﻿Temochloa (Poaceae, Bambusoideae), a newly-recorded bamboo genus for China and Vietnam, with new taxa and a re-interpretation of flowering structures

﻿Abstract Neomicrocalamus and Temochloa are closely-related genera for which ‘inflorescence’ structures were incompletely understood and difficult to reconcile. After re-examining the inflorescence morphology, the so-called ‘spikelets’ of both genera as described should instead be recognised as pseudospikelets with mostly inactive axillary buds. The new bamboo taxa, comprising two varieties of a new species, are placed in Temochloa, representing a new genus record for China and Vietnam.

When Keng (1983) established the genus Neomicrocalamus, he considered that the 'inflorescence' is semelauctant (determinate, characterised by true spikelets), meaning it completes its development within a single grand period of growth (McClure 1966).However, at the same time, he considered that the inflorescence basic unit is a pseudospikelet (typically with basal bracts subtending buds that will produce similarly branching lateral units, these then repeating the process, giving what has been called an iterauctant or indeterminate development; McClure (1934McClure ( , 1966)) Stapleton (1994) held the same opinion as Keng.These descriptions are very confusing, as the semelauctant and iterauctant conditions would seem to be fundamentally contrasting.In the protologue of Temochloa, it was diagnosed as having a determinate 'inflorescence with bracteate and prophyllate branches' (Dransfield 2000).
With some newly-collected flowering material from China and Vietnam, it becomes possible for us to re-examine the 'inflorescence' structure of Neomicrocalamus and Temochloa.Furthermore, during the examination, we recognised that some taxa are new to science.These taxa are described and illustrated here.

Materials and methods
Flowering material was dissected under a stereomicroscope (Mshot-MZ101) and images were taken with the camera attachment (Mshot-MSX2).Morphological comparisons and descriptions were based on the relevant literature including protologues, as well as herbarium specimens and living plants.

Results
The newly-discovered bamboo plants are characterised by short-necked pachymorph rhizomes, scrambling culms, solitary and almost circular primary branch buds, branch complement with many short and subequal branches with an occasional dominant central branch that reiterates and approaches the size of the culm, pseudospikelets with 2-4 fertile florets; 6 stamens with emarginate anther apices, 3 stigmas and caryopses.

Flowering structures in the three groups
The pseudospikelet should be interpreted as a condensed flowering branch which terminates in a spikelet proper and is basally supplied with a prophyll and bracts (reduced sheaths) subtending buds (McClure 1934(McClure , 1966)).Typically, these subtended buds are able to develop as secondary pseudospikelets, which in turn lead to higher-order lateral pseudospikelets.This is termed an iterauctant or indeterminate development (McClure 1966).In contrast, a semelauctant or determinate development refers to a single episode of development of the flowering branch, conventionally associated with true spikelets (McClure 1966).However, it should be noted that McClure (1966) mentioned that the axillary buds of primary pseudospikelets of Arundinaria prainii Gamble, equivalent to Neomicrocalamus prainii (Gamble) Keng f., may remain dormant.Perhaps because of this, Keng (1983) employed the term 'semelauctant' for the 'inflorescence' of Neomicrocalamus.Afterwards, in the Flora Reipublicae Popularis Sinicae, inflorescence development was not discussed, but it was stated that, although axillary buds of the pseudospikelet type are present, they have never been known to develop as secondary pseudospikelets (Geng and Wang 1996).Probably for much the same reason, the 'inflorescence' of Temochloa was also regarded as being of a determinate nature, but its description in the protologue states it is "usually comprising one spikelet, in inflorescences with more than one spikelet . . . the axis [is]. . .terminated by a spikelet and a single branch . . .borne at each node, subtended by a bract/sheath and a prophyll . .." (Dransfield 2000).It should be noted that Dransfield (2000) avoided using the term 'semelauctant' in her very careful description.
In fact, according to our observation, pseudospikelets of Neomicrocalamus do produce secondary pseudospikelets, although it happens only occasionally (Fig. 1A).Therefore, the pseudospikelet development of Neomicrocalamus is still iterauctant rather than semelauctant, the latter conventionally linked to true spikelets.It would be confusing if the semelauctant condition is also said to develop pseudospikelets.Hence, the 'spikelets' of both Neomicrocalamus and Temochloa could be properly recognised as a variant (possibly extreme reduction) of the typical pseudospikelet structure, in which axillary buds are not as active as in the typical condition.

The new species of Temochloa
The morphological characters of our newly-discovered bamboos are more similar to Temochloa.The primary branch buds of both the newly-discovered bamboos and Temochloa are nearly circular and their culm leaf sheaths are shallowly grooved, whereas Neomicrocalamus has lanceolate buds and plane culm leaf sheaths (Dransfield 2000;Li and Stapleton 2006).Besides, the anther apices of the newly-discovered bamboos are emarginate (Fig. 1L, left), while those of Neomicrocalamus are conspicuously cuspidate due to prolongation of the connective (Fig. 1L, right).However, to date, some characters, including stamens, of Temochloa are still unknown.
On the other hand, the newly-discovered bamboos and Temochloa occur at very low elevations, 50-250 m, rarely reaching 700 m, whereas Neomicrocalamus taxa have hitherto only been found above 1000 m (Dransfield 2000;Li and Stapleton 2006).
The similar morphology and distribution, in terms of elevation, of the newly-discovered bamboos and Temochloa are also commensurate with their close phylogenetic affinities.The phylogenetic evidence indicates that the newly-discovered bamboos originate from introgressive hybridisation between Temochloa liliana S. Dransf.and N. prainii (Cai et al. 2023; the newly-discovered bamboo accessions referred as BH85, 2018VNB018 and 2018VNB040).The newly-discovered bamboos inherited most (80.7%) of its genome from T. liliana; therefore, genetically, these two groups are closer related (Cai et al. 2023).
Phenology.New shoots around May.
Etymology.The specific epithet refers to its elegant habit.Chinese name.雅竹 (yǎ zhú).Distribution and habitat.This species occurs in the limestone area of southwest Guangxi, China and northeast Vietnam, at elevations of 210(-700) m.
Conservation status.Up to now, T. elegans is known from only two locations in China and Vietnam.It is not very common at those locations so the number of mature clumps appears to be limited.The Vietnamese population is well protected in the Nature Reserve, while the Chinese population is distributed along

Figure 1 .
Figure 1.Floral morphology of Temochloa elegans A pseudospikelets of Neomicrocalamus sp.showing the secondary pseudospikelet (left) developed from the base of the primary pseudospikelet (right) B rachilla segment C prophyll (abaxial view) D bracts (abaxial view) E lemma (abaxial view) F palea (left: back view; right: side view) G lodicules H pistil I young fruit subtended by a palea J mature fruit K apical not fully developed floret L anther apices (left from Temochloa elegans; right from Neomicrocalamus sp.included for comparison) M anthers.

Figure 2 .
Figure 2. Morphology of Temochloa elegans A habit B foliage leaves C branch complement without a dominant central branch D branch complement with a developing dominant central branch E primary branch bud F culm internode section G culm leaf (upper half) H culm leaf (lower half) I flowering branches J spikelet proper.