Appendices A, B, C, are only available in electronic form at http://www.aanda.org

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

In XEST, there are only five objects with a M6V spectral type (namely, V410 X-ray 3, MHO 5, MHO 8, KPNO-Tau 14, and CFHT-Tau 12), which have luminosities higher than 0.03 $L_\odot$ , placing them above the stellar/substellar boundary (Baraffe et al. 1998).

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... candidate

V410 Anon 13, V410 X-ray 3, and MHO 5, which were considered as BD candidates detected in X-rays by ROSAT (Neuhäuser et al. 1999), have now more reliable spectral types (M5.75V, M6V, and M6V, respectively; Briceño et al. 2002), which combined with their luminosities place them above the stellar/substellar boundary. Therefore these very low-mass stars were not included in our sample of BDs. However, they were detected in the XEST and will hence be used in Sect. 5 as TMC low-mass stars detected in X-rays.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

...Hong et al. (2004)

The version 1.7 of the IDL and Perl quantile analysis softwares of Hong et al. (2004) is available at http://hea-www.harvard.edu/ChaMPlane/quantile

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... SHERPA

SHERPA is a part of the CIAO package.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... BD

The BD 2MASS J0414 is located on the highly structured PSF wings of the bright X-ray source V773 Tau, which are not well-modeled, and consequently the contribution of V773 Tau in the extraction region of this BD is not well estimated. Therefore, we assumed that the number of counts observed in the BD extraction region ( $6\hbox{$^{\prime\prime}$ }$ -radius here instead of $10\hbox{$^{\prime\prime}$ }$ -radius) contains a negligible number of counts from the BD, and so the number of observed counts is identical to the number expected from background alone.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... luminosities

For the low-mass (proto)stars, we use the X-ray luminosities in the 0.3-10 keV energy band derived from the DEM method in Güdel et al. (2007). For a BD plasma with 0.3 times the solar elemental abundances and a typical temperature of 1 keV, the difference of X-ray luminosity in the 0.3-10 keV energy band and in the 0.5-8 keV energy band is only 0.06 dex, and hence can be neglected.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... respectively

To compute ratio errors we considered that the number of X-ray detected BDs and the number of (non)accreting BDs are not exactly known, and we combined both Poissonian errors using Gaussian propagation, i.e. $\Delta(a/b)=a/b \times \sqrt{(\Delta{a}/a)^2+(\Delta{b}/b)^2}$ with $\Delta{a}=\sqrt{a}$ and $\Delta{b}=\sqrt{b}$ .

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... ASURV

Namely the Gehan and Peto-Peto generalized Wilcoxon tests, and the Logrank test, which are standard methods of univariate survival analysis as described by Feigelson & Nelson (1985).

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... BDs

We suppressed from the young BD sample of Preibisch et al. (2005b) COUP 344 = HC 722, which was identified as a foreground (old) dwarf object in Slesnick et al. (2005)'s erratum. This object, detected in COUP only during an X-ray flare, was also discussed in Preibisch et al. (2005b) but kept in the reference sample.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... version 3.67

Descriptions and codes for CTI correction and acis_extract can be found at http://www.astro.psu.edu/users/townsley/cti and http://www.astro.psu.edu/xray/docs/TARA, respectively.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.