A Lee lance corporal who died after collapsing on an SAS selection march would have survived if he had been halted at the last checkpoint, an inquest has heard.

Heat illness expert George Havenith also claimed "inadequate" planning for the treatment of casualties contributed to the deaths of James Dunsby, Craig Roberts and Edward Maher.

In detailed evidence to the inquest in Solihull, Professor Havenith said stopping the exercise on July 13 2013 should have been considered as early as 12.14pm.

The inquest has been told that Corporal Dunsby and lance corporals Maher and Roberts, who were wearing GPS trackers, collapsed between 2.16pm and 3.20pm on the Brecon Beacons.

After outlining the effects of heat on military personnel, Mr Havenith said the 16-mile SAS "selection" test should have been subjected to increasingly urgent and dynamic risk assessments due to several men withdrawing with heat-related conditions.

Previous witnesses have described how two soldiers - known by the codenames 2J and 2P - were withdrawn with heat illness at 12.14pm and 12.46pm.

Commenting on procedures contained in Army guidelines, Mr Havenith told the inquest the withdrawal of Soldier 2J should have led commanders to take a Wet Bulb Globe Test temperature reading and consider stopping the exercise.

Mr Havenith, a professor of environmental physiology and ergonomics at Loughborough University, added: "With the second one, this again, with more urgency, should have led to a dynamic risk assessment.

"Stopping should have been urgently considered."

Giving his view of the withdrawal of Soldier 4E, who was hallucinating due to the effects of heat, at 4.26pm, Mr Havenith added: "The exercise should have been stopped before the third occurrence of heat illness.

"You should have a heightened level of alert if you see things that are unusual compared to a normal exercise."

Addressing each of the soldiers who died individually, Mr Havenith said the "expected outcome" for L/Cpl Roberts and L/Cpl Maher would have been survival, had they been halted and monitored at the route's checkpoint 5.

L/Cpl Maher, who was born in Hampshire, and L/Cpl Roberts, who lived in London but was originally from north Wales, both faced a steep incline after leaving the checkpoint.

In a report presented to the inquest, Mr Havenith said L/Cpl Roberts passed through checkpoint five at 1.14pm and then underwent 70 minutes of further activity, including a very steep ascent.

Mr Havenith said of the reservist: "Before the steep climb it is highly unlikely that the heat illness he suffered from was in an advanced state."

Under questioning from the MoD's barrister, Jonathan Hall, Mr Havenith said there was a good chance that L/Cpl Roberts and L/Cpl Maher did not exhibit clearly recognisable signs of heat exhaustion at the checkpoint.

The inquest heard that the purpose of the near-nine hour test march - carrying a 49lb back-pack and a 5lb gun - was to "drive people to the limit" as part of a selection process.

Met Office readings taken in the village of Libanus, near the march area, showed temperatures peaking at 28.1 Celsius on July 13.

A Wet Bulb Globe Temperature reading of 31.2 degrees WBGT was also recorded at noon on July 13 at Sennybridge Camp, where the reservists taking part in the march were based.

An Army cadet group at Sennybridge asked for the WBGT reading, but it was not requested by those in command of the SAS "test week" march.

Mr Havenith told the inquest the WBGT reading at Sennybridge, which factored in humidity, meant conditions at the camp were seven degrees above an official MoD threshold level governing activities in hot conditions.

The expert added: "A value of seven degrees above the limit, I think it should be clear to anyone that that's a huge excess of the limit."