment contained one of four treatments. When the second compartment contained a half-sibling or C. edentata seedling from the same habitat, root growth did not differ between compartments. When the second compartment contained a C. edentata seedling from a habitat 130 km away or a C. nitida seedling, root dry weight and root length were increased above that of compartments with half-siblings. Whole plant growth also increased in seedlings that experienced competition with the conspecific plant from a distant habitat or from a different species. The results validated the ability of Cycas edentata roots to recognize and then use a plastic response to neighbor root identity. In horticultural settings, planting cycad plants where neighbors are close relatives may reduce plant growth, and planting to ensure neighbors are not close relatives may increase plant growth. The ability of plants to distinguish self roots and kin roots from non-kin roots is shown in the growing body of research on plant identity recognition. In many species studied to date, roots respond differently to roots of the same individual, roots of a close relative, or roots of non-related competitors (Falik et al., 2003; Gruntman and Novoplansky, 2004; Holzapfel and Alpert, 2003; Mahall and Callaway, 1992, 1996; Maina et al., 2002). These responses are studied in the context of competition reduction among different parts of the same plant or those of a close relative. Kin selection theory is partly founded in the conception that individuals increase their collective fitness by behaving in a manner that increases the fitness of related individuals (Hamilton, 1964a, 1964b). This is realized by reducing investment in competition, which can increase the individual’s fitness because fewer resources are consumed for competing with neighbors and can similarly increase the neighbor’s fitness through altruism (Axelrod and Hamilton, 1981). The most prevalent response variables in this body of research are various root growth traits. Other responses to belowground neighbor identity reported to date include plant biochemistry (Broz et al., 2010) and pollen quality (Lankinen, 2008). Identity recognition capabilities of cycads have not been studied to date. A greater understanding of this phenomenon may be useful for planning the layout of garden and conservation plantings. The purpose of this study was to determine the presence of kin recognition capabilities in roots of a representative cycad species. Cycas edentata is a widespread cycad species that colonizes islands throughout its natural range (Lindström et al., 2008). It is one of two coastal species in the Philippines, and I selected this species because seeds from distinct islands could be easily secured to provide con-specific individuals. I predicted that root growth would be minimized when seedlings were forced to interact with close relatives and would be maximized when forced to interact with non-relatives. Materials and Methods Cycas edentata seeds were purchased from private property owners in the Philippine islands of Mindoro and Marinduque in Nov. 2010. Cycas nitida seeds were purchased from a private property owner in Samar Island in Feb. 2011. All of the trees were growing naturally within minimally managed copra farms. Four groups of seeds were used for this study: 1) the test group of seedlings originated from one mother tree in the Mindoro habitat. A portion of these was used to prepare the split root subjects, and a portion was used to generate competitor seedlings representing half-sibling competitors (half-sibs hereinafter, pollen source was unknown for these open-pollinated seeds); 2) a second mother tree in the same Mindoro habitat and located 53 m to the north represented competitors from a different mother but the same habitat (same habitat); 3) one mother tree in the Marinduque habitat was located 130 km to the northeast and represented competitors of the same species but a distant habitat (different habitat); and 4) one C. nitida mother tree in the Samar habitat was located 350 km to the east and represented competitors from a con-generic population (different species). The sarcotesta was removed and the cleaned seeds were stored until 25 Aug. 2011, when all seeds were planted in a bed of perlite. The germinated seedlings were removed from the perlite on 3 Oct. 2011. At this stage, a robust taproot from 110 to 120 mm in length had developed but no leaf growth had ensued (Fig. 1A). The split-root seedlings were crafted by starting with a transverse cut to standardize taproot length to 50 mm and then by splitting the entire taproot into two halves with a radial longitudinal cut (Fig. 1B). All cuts were made with a razor blade. The split-root containers were created by adjoining two square 0.625-L pots. The split roots were positioned onto the ridge that separated each of the two containers such that each container received half of the split root. The containers were filled with washed river sand. The remainder of the seedlings were planted singly in 1.8-L pots in the same medium and held until the beginning of the competition treatment. The split-root seedlings were grown under 75% shade for 1 month and then the shadecloth was removed. The nursery bench received 4 h of morning sun and then filtered shade from a Sandoricum koetjape Merr. tree for the remainder of the photoperiod. The sand was watered daily. Root competition. Each of the split-root seedlings was bare-rooted to visually inspect root development on 13 June 2012. Seedlings with heterogeneous fine root development in each compartment were retired from the study, and the seedlings with homogeneous amounts of fine roots in the two halves (estimated visually) were re-planted into larger split-root containers with 2.6-L capacity in each half. Seedlings that had been grown singly were bare-rooted and planted as competitors. Every replication received a half-sib competitor in one of the two compartments. The second compartment received one of four seedling types: 1) a half-sib seedling as the control; 2) a C. edentata seedling from the same habitat; 3) a C. edentata seedling from the distant habitat; or 4) a C. nitida seedling (Fig. 1C). All plants were provisioned with 75% shade for 1 month and then were grown under the ambient light conditions described previously. A completely randomized design was used with four treatments and nine replications, and a border was provided by C. edentata seedlings in single pots. To maximize root competition, no fertilizer was applied for the duration of the experiment. The river sand medium was impoverished as indicated by soil analyses that revealed total nitrogen was 4.4 ± 0.6 mg·g, available phosphorus was 5.01 ± 0.49 mg·kg, and exchangeable potassium was 20.11 ± 1.02 mg·kg (mean ± SE, n = 3). The plants were irrigated every 2 d in the absence of rain and maintained until 16 May 2013. Ending measurements. All three plants in each replication were bare-rooted. The two Received for publication 05 Aug. 2013. Accepted for publication 23 Aug. 2013. To whom reprint requests should be addressed; e-mail [email protected]. 1266 HORTSCIENCE VOL. 48(10) OCTOBER 2013 competitor seedlings were not used for measurements and were removed from the experiment. Measurements on the split-root seedling began by counting the number of leaves and number of leaflets for each leaf. Stem height and diameter were measured. Fine roots under 2 mm in diameter were removed from coarse roots above 2 mm in diameter, and their length was determined using the line intersect method (Smit, 2000; Tennant, 1975). Length of roots greater than 2 mm was measured directly. Aboveground components were separated into leaflets, petiole + rachis, or stem sections. Each root compartment was separated into coarse roots, coralloid roots, or fine roots. Each of these sections was dried at 75 C in a convection oven and then dry weight was measured. Specific leaf area was calculated from area and dry weight measurements on 10 leaflets and then total leaf area for each replication was calculated from specific leaf area and total leaflet dry weight. Total plant dry weight was calculated by the sum of all components, and the root/shoot proportion was calculated from the components of belowground and aboveground plant parts. The influence of competitor identity on root development was calculated by dividing the dry weight or length for fine roots in the half-sib compartment into dry weight or length for fine roots in the second compartment. For this variable, a value above 1 indicated greater dry weight or length of roots in the second compartment, whereas a value below 1 indicated greater dry weight or length of roots in the half-sib compartment. All response variables were subjected to one-way analysis of variance with four treatment levels and nine replications.